National Academies Press: OpenBook
« Previous: Part 1 - Guidebook
Page 76
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 76
Page 77
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 77
Page 78
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 78
Page 79
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 79
Page 80
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 80
Page 81
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 81
Page 82
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 82
Page 83
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 83
Page 84
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 84
Page 85
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 85
Page 86
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 86
Page 87
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 87
Page 88
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 88
Page 89
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 89
Page 90
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 90
Page 91
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 91
Page 92
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 92
Page 93
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 93
Page 94
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 94
Page 95
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 95
Page 96
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 96
Page 97
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 97
Page 98
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 98
Page 99
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 99
Page 100
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 100
Page 101
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 101
Page 102
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 102
Page 103
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 103
Page 104
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 104
Page 105
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 105
Page 106
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 106
Page 107
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 107
Page 108
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 108
Page 109
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 109
Page 110
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 110
Page 111
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 111
Page 112
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 112
Page 113
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 113
Page 114
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 114
Page 115
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 115
Page 116
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 116
Page 117
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 117
Page 118
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 118
Page 119
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 119
Page 120
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 120
Page 121
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 121
Page 122
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 122
Page 123
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 123
Page 124
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 124
Page 125
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 125
Page 126
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 126
Page 127
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 127
Page 128
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 128
Page 129
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 129
Page 130
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 130
Page 131
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 131
Page 132
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 132
Page 133
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 133
Page 134
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 134
Page 135
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 135
Page 136
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 136
Page 137
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 137
Page 138
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 138
Page 139
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 139
Page 140
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 140
Page 141
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 141
Page 142
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 142
Page 143
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 143
Page 144
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 144
Page 145
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 145
Page 146
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 146
Page 147
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 147
Page 148
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 148
Page 149
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 149
Page 150
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 150
Page 151
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 151
Page 152
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 152
Page 153
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 153
Page 154
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 154
Page 155
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 155
Page 156
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 156
Page 157
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 157
Page 158
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 158
Page 159
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 159
Page 160
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 160
Page 161
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 161
Page 162
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 162
Page 163
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 163
Page 164
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 164
Page 165
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 165
Page 166
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 166
Page 167
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 167
Page 168
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 168
Page 169
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 169
Page 170
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 170
Page 171
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 171
Page 172
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 172
Page 173
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 173
Page 174
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 174
Page 175
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 175
Page 176
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 176
Page 177
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 177
Page 178
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 178
Page 179
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 179
Page 180
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 180
Page 181
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 181
Page 182
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 182
Page 183
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 183
Page 184
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 184
Page 185
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 185
Page 186
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 186
Page 187
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 187
Page 188
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 188
Page 189
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 189
Page 190
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 190
Page 191
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 191
Page 192
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 192
Page 193
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 193
Page 194
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 194
Page 195
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 195
Page 196
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 196
Page 197
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 197
Page 198
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 198
Page 199
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 199
Page 200
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 200
Page 201
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 201
Page 202
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 202
Page 203
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 203
Page 204
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 204
Page 205
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 205
Page 206
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 206
Page 207
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 207
Page 208
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 208
Page 209
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 209
Page 210
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 210
Page 211
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 211
Page 212
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 212
Page 213
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 213
Page 214
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 214
Page 215
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 215
Page 216
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 216
Page 217
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 217
Page 218
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 218
Page 219
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 219
Page 220
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 220
Page 221
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 221
Page 222
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 222
Page 223
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 223
Page 224
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 224
Page 225
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 225
Page 226
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 226
Page 227
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 227
Page 228
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 228
Page 229
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 229
Page 230
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 230
Page 231
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 231
Page 232
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 232
Page 233
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 233
Page 234
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 234
Page 235
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 235
Page 236
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 236
Page 237
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 237
Page 238
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 238
Page 239
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 239
Page 240
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 240
Page 241
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 241
Page 242
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 242
Page 243
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 243
Page 244
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 244
Page 245
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 245
Page 246
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 246
Page 247
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 247
Page 248
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 248
Page 249
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 249
Page 250
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 250
Page 251
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 251
Page 252
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 252
Page 253
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 253
Page 254
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 254
Page 255
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 255
Page 256
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 256
Page 257
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 257
Page 258
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 258
Page 259
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 259
Page 260
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 260
Page 261
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 261
Page 262
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 262
Page 263
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 263
Page 264
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 264
Page 265
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 265
Page 266
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 266
Page 267
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 267
Page 268
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 268
Page 269
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 269
Page 270
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 270
Page 271
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 271
Page 272
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 272
Page 273
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 273
Page 274
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 274
Page 275
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 275
Page 276
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 276
Page 277
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 277
Page 278
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 278
Page 279
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 279
Page 280
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 280
Page 281
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 281
Page 282
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 282
Page 283
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 283
Page 284
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 284
Page 285
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 285
Page 286
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 286
Page 287
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 287
Page 288
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 288
Page 289
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 289
Page 290
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 290
Page 291
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 291
Page 292
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 292
Page 293
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 293
Page 294
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 294
Page 295
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 295
Page 296
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 296
Page 297
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 297
Page 298
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 298
Page 299
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 299
Page 300
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 300
Page 301
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 301
Page 302
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 302
Page 303
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 303
Page 304
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 304
Page 305
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 305
Page 306
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 306
Page 307
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 307
Page 308
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 308
Page 309
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 309
Page 310
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 310
Page 311
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 311
Page 312
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 312
Page 313
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 313
Page 314
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 314
Page 315
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 315
Page 316
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 316
Page 317
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 317
Page 318
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 318
Page 319
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 319
Page 320
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 320
Page 321
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 321
Page 322
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 322
Page 323
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 323
Page 324
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 324
Page 325
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 325
Page 326
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 326
Page 327
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 327
Page 328
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 328
Page 329
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 329
Page 330
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 330
Page 331
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 331
Page 332
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 332
Page 333
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 333
Page 334
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 334
Page 335
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 335
Page 336
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 336
Page 337
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 337
Page 338
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 338
Page 339
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 339
Page 340
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 340
Page 341
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 341
Page 342
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 342
Page 343
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 343
Page 344
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 344
Page 345
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 345
Page 346
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 346
Page 347
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 347
Page 348
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 348
Page 349
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 349
Page 350
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 350
Page 351
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 351
Page 352
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 352
Page 353
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 353
Page 354
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 354
Page 355
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 355
Page 356
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 356
Page 357
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 357
Page 358
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 358
Page 359
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 359
Page 360
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 360
Page 361
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 361
Page 362
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 362
Page 363
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 363
Page 364
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 364
Page 365
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 365
Page 366
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 366
Page 367
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 367
Page 368
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 368
Page 369
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 369
Page 370
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 370
Page 371
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 371
Page 372
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 372
Page 373
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 373
Page 374
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 374
Page 375
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 375
Page 376
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 376
Page 377
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 377
Page 378
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 378
Page 379
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 379
Page 380
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 380
Page 381
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 381
Page 382
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 382
Page 383
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 383
Page 384
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 384
Page 385
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 385
Page 386
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 386
Page 387
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 387
Page 388
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 388
Page 389
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 389
Page 390
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 390
Page 391
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 391
Page 392
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 392
Page 393
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 393
Page 394
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 394
Page 395
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 395
Page 396
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 396
Page 397
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 397
Page 398
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 398
Page 399
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 399
Page 400
Suggested Citation:"Part 2 - Toolbox ." National Academies of Sciences, Engineering, and Medicine. 2018. Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox. Washington, DC: The National Academies Press. doi: 10.17226/24991.
×
Page 400

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

P a r t 2 Toolbox

Tools s e c t i o n 1

C o n t e n t s 83 Synopses of Tools 88 Tool 1 Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 102 Tool 2 Using Public Use Microdata Samples to Profile Transportation Characteristics and Differences 110 Tool 3 Using the National Household Travel Survey to Profile Transportation Characteristics and Differences 113 Tool 4 Preparing, Implementing, and Assessing a Public Involvement Plan 140 Tool 5 Using Focus Groups in Assessing the Impact of Tolling on Environmental Justice Populations 151 Tool 6 Designing and Executing Surveys to Assess Attitudes and Travel Behavior for Environmental Justice Analyses and to Monitor Implementation 178 Tool 7 Using Travel Demand Models for Environmental Justice Assessments 201 Tool 8 Applying a Select Link Analysis to Assess Trip Patterns 211 Tool 9 Analyzing the Value of Time/Willingness to Pay in Environmental Justice Assessments 220 Tool 10 Assessing User Costs and Household Burden Effects 228 Tool 11 Evaluating Disproportionate Effects with Quantitative Methods 235 Tool 12 Instituting Cash Replenishment Options for Unbanked and Underbanked Populations 241 Tool 13 Recycling Tolling Revenue through Transit Investment and Low-Income Assistance as Forms of Mitigation 253 Tool 14 Examining Spatial Patterns and Distribution of Users on Existing Tolling Facilities

83 Synopses of Tools Developing a Socioeconomic Profile and Community Characteristics Inventory for EJ Assessments A demographic profile of social and economic characteristics such as income, race and ethnicity, disability, age, limited English proficiency, educational attainment, time leaving home for work, and “zero-car” households provides important building blocks about communities for impact assessment. Creating a profile contributes to developing a substantive understanding of the needs of the affected populations, including low-income and minority populations and other traditionally underserved populations. Through the detailed social and economic profile, the practitioner can begin to consider the unique challenges or barriers to participation likely to be faced in trying to engage various population segments in specific communities or region wide. It is a foundational tool for the development and imple- mentation of a thoughtful and inclusive public involvement plan. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement Using Public Use Microdata Samples to Profile Transportation Characteristics and Differences The Public Use Microdata Sample (PUMS) are computerized files that contain a sample of individual records from the U.S. Census and the American Community Survey, with identifying information removed, that show the population and housing characteristics of the units and people represented on the data records. The sample data supports the assessment of transportation characteristics, including mode of travel to/from work; travel time; vehicle ownership; and assessment of how travel behavior varies by race, income, and demographic categories. Using PUMS allows for a uniquely detailed analysis of household, income, and transportation characteristics at varying geographic scales with a high degree of precision. This dataset is very well-suited for examining the differences in journey-to-work and vehicle ownership character- istics by income levels or racial categories, which can contribute to a better understanding of the financial and mobility challenges within a state or metropolitan area experienced by households and individuals—an important context for considering potential effects of toll road pricing. Toolbox Step: Step 4—Scope Approach to Measure and Assess Impacts Using the National Household Travel Survey to Profile Transportation Characteristics and Differences The National Household Travel Survey (NHTS) is a nationally representative travel diary survey containing information on more than 150,000 U.S. households, the characteris- tics of the households’ members, details of the trips they take on a survey day, as well

84 Assessing the environmental Justice effects of toll Implementation or Rate Changes: Guidebook and toolbox as the characteristics of the vehicles they own and their attitudes toward several transportation issues. Using NHTS allows for a very detailed analysis of households’ travel patterns among population subgroups and across a variety of geographic characteristics. For instance, research- ers can examine the differences in transportation choices between low- and higher-income households and across racial and ethnic groups. NHTS contains some information on transpor- tation expenditures, including tolls paid and examines various trip purposes in addition to the work trip. Toolbox Step: Step 4—Scope Approach to Measure and Assess Impacts Preparing, Implementing, and Assessing a Public Involvement Plan Developing a public involvement plan (PIP) is an important step in identifying the needs and concerns of the public, including low-income and minority populations in transportation decision-making processes. PIP serves as a procedural guide for agencies and practitioners that describes key considerations and strategies for carrying out public participation activities. PIP should guide all stages of trans- portation decision-making, but is particularly relevant in statewide and metropolitan planning, project design, and environmental review stages of decision-making. The essence of effective environmental justice (EJ) practices has been distilled by the U.S. DOT and FHWA in the Fundamental Principles of Environmental Justice. One of the practices is to ensure the full and fair participation by all potentially affected communities in the trans- portation decision-making process. PIP offers a platform for clearly articulating the agency and practitioner’s planned approaches for ensuring the full and fair participation of low-income and minority populations and other populations on toll-related studies that are likely to be effective in overcoming barriers to participation. Toolbox Steps: Step 3—Recognize the Relevant Decision-Makers and Stakeholders, Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies, Step 7—Document Results for Decision-Makers and the Public, Step 8—Conduct Post-Implementation Monitoring Using Focus Groups in Assessing the Impact of Tolling on EJ Populations Focus groups are a qualitative research method involving small group dis- cussions led by a trained moderator (also referred to as a facilitator). The qualitative nature of focus groups comes from the data typically being col- lected in the form of “why” behind people’s thoughts, feelings, and behaviors, rather than the quantitative number of people who think, feel, or behave in a particular way. Assessment of impacts of tolling on low-income and minority populations as part of an EJ analysis requires not only valid information on the number of people affected but also an understanding of the essence of those impacts and what meaningful mitigation would be for those affected. Consequently, focus groups are often used in conjunction with quantitative survey research. Toolbox Steps: Step 4—Scope Approach to Measure and Assess Impacts, Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post-Implementation Monitoring Designing and Executing Surveys to Assess Attitudes and Travel Behavior for EJ Analyses and to Monitor Implementation Surveys are a common means of assessing public attitudes toward tolling facilities, travel behavior, and a willingness to pay for the use of managed lanes and tolling facilities. Properly designed surveys can be used to examine how tolling solutions are perceived in terms of fairness and how they may affect low-income and

synopses of tools 85 minority travelers compared to other populations. This information is important to gather to determine potential EJ impacts and the effectiveness of mitigation strategies. Specific consider- ations should be taken into account when designing and implementing surveys to be responsive to the protected population focus of EJ. To support this process, this tool inventories topics and questions that have been typically used to conduct travel behavior surveys on toll-related projects—whether implemented before or after or with panels maintained in both time periods— as well as survey collection methods and findings that have resulted from their implementation. Toolbox Steps: Step 4—Scope Approach to Measure and Assess Impacts, Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post-Implementation Monitoring Using Travel Demand Models for EJ Assessments Travel demand models are used to forecast and simulate vehicle and person trips on the highway and multimodal transportation systems. TDMs can estimate the quality of the level of flow (travel time, speed, levels of service for regional transportation systems, and individual links in the network). In a tolling context, TDMs are used to assess the impact of tolling projects to capture the systemwide effects of travelers switching to other routes to avoid tolls and to forecast usage of the toll facility under different growth and revenue sce- narios. Revenue forecasts depend on travel demand forecasts and their underlying assumptions. With regard to tolls, TDMs generally model traveler responses to pricing, which can shed light on how various populations may be affected by the pricing change. An understanding of EJ considerations should inform methods taken by travel demand modelers at many of the decision- making stages associated with pricing projects. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post-Implementation Monitoring Applying a Select Link Analysis to Assess Trip Patterns A select link analysis is a travel demand procedure that shows where trips that traverse a selected link start and end. The beginning and end points of the trips are displayed in various formats, including volumes and percentages. The displays could be shown by bandwidth, color, or numerically. Select link analysis is effective for EJ analysis because it allows the user to apply available out- put of travel demand models to assess and compare the impacts that toll facilities or similar types of transportation improvements will have on travel time and accessibility measures. Once the zones that contain high concentrations of low-income or minority populations are identified, select link analyses can be set up and quickly analyzed to compare scenarios with and without toll facilities in the network. Toolbox Step: Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post-Implementation Monitoring Analyzing the Value of Time/Willingness to Pay in EJ Assessments The value of time (VOT) and its measurement is a key factor and variable in trans- portation planning, project evaluation and asset calculation, and performance. VOT can be influenced by several factors, including trip purpose, trip timing, trip urgency, work schedule flexibility, and sociodemographic characteristics, among other factors. VOT and its accurate measurement can inform an assessment of EJ effects through its close consideration of the factors that influence potential differences in travel behavior. For EJ, VOT by income

86 Assessing the environmental Justice effects of toll Implementation or Rate Changes: Guidebook and toolbox distribution and trip purpose are important measures to estimate within a region or corridor. VOT can be used to inform the development of toll pricing and account policies that are sensi- tive to the behavioral differences of low-income travelers. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post-Implementation Monitoring Assessing User Costs and Household Burden Effects User cost assessments examine the economic or financial impact on households of various transportation options, including toll road pricing and their “burden” effects on household budgets. The approach estimates the cost of using the tolling facility for low-income and non-low-income households, compares how these costs may differ, and how they may be experienced differently within and between various household income levels. The frame of analysis considers possible regressive effects of travel alternatives (including but not limited to toll road pricing) for individual and/or groups of low-income households. Addition- ally, as transportation costs rise, a potential risk of social exclusion exists for low-income house- holds as a consequence of tolling. The outcome of social exclusion is that affected individuals or groups are precluded from participating fully in the economic, social, and civic activities of the society in which they live. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement, Step 6—Identify and Assess Mitigation Strategies Evaluating Disproportionate Effects with Quantitative Methods Although Executive Order 12898 directs federal agencies to develop a strategy to identify and address the disproportionately high and adverse human health or envi- ronmental effects of their actions on low-income and minority populations, no guid- ance is provided as to the criteria for the determination of “disproportionately high and adverse” effects. The Council on Environmental Quality’s 1997 guidance on EJ is purposefully vague to allow agencies to integrate analyses of EJ concerns in an appropriate manner. The Texas Depart- ment of Transportation (DOT) and the Washington State DOT have developed quantitative methods to evaluate disproportionately high and adverse effects. Texas DOT’s methods for qualitatively evaluating disproportionately high and adverse effects include statistical analyses to determine if (1) the measured impacts associated with the tolled alternative are statistically sig- nificantly higher than the measured impacts associated with the non-tolled alternative and (2) the impact is statistically significantly higher in zones with greater concentrations of low-income or minority populations. Washington State DOT’s method considers percentage differences for effects rather than statistical tests to determine disproportionate impacts. The two examples briefly describe analytical steps and criteria used to make a determination of disproportionately high and adverse effects on low-income and minority populations. The methods and potential limitations of using solely quantitative methods to make a determination are discussed. Toolbox Steps: Step 6—Identify and Assess Mitigation Strategies Instituting Cash Replenishment Options for Unbanked and Underbanked Populations Cash replenishment options address the possible barrier to use all-electronic tolling facilities because of the absence of credit or debit cards to purchase or make a needed deposit to acquire a transponder or to replenish an account. One of the key consid- erations for addressing EJ considerations with toll implementation is equitable access to the toll facility. As agencies adopt all-electronic tolling systems, individuals without a credit card or bank

synopses of tools 87 account face the prospect of being excluded or financially penalized for their use. Populations in this circumstance may experience a form of social exclusion (i.e., compromised access to jobs and other amenities critical to sustaining a household’s quality of life). Toolbox Step: Step 6—Identify and Assess Mitigation Strategies Recycling Tolling Revenue through Transit Investment and Low-Income Assistance as Forms of Mitigation Federal law requires toll revenues to be used first to cover the costs of developing, operating, and maintaining the toll facility and providing a return on investment to any private investment partner. Excess revenues may be used for any transportation purpose within the subject corridor for which federal funds would normally be used. Toll pay- ment assistance or infrastructure and transit advancements that benefit EJ communities, a tech- nique known as revenue recycling, are ways to mitigate adverse financial and physical effects of the tolling projects. Recycling toll revenues into transportation infrastructure and services can increase passenger throughput in the corridor and potentially benefit all users, not just those who can afford to pay the fee as single-occupancy drivers, particularly if there is a non-tolled general purpose lane. Subsidizing public transportation services or its needed infrastructure and equipment can offer an option for low-income populations seeking mobility and access along tolled corridors. Dedicating a portion of toll revenues to finance transportation improvements, such as more frequent transit services or vanpools, may offer benefits that offset adverse impacts of tolling and toll facilities for low-income travelers. Toolbox Steps: Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post- Implementation Monitoring Examining Spatial Patterns and Distribution of Users on Existing Tolling Facilities Public records of usage patterns and the income distribution of users of existing tolled facilities can be studied to analyze the distribution of tolling burdens and benefits by income level and geographic area. Although using public tolling data from agencies is possible to analyze the equity and EJ effects of various tolling systems, the practice is not as common as tools that often rely on opinions solicited in surveys or drawn from transportation modeling output. This tool presents case examples to show how available records data can be used to analyze usage patterns, the demographics of users, and the methods and measures used to assess disparities in usage. The research may be used to monitor the equity dimension of existing operations or advocate for mitigation solutions to ensure mobility and access for low-income or other disadvantaged populations. Toolbox Step: Step 8—Conduct Post-Implementation Monitoring

88 What Is It? Demographic profiles of social and economic charac- teristics such as income, race and ethnicity, disability, age, limited English proficiency, educational attainment, time leaving home for work, and “zero-car” households provide important building blocks about communities for impact assessment. Why Is It Effective in Environmental Justice Analysis? Creating a profile contributes to developing a substan- tive understanding of the needs of the affected popula- tions, including low-income and minority populations and other traditionally underserved populations. While minority and low-income populations, in accor- dance with the Environmental Justice (EJ) Executive Order (EO) 12898 and subsequent U.S. DOT Order 5610.2(a) and FHWA Order 6640.23A on Environmental Justice, represent the formal demographic focus of community identification for EJ analyses, there is good reason—in recognition of the definitional overlap between EJ and Title VI—to look more broadly at how projects affect other traditionally under- served populations, including persons with limited English proficiency, foreign-born populations, and persons with disabilities and their transportation needs. Through a detailed social and economic profile, the prac- titioner can begin to assess the unique challenges or barriers to participation likely to be faced in trying to engage various population segments in specific communities or regionwide. It is a foundational tool for the development and implemen- tation of a thoughtful and inclusive PIP. It is a pre-requisite element for preparing plans for working with low-literacy and limited English proficiency populations. t o o l 1 Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • U.S. Census Bureau’s American FactFinder and American Community Survey • EPA Environmental Justice Screen • U.S. Housing and Urban Development AFFH Data and Mapping Tool • Threshold Analysis • Environmental Justice Index Method Affected Populations • Low-Income • Minority • Limited English Proficiency • Foreign-Born Populations • Persons with Disabilities Examples Featured • Metropolitan Transportation Commission, San Francisco, I-680 Corridor • Washington State DOT, SR 520 Bridge Replacement • North Central Texas Council of Governments, EJ Index

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 89 What Are Some Techniques for Implementing This Tool? There are many data sources and methods for conducting a preliminary screening to deter- mine whether minority or low-income populations may be affected by the tolling project and to inform the preparation of a detailed social and economic profile. Defining Study Area Boundaries. The location and potential function of the tolled or to- be-tolled facility in making connections to opportunities—employment centers, education, health care, shopping—should be considered in defining initial study area boundaries. The initial boundary of a study area for investigating the connections among these features should be relatively large so as not to unintentionally exclude important connections from the analysis. Among the considerations for setting the boundary are: • the location of the tolled or to-be-tolled facility; • the local areas served by the facility and its connecting roads; • the location of minority and low-income communities having access to the facility via con- necting roads and feeding network of roadways; and • the location of opportunity destinations served by the facility and its connecting roadway network. The extent to which a connecting road or roads is itself tolled should be identified so that the cumulative impact of the subject tolling action with other past, present, and reasonably foresee- able tolling actions can be considered. Sources of information on these features typically include U.S. Census products, geographic information system (GIS) data layers, regional travel models, map software, and interviews with relevant specialists, for example, municipal planners, department of health representatives, parks and recreation department representatives, etc. One particularly useful U.S. Census product for establishing the context and boundary of the study area is the American Community Survey (Journey-to-Work) Worker Flow datasets which can be helpful in seeing typical commuting patterns within a metropolitan area. Data Tools and Sources. The U.S. Census Bureau’s American FactFinder and American Community Survey (ACS) provide important sources of demographic information for this profile (see text boxes, Using the U.S. Census Bureau’s American Community Survey and Using the U.S. Census Bureau’s American FactFinder). Table 1 identifies a set of demographic variables from the ACS, including relevant ACS data tables, example metrics, and the most detailed small area geographies reported for each of the data variables. While the U.S. Census, particularly the ACS, are valuable data sources for the characterization of populations, other data sources may also be consulted to inform the identification of popula- tions in an affected environment and to prepare the PIPs and impact assessments. For example, local housing agencies may be able to supply information on low- and moderate-income afford- able and senior citizen housing projects. Workforce development agencies, vocational training institutions, social service providers, and community-based or advocacy-based organizations may be able to share data on client needs, human services transportation providers, or small clusters or pockets of populations that may not be revealed through census data. The U.S. Depart- ment of Education’s National Center for Education Statistics and state educational departments may have data on student enrollment and participation in the free and reduced price lunch pro- grams that can be a useful supplement to census materials on race and income in various schools and school districts.

90 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Using the U.S. Census Bureau’s American Community Survey The American Community Survey (ACS), a product of the U.S. Census Bureau, pro- vides current demographic, social, economic, and housing information about the country’s communities. ACS collects and produces population and housing data drawn from surveys every year instead of every 10 years through the U.S. Census Bureau. Approximately 3 million households participate in the ACS annually. Source. ACS is published by the U.S. Census Bureau and can be accessed through multiple sources, including American FactFinder, an online clearinghouse main- tained by the U.S. Census Bureau. Geographies. ACS publishes single-year data for all areas with populations of 65,000 or more. Areas with populations less than 65,000 will require the use of multiyear estimates to reach an appropriate sample size for data publication. In 2008, the U.S. Census Bureau began releasing 3-year estimates for areas with populations greater than 20,000. The first 5-year estimates for all census tracts and block groups were released in December 2010, based on data collected from January 1, 2005, to December 31, 2009. With this and subsequent releases, data became available for every state, county, city, town, and place as well as census tracts and block groups. The multiyear estimates will be updated annually, with data published for the largest areas in 1-, 3-, and 5-year formats, and for those meeting the 3-year threshold in both 3- and 5-year formats. Even the lesser populated communities will be able to obtain ACS data based on 5-year estimates annually. Demographic Variables. ACS provides an excellent starting point for accessing census data to profile communities and support demographic mapping. Low- income, minority, limited English proficiency, zero-car, elderly, and disabled persons, among other census-related variables, can be used to reveal social, economic, and travel patterns. Pros/Cons. ACS data is an up-to-date source of population information since it is collected and published every year instead of every 10 years. One- and three- year estimates are generally available with a one to two year lag. For small areas and population groups of 20,000 or less, it takes five years of continued sampling to provide estimates at the local level. The quality of the data at the local level should be closely investigated; many analysts are not familiar with how to use this relatively new but potentially highly valuable and informative data product. ACS data will describe a period of time and require data for 12 months, 36 months, or 60 months to do so. Smaller sample sizes for 5-year ACS estimates will reduce the reliability of estimates. Value. ACS data sets make it possible to prepare a relatively current demographic profile of a community that can be applied to planning or project development studies. The latest release of the 5-year estimates can help identify the presence of minority and limited English proficiency populations, among other tradition- ally underserved populations, at the sub-municipal geographic levels for plan- ning and project development studies, as well as also be referenced for policy research. ACS can generally support profile characterization at both regional and small-area levels.

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 91 Using the U.S. Census Bureau’s American FactFinder American FactFinder is an easy-to-use portal to find population, housing, eco- nomic, and geographic data collected by the U.S. Census Bureau. It is a product of the U. S. Census Bureau and provides access to data collected through the Decen- nial Census, American Community Survey (ACS), Puerto Rico Community Survey, Population Estimates Program, Economic Census, and Annual Economic Surveys. Source. From the 2010 Decennial Census, the total population as well as racial and ethnic data has been compiled in Summary File 1 (SF1) (i.e., data that was col- lected in the short-form census questionnaire distributed to all households). This information is a 100 percent sample of the population. The SF1 file shows detailed tables on age, sex, households, families, relationship to householder, housing units, detailed race and Hispanic or Latino origin groups, and group quarters. Most tables are shown down to the block or census tract level. Some tables are repeated for nine race/Hispanic or Latino origin groups. The nine groups are (1) White alone, (2) Black or African American alone, (3) American Indian and Alaska Native alone, (4) Asian alone, (5) Native Hawaiian and Other Pacific Islander alone, (6) Some Other Race alone, (7) Two or More Races, (8) Hispanic or Latino, and (9) White alone, Not Hispanic or Latino. The long form is generally being discontinued for most U.S. populations (except territories such as the Virgin Islands and Guam) in favor of the ACS. Geographies. National data aggregated at the state, metropolitan/micropolitan statistical area, county, place, census tract, block group, and block level, among others. Demographic Variables. Low-income, minority, linguistic isolation, zero-car, elderly, and disabled populations, among other social, economic, and travel indicators. Pros/Cons. Provides access to multiple data sources and allows users to create and save detailed and custom tables. ACS data is updated with a lag of 1 to 2 years. Requires some familiarity with data contained in census surveys and does not provide flat data files, limiting the utility of the portal data as a source for performing cross-tabulations (such as with PUMS). Value. This website can be used to prepare a demographic profile of a community that can be applied to planning or project development studies. Depending on how long it has been since the release of the last decennial census, it will become increasingly appropriate to use ACS data, also available on the U.S. Census website, to prepare a demographic profile with intercensal survey estimates. In preparing the profile of a community, practitioners should closely consider whether the subject communities currently bear, or have historically borne, environmental stressors that affect community quality of life and public health. Similarly, the cumulative effects of prior land use and zoning policies, disinvestment, and the past siting of locally unwanted facilities can contribute to patterns of racial or ethnic segregation, income segregation, or isolation of com- munities. Several online tools have been developed that can be informative in understanding how subject communities compare with the region on key measures of health, safety, and the

Variable Example Metrics U.S. Census Product Lowest Geographic Level Data Tables Reporting Variables Household and family type Number of family households 2013 ACS 5-year estimates Block Group B11001 Household Type (Including Living Alone) Number of female householders with children, no husband present Mobility Number of workers who drove to work in a car, truck, or van 2013 ACS 5-year estimates Block Group B08301 Means of Transportation toWork. Universe: Workers 16 years and over Number of workers who took public transportation to work (excluding taxicabs) Disability Number of females with a disability that did not work in the past 12 months 2013 ACS 5-year estimates Block Group C2303 Sex by Disability Status by Full-TimeWork Status in the Past 12 Months for the Population 16 to 64 Years Number of females without a disability that did not work in the past 12 months Work status (part-time, full-time) Number of females worked in the past 12 months who usually worked 1 to 14 hours per week for 50 to 52 weeks 2013 ACS 5-year estimates Block Group B23022 Sex by Work status in the Past 12 Months by Usual Hours Worked per Week in the Past 12 Months by Weeks Worked in the Past 12 Months for the Population 16 to 64 Years Number of females worked in the past 12 months who usually worked 35+ hours per week for 50 to 52 weeks Education (school enrollment and educational attainment) Number of persons with no schooling completed 2013 ACS 5-year estimates Block Group B15002 Sex by Educational Attainment for the Population 25 Years and Over Number of persons with master ’s degree Table 1. Use of census data for demographic mapping: race, income, and other social characteristics.

Vehicle availability Owner occupied housing units with no vehicle available householder 15 to 34 years 2013 ACS 5-year estimates Block Group B25045 Tenure by Vehicles Available by Age of Householder Owner occupied housing units with one or more vehicles available householder 35 to 64 years Language (limited English proficiency) Limited English speaking households that speak other languages at home 2013 ACS 5-year estimates Block Group B16002 Household Language by Household Limited English Speaking Status Limited English speaking households that speak Spanish at home Race (race and ethnicity) Population of Non-Hispanic Black or African American People 2013 ACS 5-year estimates Block Group B03002 Hispanic or Latino Origin by Race Population Hispanic or Latino White Alone People Income (percent of poverty level) Population with income 50% of poverty status 2013 ACS 5-year estimates Block Group C17002 Ratio of Income Poverty Level in the Past 12 Months. Universe: Population for whom Poverty Status is Determined Population with income 200% of poverty status Female headed households with children Female householder, no husband present: children under 5 years 2013 ACS 5-year estimates Block Group B09002 Own Children Under 18 Years by Family Type and Age Female householder, husband present: children under 5 years Median household income Median household income in the past 12 months (in 2013 inflation-adjusted dollars) 2013 ACS 5-year estimates Block Group B19013 Median Household Income in the Past 12 Months (In 2013 Inflation-Adjusted Dollars) Foreign-born status Population born in Western Europe 2013 ACS 5-year estimates Census Tract B05006 Place of Birth for the Foreign-Born Population in the United States Population born in Kenya Source: U.S. Census Bureau, 2015

94 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox “geography of opportunity” such as school proficiency and proximity to jobs (see text box, Using Available Tools from EPA and the U.S. Department of Housing and Urban Development). Patterns of disparity may be identified from a review of these tools at an early stage of the study, and the tools can also be a resource for identifying some key assets and resources that may be important to consider in planning for meaningful public involvement processes. Using Available Tools from EPA and the U.S. Department of Housing and Urban Development EJSCREEN: Environmental Justice Screening and Mapping Tool. The EPA EJSCREEN tool can be used for quickly compiling information on community health and environmental stressors. In addition to providing a handful of U.S. Census Bureau American Community Survey (ACS) demographic indicators, the EJSCREEN tool presents data and mapping information on select environmental indicators that examine sources of exposure to envi- ronmental pollutants, such as nearby hazardous waste sites or traffic, ambient levels of air pollutants, such as particulate matter (PM2.5), ozone and diesel particulate matter, or air toxics-related cancer risk or a hazard index, which summarizes the ratios of ambient air toxics levels to health-based reference concentrations. Affirmatively Furthering Fair Housing Data and Mapping Tool. The U.S. Department of Housing and Urban Development (HUD) recently updated its Affirmatively Furthering Fair Housing (AFFH) Regulations, in part, to address issues with fair housing compliance and enforcement. HUD grantees must prepare an Assess- ment of Fair Housing (AFH) to HUD and then incorporate the AFH’s findings into their Public Housing Assessment and Consolidated Plans. All HUD-funded recipients are required to prepare an AFH, including public housing authorities and state and local governments receiving funding from various HUD programs. Grantees are expected to “take steps to address the issues of segregation and related barriers, particularly as reflected in racially and ethnically concentrated areas of poverty.” HUD has created an online AFFH Data and Mapping Tool (AFFH-T) with tables and maps customized to grantee communities. The data and maps explore the relationships between demographics (e.g., race, poverty, public housing programs, limited Eng lish proficiency, and disability) and measures of opportunity such as job proximity, labor market, school proficiency, transportation costs, and environmental health. Identify Spatial Concentrations of EJ Populations. The geographic scale of the units of analysis varies depending on the study purpose (e.g., statewide or metropolitan planning, corridor-level planning, project environmental review, or program evaluation after implemen- tation). Thus, the unit of analysis may range from metropolitan statistical area (MSA)-level or county-level aggregation for regional benchmarking of socioeconomic trends and patterns, to somewhat coarse travel analysis zones for travel demand modeling, to more granular treatments at the tract level, block group, and the block level. The scale of geographic analysis is impor- tant because it will affect how detailed and geographically specific the profile of the potentially affected populations in the impacted areas may be. Minority groups and low-income populations are unlikely to be uniformly distributed over a subject region, but rather may be spatially concentrated—to a greater or lesser extent depending on the region—in specific communities or areas. Therefore, the practitioner should recognize that the more aggregate the level of the analysis, the higher the probability that pockets of low-income and minority populations may be overlooked. Effort should be made to ensure that the selection of geographic boundaries accurately and fairly represents the presence and concerns of minority and low-income persons in the affected communities. Practitioners should also be prepared to adjust their assumptions about geographic boundaries if new information comes to light. In practice, many planning and environmental studies related to tolling effects rely on data- driven “threshold” evaluation methods to facilitate their screening or identification of an

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 95 “EJ community” or “community of concern,” although this approach is not always appropri- ate. Thresholds are the comparison values used to make a determination of whether protected populations are located within a study area. The level of EJ concern can be generally assumed to be higher in places that exceed the threshold value (see text boxes, Identifying Minority Popula- tions and Identifying Low-income Populations). Identifying Minority Populations Drawing on the Guidance for Federal Agencies on Key Terms in the Executive Order 12898 on Environmental Justice, Appendix A, several guiding principles and considerations for identifying affected minority populations have emerged. In a 2016 Federal Interagency Working Group report, Promising Practices for EJ Methodologies in NEPA Review, it was noted that the identification of minority populations can be accomplished in various ways, including the following: • No Threshold Analysis. While the use of thresholds to identify minority populations is an established method, it may fail to reveal affected protected populations and this is particularly of concern during the NEPA phase. The No-Threshold analysis seeks to identify all minority populations regardless of popula- tion size. It can be relevant in cases where there are small pockets, transient populations or geographically dispersed populations, or seasonal or migrant workers. The approach should be informed by selecting an appropriate geographic unit of analysis (e.g., census block, block group) and determining the total number of minority individuals (all individuals other than non-Hispanic whites) and the percent minority for each geographic unit of analysis within the affected environment. Mapping and tables should be prepared as appropriate. A written rationale should explain the selection of the geographic unit of analysis, the reference community, and other methods used to identify minority populations. • Fifty Percent Analysis. The Fifty Percent analysis can be conducted to initially identify the extent to which minority populations reside within the affected environment. An aggregate of minority populations over 50 percent for the entire affected environment has been used as indicator for increased scrutiny in the EJ analysis under EO 12898 and may be appropriate (e.g., to assess majority-minority populations). The Fifty Percent analysis can function as a direct measure to ensure the identification of minority populations when they comprise a majority of an appropriate geographic unit of analysis (e.g., block group) regardless of whether the Meaningfully Greater analysis has a similar outcome. Agencies may also conduct the Meaningfully Greater analysis, regardless of the results from the Fifty Percent analysis. • Meaningfully Greater Analysis. In a Meaningfully Greater analysis, agencies identify affected protected populations in situations where a large percentage of the residents is composed of minority individuals. Depending on the study purpose, relevant demographic information on minority population(s) is collected for the appropriate primary study area using an appropriate geographic unit of analysis (e.g., census tract, block group, or block) as well as for a larger scale reference community (e.g., municipal, county, region, or state) to identify the minority population patterns. There are no precise criteria for selection of the appro- priate meaningfully greater threshold for comparison (e.g., 10 or 20 percent greater than the reference community) and it varies by agency. The Meaningfully Greater analysis requires use of a reasonable, subjec- tive threshold. Mapping and tables should be prepared as appropriate. A written rationale should explain the selection of the geographic unit of analysis, the reference community, the meaningfully greater threshold, and other methods used to identify minority populations. Regardless of method for identifying minority populations, it is important to recognize that FHWA expects a thorough consideration. As stated in FHWA’s Frequently Asked Questions on its website: “It is not correct to suggest that if minority populations are small (such as less than 50%) that there is no need to assess whether there is an Environmental Justice issue. Environmental justice determinations are made based on effects, not population size. It is important to consider the comparative impact of an action among different population groups.”

96 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Identifying Low-Income Populations The 2016 Federal Interagency Working Group report, Promising Practices for EJ Methodologies in NEPA Review, listed several guiding principles and steps for the identification of low-income populations, which are para- phrased below: • Use current U.S. Census Bureau poverty thresholds and U.S. Department of Health and Human Services poverty guidelines. Agencies and practitioners should consider the publication date for poverty data that is used in the U.S. Census Bureau’s poverty thresholds and the U.S. Department of Health and Human Ser- vices’ poverty guidelines or other agency-specific poverty guidelines. Using the most current poverty data is preferable, but agencies should also consider whether there are differences in the dates for local, state, and national data. • Refinements to low-income status determinations may be appropriate. Use of local data sources on poverty may be more current than the U.S. Census Bureau’s American Community Survey (ACS) or other periodically collected data sources. • Consider various criteria for defining low-income populations and appropriate thresholds. There are several ways to assess low-income thresholds, such as identifying the proportion of individuals below the poverty level, households below the poverty level, and families with children below the poverty level. Using more than one method may be appropriate for making a determination. • Recognize that low-income populations may be clustered or dispersed and, generally, may not be evenly distributed throughout the general population. Selecting a geographic unit of analysis for the primary study area and reference community (e.g., county, state, or region) without sufficient justification may not accu- rately portray low-income population percentages or could arbitrarily dilute their representation within the selected unit of analysis. If there are small clusters or dispersed populations, the identification of low-income populations may warrant the use of alternative criteria—an approach that warrants documentation and lays out the rationale for the identification of the low-income population that is analogous to the “no-threshold” method for race. • Low-income status need not always be capped at the poverty level. In some instances, it may be appropriate for agencies to select a threshold for identifying low-income populations that exceeds the poverty level. In accordance with these principles, the specific steps for conducting the low-income threshold analysis and the alternative criteria analysis are described in the Working Group report. Preparing a strong written rationale for the selection of data sources and other methods used to identify low-income populations is essential regardless of whether the alternative criteria analysis or low-income threshold criteria analysis is used. While threshold methods can be effective in categorizing communities or places, the practi- tioner should recognize that the method is subjectively defined in terms of its selection of ref- erence geographies and threshold values. In specific circumstances, using a threshold method can be efficient for addressing state, regional, or other large-scale planning-level analyses and acceptance for the method can be usefully fostered through careful technical research as well as through expert panel or public participation processes. However, other methods for identifying and understanding the locations, activities, needs, and effects of tolling actions on EJ popula- tions are generally warranted for environmental or corridor-specific studies and can certainly supplement the threshold method identification approach. Other Methods for Identifying EJ Populations and Activity Spaces. To support planning and project-related EJ assessments, other methods are used to identify areas in communities where protected populations reside, may reside in the future, or to analyze “activity space.”

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 97 Activity space may be understood as the destinations to which potentially protected popula- tions travel to access jobs, education, and other opportunities or to reach places that are valued for cultural, community, and family life. NCHRP Report 532: Effective Methods in Environ- mental Justice Assessment described 12 such methods for identifying potentially affected pro- tected populations (Forkenbrock and Sheeley, 2004). In Table 2, a summary table from NCHRP Report 532 is shown highlighting key considerations and the context for their use, required data needs, and expertise. These listed methods reaffirm the importance of drawing on local sources of knowledge and public outreach, use of field visits, and conducting surveys to ensure that analysis is more than a desktop exercise. The last three methods referenced apply to analyzing the activity space; their use for specific plans and environmental studies is generally influenced by considerations related to the spatial nature of the effects, the perceived complexity of the effects, and their perceived importance. What Are Its Limitations? There is no uniform method for defining spatial concentrations of EJ populations (i.e., low- income and minority populations that may be affected by a plan or project). Often, the criteria used to identify high concentrations of low-income and/or minority populations will refer to guidelines established by the Council on Environmental Quality (CEQ) in its 1997 report, Envi- ronmental Justice Guidance under the National Environmental Policy Act. Consistent with the CEQ guidance, spatial concentrations of EJ populations or EJ communi- ties are likely to be identified when selected small area census geographies (e.g., census blocks, block groups, or tracts) are mapped and categorized in one of the following ways: • The minority population exceeds 50 percent of the population in a potentially impacted area. • The minority or low-income population percentage is “meaningfully greater” in the potentially impacted area than the minority or low-income population in the general population (i.e., the region of comparison) or other appropriate geographic area. • If there is more than one minority group present and the minority percentage, as calculated by aggregating all minority persons, meets one of the above-stated thresholds. U.S. DOT and FHWA have not adopted an enforcement approach that defines specific thresh- olds for identifying low-income or minority populations, although they do require that minority populations be examined separately from low-income populations. On FHWA’s “Questions and Answers on Environmental Justice” webpage, it is cautioned that the terms “minority” and “low-income” should not be combined. There are minority populations of all income levels; similarly, low-income populations may be minority, non-minority, or a mixture in a given area (FHWA, 2015b, Question 8). When identifying EJ populations, agencies may make their own determinations and assump- tions, but these should be conveyed in their planning and environmental documents. The identification of EJ persons or populations must be developed on a “case-by-case basis— depending on context, this may mean one person, multiple families, or entire communities” (FHWA, 2015a). Moreover, determinations are to be made based on effects, not population size. The comparative impact of an action among different population groups is of importance. FHWA does not dismiss the possibility of a disproportionately high and adverse effect of a proposed action even when the minority or low-income populations in an area are small. Most pointedly, “[I]t is not correct to suggest that if minority or low-income populations are small (such as less than 50%) that there is no need to assess whether there is an Environmental Justice issue” (FHWA, 2015b, Question 9).

98 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Method Assessment Level Appropriate Uses Use When Data Needs Expertise Required 1. Local knowledge and public input All Recommended in all situations Initial evaluation of potential for distributive effects and to assure quality of findings of other methods Low Local area/ community involvement 2. Threshold analysis Screening/ detailed Regional plans, State Transportation Improvement Program (TIP)/TIP, system assessment Demographic patterns must be evaluated for large areas Low GIS, census data 3. Spatial interpolation Screening/ detailed Corridor/project Demographic patterns must be evaluated for small areas or population patterns must be evaluated for finite areas of effect Medium GIS, census data 4. Field survey Detailed Corridor/project Detailed residence, business, and public space location information is required Low/ medium Global Positioning System (GPS) & photo interpretation can be useful 5. Customer survey Detailed All System users could experience distributive effects Medium/ high Survey design 6. Population surfaces Detailed Regional plans/ corridor/project Scenario modeling or integration with grid- based modeling packages is required High GIS, census data 7. Historic data review Detailed All Past projects or investment plans are at issue, or when population trends are needed Medium/ high GIS, census data 8. Population projection Detailed Regional plans, STIP/TIP Planning horizon is five years or more High Census data, statistical modeling 9. Environmental justice index Screening/ detailed All Combined analysis of multiple demographic factors is needed Medium/ high Census data, GIS 10. Personal interviews Screening/ detailed Regional plans/ corridor/project Analysis of a relatively well-defined impact area Low/ medium Interview techniques 11. Abbreviated diary Detailed Corridor/project Analysis of movement along a corridor is needed Medium Sampling, surveys 12. Space-time activity analyses Detailed Corridor/project Analysis of movement along a corridor is needed High Sampling, surveys, GIS, GPS Source: Forkenbrock and Sheeley, 2004 Table 2. Summary of methods for identifying protected populations.

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 99 What Types of Resources and Costs Are Required? As summarized in Table 2, data needs and cost vary greatly across methods of analysis depend- ing on their use of secondary and primary data sources. The analysis should be complementary with PIP (see tool, “Preparing, Implementing, and Assessing a Public Involvement Plan”) and the principles of fostering meaningful engagement; the level of detail should also be appropriate for the issues involved and the stage of decision-making. Who Has Used It Successfully? Depending on the region and project study purpose, definitions and methods for identifying affected populations or communities of concern and application of geographic study boundaries will vary. First, the selection of the “study area” in many of the studies refers to a buffered area surrounding the proposed improvement/toll implementation. Often these buffered areas range in size, from as small as 1,500 feet to 0.5 mile. While many current and/or potential facility users may begin future trips from this area, it represents a small segment of the catchment area of potential users as it relates to travel behavior in tolling projects. Example 1: Interstate 680 Corridor, Metropolitan Transportation Commission, San Fran- cisco Bay Area. The overall study area was determined by mapping the percentage of trips and the number of trips originating in each traffic analysis zone (TAZ) that use one or more of the roadways that would include express lanes within the Metropolitan Transportation Commission (MTC) Regional Express Lanes Program (or the MTC Program) for a portion of the trip. The analysis suggested that Alameda, Contra Costa, and Solano Counties represented an appropriate study area for the MTC Program. Then, for the project-level analysis, three study areas were defined and considered in the analysis: • Direct Impact Area (DIA). The area most likely to experience the potential direct impacts from the project construction and operation; this includes all census tracts within 0.25 mile of the I-680 corridor. • Extended Resource Area (ERA). The likely users of the proposed express lane facility; this includes the entirety of Alameda and Contra Costa Counties. • Region of Comparison (ROC). The study area identified for the MTC Program, used to deter- mine if potential project-related adverse impacts are disproportionate in comparison to the greater area; this includes Alameda, Contra Costa, Solano, and Santa Clara Counties. (Santa Clara County was added because part of the express lane extends into this county.) MTC’s regional travel demand model (TDM) (Travel Model One) was used to review regional travel patterns and identify the area most affected by the express lanes within the MTC Program. The analysis was performed using a select link analysis that simulates the travel patterns of all Bay Area residents on a typical weekday and estimates the traffic flow on every major roadway in the region. The U.S. Census Bureau and ACS data were collected to determine the census tracts that comprise the Direct Impact Area that include EJ populations of concern. For race and ethnicity data, decennial data from 2010 were available for the full population. For poverty information, only sample data were available; ACS five-year estimates (2006 to 2010) were used to identify low-income populations. As a majority-minority region, MTC identified areas with concentrations of minority popula- tions in geographic units where 70 percent or more of the population was identified as minority. As a high cost of living region, concentrations of low-income persons were identified where 30 percent or more of individuals within a geographic unit are below 200 percent of the poverty

100 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox level. MTC used 200 percent of the poverty level to account for the region’s high cost of living relative to the nationwide federal standard. The appropriateness of these determination thresh- olds were analyzed in the report. Example 2: SR 520: I-5 to Medina Bridge Replacement, Washington State Department of Transportation, Central Puget Sound Region. The environmental analysis used three study areas: the project study area, the Evergreen Point Bridge travelshed study area, and the Pontoon Construction and Transport study area. • Project Study Area. The project study area was used to determine the effect of project construc- tion and operation on the human environment within a specified distance of the construc- tion limits, including the effects on the residents and the people who worked in the project study area. • Travelshed Study Area: The Evergreen Point Bridge travelshed study area was used to under- stand the effects of tolling on bridge users. The Evergreen Point Bridge travelshed study area included the geographic area from which traffic on the Evergreen Point Bridge originates. • Pontoon Construction and Transport Study Area. This geographic area included the sites that Washington State Department of Transportation (DOT) was evaluating for construction of the supplemental stability pontoons required for a new six-lane floating bridge and also included the haul route that would be used to transport the pontoons from the production site to the bridge construction site. EJ analysts used six approaches to collect information on low-income and minority populations: 1. Travelshed determination, 2. Demographic analysis, 3. Surveys of Evergreen Point Bridge users, 4. Focus groups and Spanish language telephone interviews with Evergreen Point Bridge users, 5. Public involvement activities, and 6. Windshield survey. Example 3. EJ Index, North Central Texas Council of Governments, Dallas–Fort Worth Region. In the Dallas–Fort Worth region, minority populations account for close to 50 percent of the population. In response to this pattern, the North Central Texas Council of Governments (NCTCOG) applied an EJ index method that identifies block groups with higher “degrees of vulnerability” using three variables: persons per square mile, percent below poverty, and percent minority. The EJ index method results are applied to categorize block groups as either “protected” or “non-protected.” Similar to other mathematical indices, the EJ index has some limitations. It is useful for depicting combinations of variables as a single value, which makes it valuable for screening assessments and for mapping. However, the underlying variables comprising the index must be used if more detailed analysis is required (Forkenbrock and Sheeley, 2004). Further description of this method as it was applied by the NCTCOG is described as one of the case examples in the Toolbox (see case example, “Using an EJ Index to Identify Affected Populations, Dallas–Fort Worth Metro Region”). In addition to the U.S. Census Bureau and the EPA and HUD data tools noted above, several tools in the Toolbox can help describe data sources and methods for the identification of EJ populations. These include: • Using Public Use Microdata Samples to Profile Transportation Characteristics and Differences, • Using NHTS to Profile Transportation Characteristics and Differences, • Using Travel Demand Model for EJ Assessments, and • Applying a Select Link Analysis to Assess Trip Patterns.

Developing a Socioeconomic Profile and Community Characteristics Inventory for Environmental Justice Assessments 101 Resources FHWA. 2015a. Federal Highway Administration Environmental Justice Reference Guide. Retrieved from https:// www.fhwa.dot.gov/environment/environmental_justice/resources/reference_guide_2015/fhwahep15035.pdf. FHWA. 2015b. Questions and Answers on Environmental Justice. Retrieved from http://www.fhwa.dot.gov/ environment/environmental_justice/facts/ejfaq.cfm. Forkenbrock, D. J., and Sheeley, J. 2004. NCHRP Report 532: Effective Methods for Environmental Justice Assess- ment. Transportation Research Board of the National Academy of Sciences, Washington, D. C. HDR Engineering, Inc. 2013. MTC Regional Express Lanes Interstate 680 Corridor: Environmental Justice Technical Memorandum. Prepared for the Metropolitan Transportation Commission, Oakland, CA. NCTCOG. 2014. “Mobility 2035–2013 Update Appendix B: Social Considerations.” http://www.nctcog.org/trans/ mtp/2035/documents/AppendixB.pdf. U.S. Census Bureau. 2015. American FactFinder. Retrieved from http://factfinder.census.gov/faces/nav/jsf/pages/ index.xhtml. U.S. Department of Housing and Urban Development. 2015. Affirmatively Furthering Fair Housing Data and Mapping Tool: User Guide. Retrieved from https://www.hudexchange.info/resources/documents/AFFH- Data-Mapping-Tool-User-Manual.pdf. EPA. 2015. EJSCREEN: Environmental Justice Screening and Mapping Tool. Retrieved from https://www.epa. gov/ejscreen/how-interpret-standard-report-ejscreen. EPA. 2015. EJSCREEN Technical Documentation. Retrieved from https://www.epa.gov/sites/production/files/ 2015-05/documents/ejscreen_technical_document_20150505.pdf. EPA, Federal Interagency Working Group on Environmental Justice & NEPA Committee. 2016. Promising Practices for EJ Methodologies in NEPA Review. EPA Pub. No: 300-B-16-001. Retrieved from https://www. epa.gov/sites/production/files/2016-03/documents/iwg-promising-practices-final-033016.pdf. Washington State DOT. 2009. Environmental Justice Discipline Report: SR 520: I-5 to Medina Bridge Replace- ment and HOV Project Supplemental Draft EIS.

102 Using Public Use Microdata Samples to Profile Transportation Characteristics and Differences What Is It? PUMSs are computerized files containing samples of individual records, with identifying information removed, that show the popula- tion and housing characteristics of the units and people represented on the data records. The Integrated Public Use Microdata Series USA (IPUMS-USA) consists of more than 60 high precision samples of the American population using the U.S. Census and ACS Public Use Micro Sam- ples. IPUMS-USA, an initiative of the Minnesota Population Center, maintains a collection of microdata survey samples compiled by the federal government over more than a hundred years. It is an impor- tant source of quantitative data on changes in the U.S. population. IPUMS-USA seeks to harmonize these samples created at differ- ent times through the development of uniform codes and technical documentation to facilitate analysis of social and economic changes. IPUMS-USA provides these records to researchers through a web- based data platform. Each record in a microdata sample is a person numerically coded for all its characteristics. Records can also be categorized as households or families in order to understand the relationships between residents. The sample data supports the assessment of transportation characteris- tics, including mode of travel to/from work, travel time, vehicle owner- ship and assessment of how travel behavior varies by race, income, and demographic categories. Why Is It Effective in Environmental Justice Analysis? Using PUMS allows for a uniquely detailed analysis of household, income, and transporta- tion characteristics at varying geographic scales with a high degree of precision. This dataset is very well-suited for examining the differences in journey-to-work and vehicle ownership char- acteristics by income levels or racial categories, which can contribute to a greater understand- ing of the financial and mobility challenges within a state or metropolitan area experienced by households and individuals—an important context for considering potential effects of toll road pricing. t o o l 2 Framework Step • Scope Approach to Measure and Assess Impacts Stages in Decision-Making • Policy and Planning • Project Design and National Environmental Policy Act Tools and Techniques • Descriptive Statistics • Cross Tabulations • IPUMS-USA 1 Percent Sample Affected Populations • Low-Income • Minority Examples Featured • University of Washington, Puget Sound Region, SR 520 Bridge Replacement • Metropolitan Transportation Commission, San Francisco, Bay Bridge Congestion Pricing Study

Using Public Use Microdata Samples to Profile transportation Characteristics and Differences 103 What Are Some Techniques for Implementing This Tool? The use of PUMS data supports a highly specific and adaptable classification of the low and non-low-income households and the breakdown both within households in these categories and between categories for characteristics such as car ownership, means of commute to work, and travel time to work. Table 1 illustrates how PUMS data can be used to observe trans- portation and racial characteristics of low and non-low-income households throughout the United States. In this example, researchers used the 1 percent sample from the 2013 ACS to examine the transportation characteristics of low and non-low-income households at the national, state, or metropolitan levels. For this analysis, low-income households were defined as households whose income was at or below 100 percent of the poverty level, as determined by the total household income and family size. Non-low-income households were all households whose income was 101 percent or more of the poverty level. In 2013, the 2013 Federal Poverty Level threshold for a family of four was $23,550 and for a family of three was $19,530 within the 48 contiguous states and the District of Columbia. While not shown here, in some cases, researchers and sponsoring jurisdictions have preferred to adopt a higher threshold for defining low-income in recognition of the high cost of living within a metropolitan region, or in deference to eligibility criteria for other federal programs extended to low-income families such as the free and reduced price lunch program (i.e., free lunches for children in families at 130 percent of poverty level and reduced priced meals in families at 130 percent and 185 percent of the poverty level). Using descriptive statistics and basic pivot-table tools, characteristics such as car ownership, means of commute to work, and travel time were compared within and between the low and non-low-income household groups. The analysis demonstrates that low-income households have noticeably different travel characteristics than non-low-income households. In particular, low-income households are less likely to have a car available, have fewer cars available, and drive alone to work less. Given that a lower percent of low-income households drive to work alone and that a larger percent take public transportation or other means of commute to work, it is perhaps reasonable to assume that an increase in tolling would affect a proportionately smaller amount of low-income households than non-low-income households. Nonetheless, it is also likely that those who are affected will experience a considerably higher cost burden from increased tolling charges or fees. PUMS data can reveal similarities and differences at state or regional levels, among other geographies. In this example, income and travel patterns for the state of New Jersey and the New York Metropolitan region are shown in comparison with the nation. In this example, the New York Metropolitan region has a significantly greater share of low-income households as a share of all households and, at 31.9 percent of all households, it has at least twice as high a percentage as the nation or the state of New Jersey (Table 1). Comparing New Jersey with the nation, the patterns are broadly similar in terms of the per- centage of low-income households and the percentage of households having car ownership, although some important differences do exist: • New Jersey low-income households are more likely to have cars than nationally (1.47 vs. 1.38) while non-low-income households have fewer cars than the nation (2.06 vs. 2.35). • New Jersey’s low-income households experience longer commutes relative to non-low- income households—a pattern that differs from the national pattern and perhaps reflects the large number of jobs outside the urban areas that have a proportionately greater number of New Jersey’s low-income Black and Hispanic households.

Transportation and Racial Characteristics U.S. National Level New Jersey State Level New York City Metro Region Low-income Households Non-low- income Households National Total Low-income Households Non-low- income Households Statewid e Total Low- income Households Non-low- income Households Region wide Total Percent of regional population 15.2% 84.8% 100% 14.4% 85.6% 100% 31.9% 68.1% 100% Characteristics of households Cars available (%) 74.7% 93.8% 93.3% 72.3% 93.8% 90.7% 48.6% 88.4% 75.4% Mean number of cars available 1.38 2.35 2.01 1.47 2.06 1.98 0.87 1.94 1.53 Characteristics of workers Commuting mode Drives alone 68.1% 81.8% 80.8% 52.7% 76.0% 74.5% 32.7% 60.4% 52.6% Carpools 13.6% 9.3% 9.6% 13.3% 8.5% 8.8% 8.8% 5.8% 7.4% Public transportation 8.3% 5.0% 5.2% 18.9% 10.9% 11.4% 41.0% 27.6% 31.5% Table 1. Transportation and racial characteristics for low and non-low-income households in the United States, State of New Jersey, and the New York metropolitan region.

Race of population outside group quarters White (Non-Hispanic) 43.7% 66.1% 62.6% 36.1% 60.8% 57.2% 27.9% 67.7% 47.9% Black 20.9% 10.4% 12.0% 20.2% 10.7% 12.1% 21.6% 8.8% 15.4% Hispanic 27.0% 15.3% 17.2% 35.3% 17.0% 19.6% 38.2% 9.9% 24.0% American Indian & Alaskan Native 1.2% 0.5% 0.7% 0.3% 0.1% 0.1% 0.2% 0.1% 0.2% Asian 4.2% 5.3% 5.2% 4.8% 9.6% 8.9% 9.5% 11.4% 10.3% More than two Races or Other 3.0% 2.3% 2.4% 3.3% 1.9% 2.1% 2.6% 2.1% 2.2% Source: IPUMS-USA, 2013 ACS 1YR 1% Sample Note: NYC Metro Region is New York – New Jersey – Newark – Jersey City, NY-NJ-PA. Other commutemode 10.1% 3.9% 4.4% 15.1% 4.6% 5.3% 17.5% 6.2% 8.5% Mean commute time in minutes 23.5 26.1 25.9 31.0 26.7 30.7 33.4 36.4 35.2

106 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Comparing the New York Metropolitan region with the state of New Jersey and the nation, some other differences are apparent: • Car availability among low-income households is significantly lower within the New York Metropolitan region than either the state of New Jersey or the nation. • While public transportation plays an important role in the commute to work in the state of New Jersey, the New York Metropolitan region, which contains several northern New Jersey counties with strong work ties to New York City, exhibits an even higher dependency on public transportation relative to the nation or the state of New Jersey. These differences are particularly pronounced for low-income households. The PUMS dataset can be used to examine race differences of persons and households in poverty as well as travel characteristics. In this case, race/ethnicity data was constructed from two Census variables describing race and Hispanic origin so respondents were assigned to one of six mutually exclusive categories. In this example, the data shows that Non-Hispanic Whites comprise a plurality of low-income households in the nation and in the state of New Jersey, while Hispanics comprise the largest segment of low-income households (38.2%) in the New York Metropolitan region. For households that are not in poverty, the patterns are significantly different. Conversely, minority populations comprise a disproportionate share of the low-income households. While this analysis examines general race characteristics data available through PUMS, it is also possible to use more detailed classification schemes. For example, it is possible to examine countries of origin of Hispanic populations or Asian populations, greater dis- aggregation of multi-racial respondents, or the characteristics of individual Native American tribes, among other population segmentations. What Are Its Limitations? PUMS requires some proficiency in handling and manipulating the dataset. For example, IPUMS users will need a statistical package such as SPSS, SAS, and STATA to handle the data extracts from the website. Technical documentation is required to read that data, coded as ASCII numeric data. Given the size of the U.S. population, the sample file contains household and person weight fields. The demographer undertaking this work must apply these weights in con- ducting analyses. For example, a 1 percent sample from the 2013 ACS for the U.S. and state of New Jersey returns an unweighted and unfiltered 3,132,795 and 88,510 cases, respectively. The New York Metropolitan region returns 179,678 cases for the same 1 percent ACS 2013 sample. The 5 percent sample would be significantly larger. One of the common critiques of demographic research relating to poverty is using too conser- vative an approach to classify the highest income brackets. The highest personal income code is just under $10 million and the highest percent of poverty ratio is set at 501 percent. PUMS data capture information on the commute to work but fail to capture data on non- work-related trips, which consists of a large number of trips taken and may vary from the work trip in terms of the mode of transportation selected. PUMS data also do not provide detailed data on the residential location of the respondent. The Public Use Microdata Area (PUMA) offers a sub-regional geographic reporting unit, but has a minimum population requirement of 100,000. This results in a reporting geography that is large and inconsistent with other jurisdictional or census boundaries, limiting its utility for some research and plan- ning purposes (see text box, IPUMS-USA: Core Features and Limitations).

IPUMS-USA: Core Features and Limitations Integrated Public Use Microdata Series USA (IPUMS-USA) offers the researcher an opportunity to look closely at state and metro area travel through a detailed lens of race and income, but before it can be used for this pur- pose, it is important to understand its core features and limitations. An overview of the tool is provided below, drawing on the language of the IPUMS technical documentation, along with observations about some challenges encountered in working with the dataset to create the files shown here. Microdata. Census microdata are composed of individual records containing information collected on persons and households. The unit of observation is the individual. The responses of each person to the different census questions are recorded in separate variables. Microdata stand in contrast to more familiar “summary” or “aggregate” data that might be found through use of the U.S. Census Bureau’s American FactFinder. IPUMS does not provide such summary statistics, but rather enables users to generate customized statistics that focus on specific issues of concern to the researcher. Using Weights and Filters. Most IPUMS samples are unweighted or “flat” (i.e., every person in the sample data represents a fixed number of persons in the population). Approximately one-fourth of the IPUMS samples, however, are weighted, with some records representing more cases than others. Thus, persons and households with some characteristics are over-represented in the samples, while others are underrepresented. Users must apply sample weights to obtain representative statistics: (1) for person-level analyses, the WTPER variable must be applied to have an appropriately weighted sample (WTPER contains the population represented by each individual in the sample) and (2) for household-level analyses, the WTHH variable must be applied to give the number of households in the general population represented by each household in the sample. Because group quarters (collective dwellings) are not usually weighted properly for household-level analyses, it should gener- ally be excluded using the GQ variable. Data Format. IPUMS produces fixed-column ASCII data. Data are entirely numeric. By default, the extraction system rectangularizes the data (i.e., it puts household information on the person records and does not retain the households as separate records). This can lead to distortion in the analyses at the household level. The number of observations will be inflated to the number of person records. The user must either select the first person in each household (PERNUM variable) or select the “hierarchical” box in the extract system to get the proper number of household observations. The rectangularizing feature also drops any vacant households, which are otherwise available in some samples. Thus, when performing calculations for persons, and not households, any filter or script implemented to uniquely identify households must be turned off or eliminated in order to accurately analyze the population of persons, appropriately weighted as discussed above. Extracted Data. The researcher downloading requested microdata from the IPUMS website receives, in addition to the ASCII data file, a statistical package syntax file to accompany each extract. The syntax file is designed to read in the ASCII data while applying appropriate variable and value labels. SPSS, SAS, and STATA are supported. Geographic Limitations. As a public use data sample, measures to assure confidentiality have been made, including the suppression of identifying information. Geographic information is somewhat limited, particularly for places with populations less than 20,000. In some places, the threshold is higher, and in some places, only states or regions can be determined. Researcher Flexibility. With IPUMS limitations properly understood, the researcher will enjoy flexibility in examining race, income, and travel characteristics—among other issues—which could be informative for con- sidering the socioeconomic context in which tolling decisions are being made. For calculations based on race, for example, it is important to remember that a Hispanic origin is classified separately from other race catego- ries because it is not considered a race, but rather a place or culture of origin. In the example provided, the researchers performed a cross-tabulation of RACE (general variable) and HISPAN in order to calculate those individuals within a particular racial group that identify as Non-Hispanic versus those who identify as Hispanic. Additionally, computations to combine the various Asian racial groups could be informative and can be done inside the program through recoding or later when preparing report tables.

108 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox What Types of Resources and Costs Are Required? Given that the data can only be downloaded and used via a statistical extraction package (SPSS, SAS, and STATA), obtaining the software and a license presents an up-front cost. The user must be familiar with the technical documentation to interpret the data correctly. A major resource required to complete this analysis is a planner/demographer with experience or knowledge of both the necessary statistical extraction packages and the IPUMS data or ASCII coding. With such a skilled individual, the analysis referenced here could be completed within 2 to 5 days, depending on the user’s prior familiarity with the tool. Who Has Used it Successfully? PUMS data has been used extensively all over the country for a variety of purposes. This discussion has largely focused on how PUMS can be used to develop transportation profiles through the use of customized cross-tabulations. Users of PUMS data, as shown in the example above, can combine the separate race and Hispanic origin census questions to support environ- mental justice analyses. Nationally, PUMS has been used to support research on environmental justice through seg- mentations of commuting behavior by income or race as well as by other specific subgroups such as gender, foreign-born status, and limited English proficiency. Because of its flexibility, PUMS has been used to examine the overlapping relationships between income, race, persons with disabilities, and the elderly—a type of customized cross-tabulation unavailable through the American FactFinder (Tierney, 2012). These research efforts have been used for transportation, job access, and labor force participation-related studies. PUMS has been used by researchers to support travel demand modeling, travel survey sam- pling validation and weighting, and the development of synthetic population microsimulation (see tool, “Using Travel Demand Models for EJ Assessments”), among other transportation study purposes (Tierney, 2012). Two examples of its use in tolling are briefly mentioned below. • University of Washington researchers examined the 2007 ACS PUMS for Washington State’s Puget Sound Region. The data set included 34,106 individual records and 14,911 households for which sample weights were applied. The PUMS data was then combined with the Puget Sound Regional Council’s 2006 Household Activities Survey (HAS) to enable consideration of the geographic distribution of travel. The HAS survey contained information on the travel behavior of 4,746 households in the Puget Sound Regional Council’s region. The University of Washington researchers merged important features of the HAS survey—specifically, basic demographic information and latitude and longitude of home and work locations—to map workers’ probable commuting routes. Using fairly limited data from HAS on income, they were then able to integrate the two data sets and identify low-income and non-low-income workers who were most likely to use routes that would be subject to tolling in the future (Plotnick et al., 2011). Following this approach, the researchers could consider the scale of potential effect of the bridge tolling option on all of the metropolitan region’s low-income households—taking into consideration those who would and would not likely be commuting via the subject SR 520 Bridge as well as the particular segment of low-income households who would still require the bridge route after its implementation. • The San Francisco Metropolitan Transportation Commission used PUMS data to tabulate the income distribution of commuters by means of transportation (e.g., bus versus drive alone versus rail) in support of the Bay Bridge Congestion Pricing Study.

Using Public Use Microdata Samples to Profile transportation Characteristics and Differences 109 Resources Minnesota Population Center, University of Minnesota. No Date. Integrated Public Use Microdata Sample (IPUMS-USA), Frequently Asked Questions. Retrieved from https://usa.ipums.org/usa-action/faq#ques22. Plotnick, R. D., Romich, J., Thacker, J., and Dunbar, M. 2011. “A Geography-Specific Approach to Estimating the Distributional Impact of Highway Tolls: An Application to the Puget Sound Region of Washington State.” Journal of Urban Affairs, Vol. 33, No. 3, pp. 345–366. Ruggles, S., Alexander, J. T., Genadek, K., Goeken, R., Schroeder, M. B., and Sobek, M. 2010. Integrated Public Use Microdata Series: Version 5.0 [Machine-readable database]. University of Minnesota, Minneapolis. Tierney, K. 2012. NCHRP Synthesis 434: Use of the U.S. Census Bureau’s Public Use Microdata Sample (PUMS) by State Departments of Transportation and Metropolitan Planning Organizations. Transportation Research Board of the National Academy of Sciences, Washington, D.C.

110 What Is It? The 2009 NHTS is a nationally representative travel diary survey con- taining information on more than 150,000 U.S. households, the charac- teristics of the members of the households, details of the trips they take on a survey day, as well as the characteristics of the vehicles they own and their attitudes toward several transportation issues. NHTS (and its predecessor, the Nationwide Personal Transportation Survey) has been conducted seven times since 1969, with the most recent five iterations readily accessible online. NHTS is conducted at irregular intervals and is a repeated cross-sample; the households surveyed in the 2001 survey, for instance, are not the same households surveyed in the subsequent survey (FHWA, 2009). NHTS data are provided in several tables, including a roster of house- holds, a separate table of individuals, a table of the trips taken by house- hold members during a survey day, and a table related to the vehicles owned by the household. Unique identifiers allow researchers to link these files to gain an understanding of the overall travel patterns of house- holds. Additionally, by entering into a data agreement with NHTS pro- gram staff, researchers and practitioners can gain access to geographically specific data for individuals’ home and workplace locations. Certain states and Metropolitan Planning Organizations (MPOs) have contracted with NHTS program staff to obtain an oversample of their region. These data, in many cases, contain addi- tional questions the state or MPO staff deemed important for regional or statewide planning. These data are typically housed within the MPO; researchers and practitioners may gain access to the data through an agreement with the MPO or state office that houses the dataset. States that requested add-ons in the 2009 NHTS ranged from California and New York to South Dakota and Vermont. MPOs ranging from Chittenden County in Vermont to the Maricopa Association of Governments in Arizona have similarly contracted for add-ons. A full list of add-ons for the 2009 NHTS can be found in the NHTS 2009 User’s Guide. Why Is It Effective in Environmental Justice Analysis? Using NHTS allows for a very detailed analysis of the travel patterns of households among population subgroups as well as across a variety of geographic characteristics. For instance, researchers can examine the differences in transportation choices between lower- and higher- income households, as well as across racial and ethnic groups. NHTS contains some information Using the National Household Travel Survey to Profile Transportation Characteristics and Differences T O O L 3 Framework Step • Scope Approach to Measure and Address Impacts Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act Tools and Techniques • Descriptive Statistics • Cross Tabulations • National Household Travel Survey Affected Populations • Low-Income • Minority

Using the National Household Travel Survey to Profile Transportation Characteristics and Differences 111 on transportation expenditures, including tolls paid. In contrast to the U.S. Census Bureau’s ACS, this dataset examines various trip purposes in addition to the work trip. What Are Some Techniques for Implementing This Tool? NHTS offers a limited online analysis tool, in which users can produce cross-tabulations of vari- ables, such as the incidence of tolls paid by various income groups. NHTS staff have also produced a number of frequently requested tables as a quick reference for researchers and practitioners. However, the NHTS dataset is most powerful as microdata, downloaded and manipulated in a database management system or a statistical software package. This allows users greater flex- ibility in analyzing particular characteristics and travel patterns across population subgroups or across various geographic contexts. Table 1 includes an example of how NHTS data can be used to observe transportation char- acteristics of low- and non-low-income households throughout the United States. Among the lowest income adults (age 18+) in NHTS, 89 percent are self-identified drivers, while nearly all (97%) of the highest income adults are. Lower-income individuals, when they drive, are consid- erably less likely (5% versus 10%) to pay a toll, either by avoiding tolled facilities or by living in areas without tolls altogether. The modal split between the lowest-earning and highest-earning 25 percent of Americans is less dramatic, with nearly identical rates of driving, but somewhat Category Lowest income 25% Highest income 25% Total Is a driver (age 18+) 89% 97% 94% Paid toll on survey day (drivers only) 5% 10% 8% Travel Modes (all purposes): Car (alone) 42% 43% 43% Car (with others) 44% 44% 44% Bus 3% 2% 2% Rail 0.4% 0.3% 0.3% Bike 0.8% 1.0% 0.9% Walk 10% 9% 9% Trip Purpose: Work-related 11% 18% 16% Shopping 28% 19% 23% Family/personal 33% 32% 32% Social/recreational 27% 31% 29% Table 1. Transportation characteristics of the bottom and top income quartile.

112 Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox more transit usage among low-income households. While not shown in Table 1, among the poorest households (the lowest 10%, for instance), these differences are starker. What Are Its Limitations? Accessing NHTS data needs some proficiency in the handling and manipulating of the data- set. Data downloaded from NHTS require the use of a statistical analysis package or a database management package. The raw data are available in SAS and Dbase formats, though ASCII and SAS-Transport data (both loadable in many other software packages) are also available for analysis using other packages such as STATA and R. Given the size of samples, it is not possible to use most spreadsheet software packages to analyze the data. Unfortunately, NHTS is designed to be nationally representative, but not representative at smaller scales of geography, except in those areas that requested an add-on oversample. However, in some instances, skilled researchers may be able to design a re-weighting scheme that could render the dataset more representative at smaller scales of geography, such as a particular region. There are several critiques of demographic research using NHTS. One important critique is that income is reported in income “bins,” thus reducing the granularity of individual house- holds’ incomes to larger brackets. The highest reported income bracket (greater than $100,000) represents roughly one-quarter of all households in the dataset. Income brackets are much more granular at the lower-income end of the spectrum, using increments of $5,000 per bracket. Another criticism of NHTS data is that the racial composition of the household is reduced to a single variable—the race/ethnicity of the household respondent, or the person responsible for reporting information on all household members. As multi-racial families become increasingly prevalent in the United States, this type of data reduction has become more problematic. What Types of Resources and Costs Are Required? Using the online analysis tool—sufficient for some but not all research questions—is free, but requires basic knowledge of statistics and some rudimentary familiarity with the logic of programming. Given that the microdata—a far more powerful way of using NHTS—can only be used via a statistical or database management software package, there is an up-front cost to obtain the software and a license. When downloading the data, the technical documentation is required to read and interpret the data correctly. A major resource required to complete this analysis is a planner/demographer with experience or knowledge in the necessary statistical or database software. Many variables are more useful when re-coded, so more advanced users will find NHTS more useful than less advanced users. With a skilled individual, the analysis referenced in this tool can be completed within 2 to 5 days depending on the user’s prior familiarity with the tool. Who Has Used It Successfully? NHTS has been used extensively all over the country in a myriad of applications, typically when analyzing national trends. MPOs, state agencies, and researchers with access to regional or state level add-on oversamples have used NHTS at smaller scales of geography. Resources FHWA. 2009. National Household Travel Survey. Retrieved from http://nhts.ornl.gov.

113 Preparing, Implementing, and Assessing a Public Involvement Plan What Is It? Developing a PIP is an important step in identifying the needs and concerns of the public, including low-income and minority populations, in transportation decision-making processes. As a procedural guide for agencies and practi- tioners, the PIP describes key considerations and strategies for carrying out public participation activities. The PIP should guide all stages of transportation decision-making, but is particularly relevant in statewide and metropolitan planning, project design, and environmental review stages of decision-making. Some PIPs will be broad and cover all public involvement conducted by the transportation agency, while others will be plan or project-specific but will be con- sistent with the agency’s overall PIP. Tools and techniques for implementing PIPs necessar- ily include establishing the plan, as well as identifying vari- ous policies and practices that can be incorporated into the plan, such as those described below, as well as procedures to periodically evaluate the PIP’s effectiveness in achieving its goals and carrying out its prescribed policies and practices. Why Is It Effective in Environmental Justice Analysis? The essence of effective EJ practice has been distilled in U.S. DOT and FHWA guidance in the “Fundamental Prin- ciples of Environmental Justice,” which embodies concerns for how transportation decisions are made; their effects on specific affected communities; and in the equitable distribu- tion of benefits that are allocated through the prioritization, investment, design and operations of plans, projects, and other activities. These fundamental principles of EJ include: • Avoid, minimize, or mitigate disproportionately high and adverse human health and environmental effects, includ- ing social and economic effects, on low-income and minor- ity populations; T O O L 4 Framework Step • Recognize the Relevant Decision-Makers and Stakeholders • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement • Identify and Assess Mitigation Strategies • Document Results for Decision-Makers and the Public • Conduct Post-Implementation Monitoring Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Prepare and Implement PIP Affected Populations • Low-Income • Minority • Limited English Proficiency • Low-Literacy Populations • Persons with Disabilities Examples Featured • Washington State DOT, SR 520 Bridge Replacement • California Department of Transportation, I-110 High Occupancy Toll Lane • Kentucky Transportation Cabinet and Indiana DOT, Ohio River Bridges • Massachusetts DOT, Turnpike All Electronic Toll Collection • Minnesota DOT, I-394 Express Lane • Metropolitan Transportation Commission, San Francisco, Regional Express Lane Network

114 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox • Ensure the full and fair participation by all potentially affected communities in the transporta- tion decision-making process; and • Prevent the denial of, reduction in, or significant delay in the receipt of benefits by low-income and minority populations. The PIP is a valuable tool for making clear the agency and practitioner’s planned approaches for ensuring the full and fair participation of low-income and minority populations, among other populations, on toll-related studies. At minimum, the PIP should consider carefully potential barriers to participation—such as those listed in Table 1—and devise strategies for overcoming such barriers. What Are Some Techniques for Implementing This Tool? There are several types of activities that can inform the preparation, implementation, and periodic evaluation of a PIP for policy and planning, project design and environmental review, and the implementation stages of decision-making processes. Drawing heavily on the research of NCHRP Report 710: Practical Approaches for Involving Traditionally Underserved Populations in Transportation Decisionmaking, the following descriptions of tools and techniques provide an overview of the engagement strategies that can be combined to support an analysis of EJ consid- erations in a tolling context. Brief references are made to select case examples of tolling-related projects that have implemented these approaches. Prepare and Implement the Public Involvement Plan. Creating a PIP should be preceded by exploratory and research activities, such as creating a demographic and economic profile or holding meetings with community leaders and organizations, which should be undertaken as early as possible, before project decisions are made. Prepare or Review the Social and Economic Profile of the Study and Project Area. The PIP itself should reflect a comprehensive identification of the various populations in the study and project areas, including its potentially affected protected low-income and minority popu- lations. The scope of the PIP should be informed by a thorough analysis of the social and eco- nomic characteristics of the study area communities. A preliminary inventory and mapping of community facilities and other notable features can be compiled from websites and secondary datasets and should include community gathering places (e.g., playgrounds, senior centers, schools, faith-based institutions) as well as natural or historic features such as important view- sheds (see tool, “Developing a Socioeconomic Profile and Community Characteristics Inven- tory for EJ Assessments”). Develop and Maintain Community Contacts Lists. Developing a community contacts list or database that includes organizations and persons that work with, or on behalf of tradition- ally underserved populations, can improve the ability of the agency and practitioner to identify EJ considerations on tolling-related projects. In creating the database, the practitioner and the agency are refining their knowledge of existing community organizations and leaders and gain- ing greater insight about which individuals and organizations have the capacity to engage low- income, minority, and other traditionally underserved populations as part of their membership or constituency. The database itself is a valuable communications management tool, making it easier to target information to community members as widely or narrowly as appropriate for a particular event. Initial Site Visits to Establish the Scope of the PIP. Verifying the quality of the demographic and community facilities data compiled from secondary data sources can be accomplished through field visits. Before the field visit, the practitioner should reach out to knowledgeable

Barriers Resulting Challenges Strategies to Address Barriers Individuals holding multiple jobs/unusual job hours • Time constraints prevent participation in community outreach activities Take outreach activities to them • Schedule community outreach activities for times of days/nights, weekdays/weekends, and locations convenient to them • Identify community events they are attending and go there • Ask large employers if you can sit in break rooms and conduct surveys • Become a vendor at shopping malls, local fairs, and other events they frequent Low level of education/literacy issues • Less understanding of the potential impacts of toll roads • Less understanding of rights • Unable to provide written responses/comments Hire consultants with special expertise in communicating with people who have low or no education • Conduct surveys orally • Use photographs to locate known landmarks on maps • Create a video that describes what is being shown at themeeting • Show before-and-after renderings • Createmorphs and 3D animations • Color code alternatives • Introduce a project, ask for their comments, and repeat the comments back to ensure that you correctly heard them and they know you listened to them Unique family structures (e.g., single parents, multi-generational families) • Time constraints prevent participation due to family obligations, such as caring for children and the elderly • Single parents often pick up their children and go straight home to feed them supper Provide accommodations for children and the elderly during community outreach activities • Hire a licensed and bonded child care provider formeetings • Provide a meal for children and adults so they do not have to go home for supper (continued on next page) Table 1. Strategies to address barriers to participation.

Barriers Resulting Challenges Strategies to Address Barriers Less likely to have modes of personal transportation (i.e., private car) • Greater difficulty getting to community outreach activities • Less concerned about toll road projects if they do not intend to use them Hold meetings at locations accessible by public transit • Ask if transportation will be needed to an event and provide a licensed driver and, if needed, wheelchair lift or bike rack • Hold events in their community and bring the chairs and tables and food • Hold events at locations they are familiar with, such as schools, parks, religious organizations, and community centers • Make sure accommodations are Americans with Disabilities Act (ADA) accessible Less access to internet, smart mobile phones, and computers; computer literacy issues • Use of websites, social media, and emails to inform and involve communities would be ineffective Distribute printed materials • Visit laundry facilities, homeless shelters, employment offices, food banks, post offices, bus stops/transit, religious organizations, parks, health clinics, grocery stores, community centers, etc. • Distribute information via local radio stations • Use flyer inserts in newspapers (e.g., Latino papers) or distribute information via school district newsletters/cultural programs Language barriers • Less ability to participate in public involvement issues • Less aware of opportunities to influence toll road project outcomes Translate public documents, notices, and hearings for limited English speaking populations • Understand that some may only speak their language and not read it • Provide translations and bilingual speakers during community outreach activities • Prepare communicationmaterials for limited English speaking populations (e.g., bilingual flyers, bilingual radio announcements) Table 1. (Continued).

Distrust of government agencies • Less likely to participate in community outreach activities Work with leaders to increase the credibility of the participatory planning process • Hire consultants with special expertise working with low-income or minority populations • Hold public meetings or events in non-governmental (or less traditional) buildings such as schools, religious facilities, and community centers • Provide opportunities for EJ communities to comment prior to making each decision • Keep the community informed, reply to public input promptly and respectfully Limited understanding of how a project will affect their lives and how participation in the process would benefit them • Less likely to participate in community outreach activities • Need to convince people of their power to influence decisions Hold informal meetings to increase public understanding of how the project may impact the community and inform the public how their input is important • Show the public how their input has informed decisions • Seek public input early in the process and make information available • Involve the communities in decisions that might affect them in approvals and implementation • Provide opportunities to comment prior to making each decision • Keep the community informed even when nothing new is happening • Reply to public input promptly and respectfully Cultural differences—discomfort with meetings • Less likely to attend public meetings and events • More likely to feel event does not invite a true exchange of information Identify preferred community outreach techniques • Open house format • One-on-one interaction • Informal small groups (continued on next page)

Barriers Resulting Challenges Strategies to Address Barriers Process offered by professional is disconnected from ways cultural groups interact with one another andmake decisions Monetary incentivemay be necessary for attendance at events Refreshments/mealsmay need to be provided Location or time of eventmay not feel accessible andmay not contribute to sense of safety • Groups segmented by religious affiliation or secular beliefs • Focus groups/groups segmented by gender, race, ethnicity, and/or nationality • Homes of individuals • Formal presentation with question and answer period • Finding the right combination of processes to provide information and build a two-way dialogue Identify less traditional venues for outreach activities • Religious facilities • School—classrooms, auditoriums, gymnasiums • Libraries • Community centers • Malls • Discount stores Cultural differences—difficulty building trust • • Inability to connect with local leaders who by definition are busy people engaged in many levels of the community Identify and work with and/or through local liaisons, trusted advocates, and recognized leaders to learnmore about cultural differences • Local religious leaders, school principals, socialworkers, health care staff • School: English as a second language coordinators, Local: ethnic organization leaders Source: Prozzi et al., 2006; Aimen and Morris, 2012 Table 1. (Continued).

Preparing, Implementing, and Assessing a Public Involvement Plan 119 persons from the community (e.g., city planners, municipal officials, community-based orga- nizations, neighborhood associations) to learn more about the area. Scheduling time to meet and conduct scoping-type interviews with select stakeholders will make it possible to dis- cover community characteristics not revealed from maps or secondary sources (see text box, SR 520: I-5 to Medina Bridge Replacement and High-Occupancy Vehicle—Initial Site Visits to Scope Efforts). SR 520: I-5 to Medina Bridge Replacement and High-Occupancy Vehicle— Initial Site Visits to Scope Efforts For an environmental study of a bridge project with variable pricing issues in the Puget Sound region of Washington State, analysts reviewed lists of social institu- tions, businesses, and public services that serve low-income populations using websites, phone directories, and comprehensive plans prepared by the region’s jurisdictions. Previously prepared technical environmental studies were also ref- erenced to identify resources of importance to low-income, minority, or limited English proficiency (LEP) populations such as social services, religious organiza- tions, schools, community centers, recreational facilities, and transit facilities. Businesses were identified that were owned by, served, or employed a number of low-income, minority, or LEP individuals. Windshield surveys were conducted to supplement and validate data. These initial investigations of resources culminated in a report that informed subsequent public outreach strategies. Source: PRR, 2009 The information obtained from the initial site visit and interviews should be woven into the collected social and economic demographics and inform the scope and scale of the PIP. Field visits provide an opportunity to hear the languages spoken on the street, experience some of the everyday transportation problems, notice the age of cars parked in residential driveways, see who works the second-shift, identify areas where people gather, and examine the absence or presence of foot traffic on the street. Community facilities that would be particularly convenient to reach for neighborhood residents and community-based organizations and other stakeholders should be noted. Most importantly, the observations and insights of the community impact practitioner and the public involvement professional should be shared with the project management team early enough to help shape the processes to come. Prepare a Limited English Proficiency Plan. Individuals who have a limited ability to read, write, speak, or understand English are considered to have limited English proficiency (LEP). An LEP plan describes the policies, services, and information that a government agency, includ- ing transportation agencies, will take to ensure that LEP persons have meaningful access to the agency’s programs and activities. An LEP plan will identify the size and locations of low-literacy populations and various foreign-born populations who may not speak English “very well” (see text box, SR 520: I-5 to Medina Bridge Replacement and High-Occupancy Vehicle—Identifying Language Translation Needs.) It will also describe the most appropriate approaches that can be taken by the governing entity to ensure that meaningful access is provided to all their programs and activities without imposing undue additional cost burdens. The need for an LEP plan is set forward in EO 13166, Improving Access to Services for Persons with Limited En glish Proficiency, which reaffirms Title VI of the Civil Rights Acts of 1964 and the prohibition of discrimination on the basis of national origin.

120 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Use “I Speak” Cards to Ensure Communications with LEP Populations. “I Speak” cards are two-sided bilingual cards that invite LEP persons to identify their language needs to transportation agency staff. Such cards, for instance, might read “I speak Spanish” in both Spanish and English. They may also include information about language access rights. These cards can be used to assist LEP people in communicating their need for interpretation and translation services. Offer Assistance for the Hearing Impaired. Some hearing impaired or deaf individuals can speak and/or read lips, while others may rely on American Sign Language or written and visual aids. Others may not be able to write or read well. The first thing to do when encountering a person with a hearing impairment is to identify how the person communicates best. The advent of telephone texting has allowed many to receive and send information of 160 characters or less through their telephones. Thus, it is important to obtain both an individual’s email address and telephone number. Telephone texting allows project information, including short surveys, to be sent to those who are hearing impaired and deaf and for them to respond in a like manner. The telephone’s vibration option provides them with notice that a message has been received. If pos- sible, an annotated agenda for any upcoming meeting or a copy of the proposed presentation with notes can be sent to an individual’s email address prior to the meeting. This will give the recipient a general idea of the topics under discussion and allow time to formulate any questions or com- ments for the project staff. When talking with individuals who are hearing impaired or deaf, the practitioner should always look directly at them and not at the individual who is signing the message or verbally relaying their response. If the person reads lips, the practitioner should not block the view of his/her face and should talk to them in a well-lighted area, speaking in a normal manner with short, simple sentences. When releasing any written information (e.g., press releases, newspaper articles, emails, websites, or newsletters), the practitioner should provide the TTY number and ask if anyone needs a signer to be present. Offer Assistance for the Sight-Impaired. Some sight-impaired or legally blind (20/200 vision) persons can distinguish colors and/or read large print while others may rely on Braille materials or their hearing. When choosing colors for a display, practitioners should be aware that some SR 520: I-5 to Medina Bridge Replacement and High-Occupancy Vehicle— Identifying Language Translation Needs In the designated study area, census tracts were identified that reported 5 percent or more of the population belonging to an ethnic group with a primary language other than English. Making reference to U.S. Department of Justice guidelines, and consistent with the Washington State Department of Transportation policy, the public involvement team used this criteria as a threshold for determining the need for translation into other languages. Thus, translations were generally warranted when an ethnic group with a primary language other than English comprises 5 percent or more of an area or represents 1,000 or more persons in an area. This criteria revealed several languages that were spoken at home by more than 5 percent of the population in Census tracts in the regional study area, including: African languages, Cambodian, Chinese, Korean, other Asian languages, other Pacific Island languages, Persian, Serbian/Croatian, Spanish, Tagalog, and Vietnamese. Source: PRR, 2009

Preparing, Implementing, and Assessing a Public Involvement Plan 121 people may be color blind and the name of the color should be written near it. The elderly may require information in a large print format. Those who have a computer with a speech compo- nent and Internet access can access websites that are Section 508 (1973 Rehabilitation Act, as amended in 1986) compliant. All federal agencies, and those agencies receiving federal funds or under contract with a federal agency, are required to comply with this law. For those who do not have computer access, radio reading services for the blind, public service announcements on radio and television, and news stories on radio and television are ways to get information to those that are sight-impaired or blind. When encountering those who are sight-impaired or blind at a public event, the practitioner should first introduce himself/herself and identify who he/she is and what role the practitioner plays on the project. The practitioner should be sure to describe information readily apparent to those who can see and should indicate that new items have been brought into the environment, describing what they are and where they have been put. The practitioner should offer to lead someone, but wait for them to accept his/her offer before proceeding, allowing them to hold his/ her arm rather than holding their arm so they can control their own movements. The practitio- ner should be descriptive when giving directions—“over there” has little meaning to someone who cannot see. The practitioner should instead say, “starting at the corner of Main Street, then going south and crossing Wales Street and Ivey Street. . . .” Practitioners should describe things from the perspective of the sight-impaired, not their own. Some people who are blind use a “clock” reference for things directly in front of them. If a blind person is accompanied by a guide dog, the practitioner should not interact with the dog while it is working (or in the harness). Offer Refreshments. Refreshments foster a more relaxed setting and put people at ease. Pro- viding food at a meeting can be a way to increase meeting attendance. It allows parents to pick up their child at a day care facility or at home and come directly to the meeting without having to eat first. When people go home first to eat supper, their willingness to attend a meeting may wane and they may remain at home. Having a meal at a meeting can provide an incentive for someone who is low-income to attend a meeting. Often having a meal at a meeting provides neighbors an opportunity to get together and becomes a reason to attend the event. Serving refreshments also provides a time and space for people unwilling to speak out in a crowd to have one-on-one discussions and ask questions in a less formal setting. When served in the middle of a meeting, refreshments can enliven, reinvigorate, or refresh a group that has become tired, bored, or frus- trated. Serving more substantial refreshments can also be a way to get around holding meetings at times that may conflict with meals. Brand Project through Clothing and Other Paraphernalia. Branding projects through clothing and other paraphernalia visually identifies members of the project team in the field or at public events. Distinctive clothing and accessories could include t-shirts, jackets, hats, and badges, among other items. These items draw attention to members of the project team, giving them an identity in places where they may not be known. Branding projects through fashion makes it easier for community members to see that outsiders have a clear purpose for being there. By making team members easily identifiable, they convey the message that they are approachable and open to receiving comments and questions from the public. This can also promote accountability among project team members because it instills in them the idea that they are representing the project to the public. Provide Information. Providing information to the public is the responsibility of transpor- tation agencies and applies to nearly all stages of transportation decision-making, including those activities pertaining to the planning, design, and implementation of toll-related facilities and operations. Low-income and minority populations, no less than other segments of the pub- lic, are entitled to interact with transportation agencies to (1) communicate their needs and concerns, (2) assess the potential impacts of government agency decisions, and (3) learn about

122 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox opportunities to influence decision-making processes. Providing information effectively may require an agency to critically assess its typical practices and adjust them, if needed, to better advertise events, describe activities in a way that clearly conveys coming changes or potential impacts, and work with affected communities, where warranted, to facilitate their informed involvement on projects that may affect them (see text box, SR 520: I-5 to Medina Bridge Replace- ment and High-Occupancy Vehicle—Strategies to Provide Information). SR 520: I-5 to Medina Bridge Replacement and High-Occupancy Vehicle— Strategies to Provide Information The public involvement team orchestrated several approaches to reach low- income and minority populations in order to provide information and invite feedback on the potential effects of a variable pricing project for its existing SR 520 Bridge, including: • Met with social service agencies throughout the Evergreen Bridge travelshed, which informed the regional study area. • Staffed project information booths at local fairs, festivals, and farmers’ markets that cater to low-income and minority populations. • Hosted public information meetings throughout the study region. Meetings were held in a wide variety of locations on different days of the week and at different times of day. • Posted flyers at transit stops and placed advertisements in publications that cater to low-income, minority, or limited English proficiency (LEP) populations. • Distributed mailings to minority-owned businesses from a purchased mailing list and offered briefings at minority-owned business coalitions. • Placed unstaffed information kiosks throughout the travelshed region at places serving target EJ populations, including libraries, community centers, community colleges, universities, senior centers, and community organizations. • Translated project materials into different languages, including Spanish, Russian, Chinese, Japanese, and Vietnamese. Some materials were translated into Amharic (Ethiopia), Laotian, Somali, and Tigrinya (Ethiopia and Eritrea). • Offered a speakers’ bureau to make presentations on tolling—how to obtain a transponder and set up an account. Source: PRR, 2009 Toll pricing schedules, account and transponder policies, and other features of tolling opera- tions can seem particularly complex for the public. Practitioners devoted to public outreach must avoid mystifying technical jargon and prepare information and outreach for “regular folks.” The challenge of presenting technical information clearly should not be minimized. The credibility of the agency’s decision-making process and the legitimacy of the proposed action can be under- mined by a poor communications strategy or insufficient transparency. Inadequate disclosure of information and insufficient candor about project issues can easily backfire, stoking opposition and controversy, which can cause the agency to lose control over its project as it endures delays, political wrangling, and even legal proceedings. New tolling operations can be poorly received by key stakeholders, the public, and the press when they are poorly briefed or when media strategies fail to prepare travelers for the coming changes.

Preparing, Implementing, and Assessing a Public Involvement Plan 123 Distribute Flyers. Flyers can effectively provide information to traditionally underserved populations because they provide flexibility in information dissemination. Flyers can be placed at community activity centers frequented by traditionally underserved populations, written in the language and tone that will best communicate to those populations. Activity centers where flyers are posted can include public buildings such as libraries and post offices, community and senior centers, places of worship, as well as local businesses such as grocery stores, hair salons, and cafes. Flyers can be posted for all to see, and copies of the flyers can be left for people to take with them. Flyers can also be distributed during “walk-throughs” in residential neighborhoods. Publicize through Local and Ethnic Media Outlets. Local and ethnic media outlets are key means for reaching populations not necessarily relying on traditional media outlets. “Local media” refers to neighborhood media such as weekly newspapers targeting a particular part of the town or a neighborhood. “Ethnic media” refers to media in a particular language, such as Spanish or Arabic, or English-language media directed to a particular ethnic group such as Asian Americans or Blacks. Local media will focus on neighborhood-related information. As a result, people in that neighborhood are very likely to read it because they know it will contain news about things that may directly affect them. Ethnic media outlets are tailored to the language and cultural interests of the group to which they are targeted. Many ethnic groups look for the media that is directed at them because they know it will have information about activities and persons that are likely to be of interest to them. The overall readership or listenership may be less than the larger, mainstream media outlets, but these outlets are relevant to particular populations and consequently the information can reach its intended audience (see text box, I-110 HOT Lane—Advertise by Holding Community Workshops). Advertise on Billboards, Marquees, and Variable Message Signs. Billboards and marquees are a way to display large-scale advertisements in highly visible places, such as alongside highways I-110 HOT Lane—Advertise by Holding Community Workshops Metro and Caltrans invited interested community members to three I-110 Con- gestion Reduction Community Workshop Meetings located along the corridor. The purpose of these meetings was to give interested parties the opportunity to participate in identifying options that would increase traveling efficiency in the corridor. • A total of 38,000 flyers advertising the community workshops were distributed along the I-110 corridor. • Metro advertised the community workshops in more than 20 foreign-language (Spanish, Chinese) and local English newspapers. • Four separate e-mail blasts to more than 1,000 persons on a contact list were notified. • An open house meeting format was followed. Metro, Caltrans, and con- sultants staffed information stations, where they discussed the project, answered questions, and recorded comments. Metro led the question and answer sessions. • One of the three meetings was held as an agenda item on the monthly South Bay Governance Council Meeting. Source: Caltrans, 2010

124 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox or on the sides of buildings. Billboards, marquees, and variable messaging signs can draw atten- tion and communicate a simple message to a captive community traveling through an area. They can be used to advertise for a public event, provide notification of upcoming construction, direct people to an online survey, or thank the community for their involvement. Because of their prominent placement and high visibility, billboards can be particularly effective in reaching groups that are not currently engaged in the topic and creating a buzz about the issue. Use Videos to Convey Information. Videos are a frequently used means for drawing atten- tion and making materials digestible to a wider audience. Videos can convey complex infor- mation on toll pricing projects such as the operations of the facilities and toll account and transponder policies. Videos may be particularly effective in reaching low-literacy and LEP per- sons. They can be disseminated widely to promote a message, frame issues of concern, or deliver information to stakeholders. They can be used to set the stage at public meetings and events, inform survey-takers or focus groups participants, or be shared with news media outlets or on social media (see Figure 1). Figure 1. Videos explaining transponder options for the Ohio River Bridges project.

Preparing, Implementing, and Assessing a Public Involvement Plan 125 Employ Visualization Techniques. Making visualizations an integral part of any presen- tation, newsletter, PowerPoint presentation, website, or newspaper article provides the public with a picture of what is actually being proposed. This increases the public’s awareness of the project, and allows individuals to consider how the project may affect their lives, communicate this information and awareness to others, and participate more fully in transportation decision- making. Depending on the size and complexity of a project and its budget, a variety of visual techniques can be used. If the project will convert high-occupancy vehicle (HOV) lanes to high-occupancy toll (HOT) lanes, before-and-after photos or visualizations of how the system may function can be used. This technique provides a relatively inexpensive way to show several alternatives at the same location and/or at different locations. If the project is a more complex widening, before-and- after pictures or a computer-generated series of different pictures (i.e., a morph) can be used. The morph representation starts out as a still photograph and then slowly adds features, such as additional lanes, a planted median, bike lanes, sidewalks, or bus pull-offs. This presentation can be repeated in 30-second cycles. If the project is a new multi-lane road, a computer-generated “3D drive through” can draw interest. The simulation can show what it would be like to drive through the new facility; however, this type of simulation is more expensive to produce. Flyover simulations are also being used. Animation, such as that shown below in Washington State, can be used to illustrate how various users might use an express toll lane and how its use could also benefit general purpose lanes (see Figure 2). Gather Feedback. Gathering feedback from all populations, including low-income and minority populations, is critical to formulating transportation solutions that will meet the needs of users and address the concerns of affected communities where facilities and services Figure 2. Washington State DOT animation illustrating its proposed I-405 Express Lanes project.

126 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox are to be sited. Traditionally underserved populations may represent a significant portion of the transportation network’s users or bear the burden of potential adverse impacts, yet oftentimes they are not heard from during the decision-making process. Tools and techniques for gather- ing feedback include methods for engaging those who do not attend traditional public events as well as creative mechanisms for collecting their input. Gathering feedback from the tradi- tionally underserved may require the broad application of a general technique (i.e., holding a meeting in every neighborhood of a study area) or efforts targeted toward specific populations (i.e., conducting focus groups). Conduct Outreach at Non-Traditional Locations. Holding formal and informal events and activities at non-traditional locations is an invaluable means for connecting to traditionally underserved populations. Depending on the targeted population, these locations will vary, but may include places of worship, community centers, social service agencies, settlement houses, senior centers, meeting rooms in apartment complexes, restaurants, hair salons or barber shops, feed stores, shopping malls, convenience stores, community fairs, sporting events, and any other place where traditionally underserved populations may congregate. Practitioners have repeatedly found that going to places where traditionally underserved populations meet, rather than waiting for them to come to an agency’s event, is more likely to make this population feel more comfortable (see text box, Louisville–Southern Indiana Ohio River Bridges—Surveys at Grocery Stores). Not all who are encountered will be interested in learning more about trans- portation matters, but there will be some in attendance that are willing to listen and give their candid feedback. Louisville–Southern Indiana Ohio River Bridges—Surveys at Grocery Stores To understand the toll-related impacts of the Louisville–Southern Indiana Ohio River Bridges projects on low-income and minority residents, project consultants went to four grocery stores in the project area. With the approval of the store management, on-site surveys were conducted with customers in exchange for an incentive. Source: KYTC and Indiana DOT, 2014a Go to “Their” Meetings. Co-sponsorship, participation in, or any other support of meetings held by advocacy groups, employers, and human service or public agencies that serve tradition- ally underserved populations, including low-income and minority populations, can effectively reach target populations “where they live.” Efforts to build partnerships with groups and agencies with expertise in working with the target groups can often build trust. Practitioners often begin by creating an asset map or database of the associations, employers, and institutions that work with the target populations in the study area. Representatives within such organizations can be called on to provide advice on issues affecting the target communities, set up meetings, or help make introductions to key individuals. It may also prove effective to work through these contacts’ media (e.g., newsletters, websites) to publicize events and exchange information. For example, practitioners may want to write short pieces for their newsletters, providing contact informa- tion or other facts about ongoing projects or request to add links to their websites regarding a proposed project or other action. It is important to recognize that the “meeting” may not actually occur at a gathering place. The real goal is to “reach people where they are” (see text box, Massachusetts DOT Turnpike

Preparing, Implementing, and Assessing a Public Involvement Plan 127 Massachusetts DOT Turnpike All-Electronic Toll Collection— Visiting the New Senior Centers On the occasion of the recent opening of Westfield, Massachusetts’ new Senior Center, Massachusetts DOT was the most popular of the exhibitors. Local seniors attending the event were provided a new E-ZPass Transponder. While the Turn- pike (including the Sumner and Ted Williams Tunnels and the Tobin Memorial bridge) currently has toll booths, MassDOT has plans to eliminate them later in 2016. Seniors who completed an application were handed a transponder to display on their vehicle. Those not opting to use the transponders will receive a bill in the mail and be charged a higher rate for video transaction once the all-electronic system is instituted. Source: LaBorde, 2016 Go to the Faith-Based Institutions. Faith-based institutions can be an effective venue for holding events and providing information to, and getting feedback from, the institution’s lead- ership and lay membership about transportation, social, or other community-related issues. Practitioners have found that working in partnership with the institution and/or seeking its endorsement can encourage participation, build support for plans and projects, or improve understanding of the effect of the project during the life of the project and future actions. Con- necting to the broader faith-based community affected by the project widens the range of con- tacts within the affected community and can act as a conduit for information exchange. Public meetings in faith-based institutions can establish the trust needed to conduct focus groups, inter- views, surveys, and the like, among various committees, boards, and subgroups (e.g., women, youth, “soup” kitchens) affiliated with the institution. The institution can also act as a partner in information dissemination and gathering. Faith-based institutions, in serving their constituents, often overlap and coordinate with human service agencies. Therefore, it is possible to find individuals in both the faith-based insti- tutions and in the social service agencies that truly understand and can empathetically express problems or issues confronted by local populations. Their knowledge and insights about the affected populations or their clientele is often effective in devising outreach and communications strategies that will make it possible to disseminate information and receive meaningful feedback (see text box, I-110 High Occupancy Toll Lane—Attending Conferences to Raise Awareness). Conduct Market Research Interviews and Focus Groups. Interviews and focus group meetings can be effective for exploring the transportation needs, travel behavior and patterns, and attitudes toward pricing of traditionally underserved populations. Especially for low- income persons and groups with limited literacy or English proficiency, understanding these needs, attitudes and travel patterns is appropriate for assessing the potential travel behavior impacts of various tolling-related policies. With proper planning, interviews and focus groups All-Electronic Toll Collection—Visiting the New Senior Centers). Feedback from the organiza- tion’s contacts and convened participants may come through informal discussions, structured interviews, or in response to presented plans or group discussions. It may help to request to be included on the agenda of meetings that the organizations may hold for their client groups. At such events, practitioners should be prepared to discuss information about the proposed proj- ect, solicit input, and describe the type of follow-up that will occur after the meeting.

128 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox can eliminate literacy, language, and cultural barriers. They can help practitioners develop a better understanding of how various population segments may use contemplated tolling facili- ties, which can vary significantly depending on the group (see text box, Minnesota DOT I-394 Express Lane Focus Groups, and tool, “Using Focus Groups in Assessing the Impacts of Tolling on EJ Populations”). Minnesota DOT I-394 Express Lane Focus Groups Minnesota DOT enlisted The Hubert H. Humphrey Institute of Public Affairs at the University of Minnesota to conduct a series of five focus groups to learn of the public’s perception and attitudes associated with implementing high- occupancy vehicle (HOV) buy-in capabilities on I-394. Source: Buckeye and Munnich, 2006 Undertake Surveys to Understand Needs, Preferences, and Impacts. Surveys and ques- tionnaires can be used to solicit information on travel behavior and attitudes toward toll pricing, understand preferences for various project alternatives or toll account options, and express views on the perceived impacts of various project alternatives, among other topics. They can be used to gather information remotely from a wide range of diverse stakeholders via telephone, email, or websites. They can also be used as a tool to improve direct commu- nications through intercept or in-person interviewing conducted in target communities or with specific stakeholders. Surveys and questionnaires are extremely versatile tools and can be implemented to gather information from a large and/or statistically significant population, as a convenience sample, or simply as a tool for starting and guiding individual conversations (see tool, “Designing and Implementing Surveys to Assess Attitudes and Travel Behavior for EJ Analyses and to Monitor Implementation”). They can also be used as tools for ascertaining the most effective ways to conduct other forms of public outreach in specific communities (e.g., What newspapers or publications do you read? What time of day is most convenient for you to attend meetings?). Use Computer-Assisted Technologies to Explore Preferences. Computer-assisted methods may, in some cases, assist in overcoming some barriers to participation by offering new avenues for participation. If structured appropriately, computer-assisted technologies can be (1) less I-110 High-Occupancy Toll Lane—Attending Conferences to Raise Awareness Metro and its outreach consultants staffed a booth at the Ward African Method- ist Episcopal Church, the 84th Session of the Southern California Annual Confer- ence; the Southern California Association of Non-Profit Housing; and Mobility 21, the 7th Annual Southern California Transportation Summit. Fact sheets in both English and Spanish were provided, and the participants were invited to submit their input via comment cards, registration forms, and exit surveys. Source: Caltrans, 2010

Preparing, Implementing, and Assessing a Public Involvement Plan 129 daunting than public meetings because supplementary, background information can be easily provided, (2) participation can occur at more convenient times (assuming online, open-hours access), and (3) online access can also help to overcome physical and/or geographic barriers. These tools can be designed to solicit input from stakeholders in a variety of settings from one- on-one individual interactions such as intercept survey using a digital tablet (see tool, “Targeting Local Grocery Stores to Administer Community Surveys, Louisville–Southern Indiana Ohio River Bridges Project”) to working with multiple respondents in the course of focus groups or neighborhood workshops. Computer-assisted methods can provide aggregate results quickly, present provocative graphics for visualization and maps, and support real-time interactivity such as electronic polling in a large group meeting. Not all populations have ready access to computers in their home or on a “smart” mobile device or have comfort with such technologies; therefore, practitioners should avoid overreliance on this approach because it may exclude some individuals or groups from participation. Build Relationships. Efforts to involve the public are sometimes criticized as being “too little, too late.” Citizens want to work with responsive public agencies that involve them in a meaningful, collaborative process from the outset, not just when they are upset and feeling left out halfway through a project. Building relationships with leaders from traditionally under- served communities, including representatives from places with higher concentrations of low- income and minority populations, can help agencies engage more effectively with traditionally underserved populations and, ideally, from the outset of the decision-making process. This may include mechanisms for initially “breaking the ice” and beginning a civil discussion, to holding a continuing dialogue throughout a project, or defining formal relationships and approaches for garnering trust and strengthening connections. Form Advisory Boards, Committees, Taskforces, and Working Groups. A group of volun- teers that meets regularly on a long-term basis to provide advice and/or support advisory com- mittees can be formed around specific geographic regions, a particular project’s stake holders, a special interest, or population group. These groups can include diverse stakeholders such as indi- vidual citizens, community-based organizations, elected officials, business owners, and others, including representatives from low-income, minority, or other traditionally underserved commu- nities (see text box, Minnesota DOT—I-394 MnPass Express Lane Project Community Task Force). Minnesota DOT—I-394 MnPass Express Lane Project Community Task Force Minnesota DOT determined that holding education and outreach processes to build greater understanding and public support for the I-394 project were needed. At the request of the Minnesota DOT staff, the governor and the lieu- tenant governor formed the I-394 Express Lane Community Task Force with members of the legislature, community leaders, interest groups, and concerned citizens. The task force was formed to help citizens and stakeholders better understand the project and its goals and provide a means for giving advice and guidance during the development of the project. At the same time, FHWA awarded a Congestion and Value Pricing pilot grant under which Minnesota DOT pursued a program of comprehensive planning, education, and outreach activities on value pricing. Source: Buckeye and Munnich, 2006

130 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Foster Understanding of Communities through Relationships with Community Organi- zations and Other Local Experts. Leveraging relationships with community organizations entails fostering—formally or informally—a working arrangement or alliance with social ser- vices organizations, faith-based institutions, community-based organizations, or other groups at the local level that regularly interact with or include members from traditionally underserved populations. These organizations can identify important individuals to contact, become an intermediary with other organizations, and act as a cosponsor on projects. In addition, these organizations can help distribute project information through their own membership, facilitat- ing input and feedback from members of the organization. In some cases, it may be very difficult to connect with underserved populations because of their work status or their difficult indi- vidual circumstances; in such cases, local experts may speak with sensitivity to these conditions or advocate on behalf of underserved populations for specific policies or projects because these experts tend to be highly informed about the community (see text box, Regional Express Lane Network—Metropolitan Transportation Commission Program). Regional Express Lane Network—Metropolitan Transportation Commission Program Informed by the demographic mapping of high concentration areas of low-income and minority populations, community-based organizations (CBOs) were selected as potential partners for performing focus groups. Participating community- based organizations were selected on the basis of their mission and connection to targeted EJ populations and their ability to host a focus group discussion and recruit 12 to 15 participants within the required time frame of the study. In put- ting together focus groups, community-based organizations were asked to screen participants who were low-income and/or minority individuals who travel in the subject corridors by public transit, carpools, or as solo drivers. Source: Resource Development Associates and Parsons Brinckerhoff, 2013 Working with the “right” organizations and individuals can ensure access to community leaders and encourage participation in planning and other transportation-related processes. Building partnerships with community organizations and other local experts can foster trust and be a valuable means for establishing long-lasting, two-way communications to begin to address critical issues interfering with effective public involvement. Recruit and Mobilize “Community Ambassadors,” “Beacons,” or “Trusted Advocates.” “Community ambassadors,” “beacons,” or “trusted advocates” are individual citizens or leaders who are capable of bridging the communication gap between agency practitioners and mem- bers of the public. They are individuals who are perceived by other members of the community as trustworthy, approachable, and effective. These ambassadors may be a member of a specific ethnic, racial, and/or cultural group with particular expertise in the culture, language, history, and values of the local community. They know who to contact and how to approach them, which makes it easier to get the word out about what is going on and how and why to partici- pate (see text box, Louisville–Southern Indiana Ohio River Bridges). A word-of-mouth approach is effective with most populations, but is especially effective with traditionally underserved

Preparing, Implementing, and Assessing a Public Involvement Plan 131 populations because the ambassador or beacon is someone they know and trust to give them good advice. The relationships are already established and people rely on the network to give them good information (see case example, “Mobilizing a Local Liaison to Recruit Community Leaders for Survey, Louisville–Southern Indiana Ohio River Bridges Project”). Assess PIP Effectiveness. The effectiveness of PIP should be periodically assessed to deter- mine if the goals and objectives set forward in the PIP were achieved. The PIP should include specific strategies for reaching low-income, minority, and other affected traditionally underserved populations (see text box, EJ Reference Guide on Documenting Public Involvement). The assess- ment should determine if the implemented involvement strategies were effective in reaching each of these populations and, in fact, whether the events and processes created opportunities for meaningful involvement. It is entirely possible that it may be necessary to do something differ- ent to reach and engage affected protected populations particularly in cases where there may be disproportionately high and adverse effects. The evaluation of the effectiveness of PIP in identifying and engaging EJ populations assumes that the plan’s developers have determined, either formally or informally, the need to reach out and involve potentially affected low-income and minority populations. The inclusion of the affected population in the development of effectiveness measures aids in identifying goals, objectives, and practices that are meaningful to both planners and the community. The assess- ment may be done at different stages of the project (e.g., project planning, detailed design and construction documents, construction) as well as after implementation in the monitoring of ongoing operations. Document “Everything.” There are many ways to “tell the story” of the public involvement, the community response, the issues that were identified, and how the issues were resolved in collaboration with the impacted community. The practitioner may choose or be directed to develop a project diary (as used by Florida DOT), a Community Impact Assessment Report (fairly common), or follow other reporting procedures using templates established by the DOT or MPO. The process, efforts, and results must be rigorously documented for stakeholder and public review and to protect and defend the integrity of the project and process. Agencies and practitioners should be mindful that thorough documentation of public involve- ment processes conducted with minority and low-income populations is an integral element of the environmental justice analysis and findings conducted during the environmental review stage. Those at FHWA responsible for environmental review will seek evidence in the narrative that sponsoring agencies have incorporated public input into analyses and decisions and that commitments are documented. Louisville–Southern Indiana Ohio River Bridges The Kentucky Transportation Cabinet hired a well-known local liaison with deep ties to and trusted relationships in the EJ communities to identify and encourage local leaders to participate in a survey to understand the impact of the proposed tolling of the Louisville–Southern Indiana River Bridges project on the local EJ communities. Source: KYTC and Indiana DOT, 2014b

132 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox What Are Its Limitations? In the context of EJ, full and fair participation means everyone has a clear path to the table and, once there, accorded the same rights as everyone else there to be heard and treated respect- fully. However, despite the good-faith efforts of practitioners, the strategies included in the PIP may prove ineffective in reaching the targeted affected protected populations or conveying information in ways that foster meaningful involvement. Moreover, there are instances where staff or decision-makers within transportation agencies prefer a specific alternative or decision, and seek to “manage” their interactions with the public toward that outcome. In such cases, they may adopt a public relations approach that can “spin” information, or adopt one-way forms of information dissemination—rather than foster two-way processes that seek to consult, involve, and collaborate with the affected public. By comparison, fostering “meaningful involvement” requires finding strategies to facilitate opportunities for participants to get more informed so that they may express their concerns and influence decisions taken by governing agencies, particularly on matters that could affect their community’s health, safety, and well-being. In this spirit, strategies should be proactive and recognize the abilities and barriers the public may face in participating, be respectful of the life styles and cultures of the varied public, and seek to overcome factors that may lead to distrust in transportation decision-making. EJ Reference Guide on Documenting Public Involvement Planning regulations in 23 CFR 450.210 and 23 CFR 450.316 require state DOTs and MPOs to document their public involvement process through Public Involvement Plans (PIPs) and participation plans, respectively. In both cases, agencies must “seek out and consider the needs of those traditionally under- served by existing transportation systems, such as low-income and minority households.” Agencies must provide sufficient opportunity for the public and interested parties to provide input on public involvement processes and participation plans. State DOTs and MPOs must establish these public involvement/participation plans before developing any subsequent planning documents. “State DOT public involvement plans and MPO participation plans should include: • Strategies for involving minority populations, low-income populations, other protected groups, and the required ‘interested parties’ in transportation decisionmaking. • Strategies to reduce participation barriers for minority and low-income populations. • Outreach to organizations representing minority and low-income populations. • Mechanisms to ensure documentation and consideration of issues raised by minor- ity and low-income populations. • Periodic review of the effectiveness of EJ strategies and tracking of mitigation measures.” Source: FHWA, 2015

Preparing, Implementing, and Assessing a Public Involvement Plan 133 What Types of Resources and Costs Are Required? The costs for the preparation and implementation of PIPs are shaped by several factors. Agen- cies will undertake very different strategies for promoting public participation depending on the stage of decision-making, goals of the project, controversy anticipated by the project, the role envisioned for the public in shaping the decision-making process, and the level of resources that the sponsoring agency is prepared to make. The “spectrum of public participation” (see Table 2), developed by the International Association of Public Participation, is one means for clarifying the agency’s goals for public participation and then aligning these goals with appropriate promises to the public and appropriate techniques. For most tolling projects moving through the environ- mental review phase, the public participation goal is likely to fall short of “empower” along the participation spectrum. But there are projects in the planning and implementation stages and instances where the project leadership may choose to commit to more collaborative forms of public engagement (e.g., in the selection of alternatives or mitigation options). Public Participation Goal To provide the public with balanced and objective information to assist them in understanding the problem, alternatives, opportunities, and solutions. To obtain public feedback on analysis, alternatives, and/or decisions. To work directly with the public throughout the process to ensure that public concerns and aspirations are consistently understood and considered. To partner with the public in each aspect of the decision, including the development of alternatives and the identification of the preferred alternative. To place final decision-making in the hands of the public. Promise to the Public We will keep you informed. We will keep you informed, listen to and acknowledge concerns and aspirations, and provide feedback on how public input influenced the decision. We will work with you to ensure that your concerns and aspirations are directly reflected in the alternatives developed and provide feedback on how public input influenced the decision. We will look to you for advice and innovation in formulating solutions and incorporate your advice and recommendations into the decision to the maximum extent possible. We will implement what you decide. Example Techniques • Fact sheets • Websites • Open houses • Public comment • Focus groups • Surveys • Public meetings • Workshops • Deliberative polling • Citizen advisory committees • Consensus building • Participatory decision-making • Citizen juries • Ballots • Delegated decisions Source: International Association of Public Participation, 2007 Increasing Level of Public Impact Inform Consult Involve Collaborate Empower Table 2. The spectrum of public participation.

134 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox In light of these varying factors, it is difficult to generalize the cost for the preparation and implementation of PIPs because the cost can vary significantly. Drafting a PIP for the Buford Highway (8 miles) project in Atlanta, for example, took just three days to prepare and was put together at a cost of $6,000. However, a far more significant outreach process, integrated into the Business 40 (1.1 miles) project in Winston–Salem, led to the initial development of a PIP that took more than 30 days and cost approximately $100,000 (Aimen and Morris, 2012). Imple- mentation of the PIP is in addition to the preparation of the PIP. Conducting surveys and other outreach can cost far more than that (see tools on focus groups and conducting surveys and case examples on citizen panels, surveys with local liaison, and surveys at grocery stores for examples on the level of effort for implementation activities). The level of effort for the preparation of a PIP is affected by the quality of previously collected and available data, the implications of that data, the field verification of that data and local condi- tions, and the extent to which additional data sources must be examined to address information gaps to plan for public outreach processes. The cost of preparing and implementing the PIP is also driven, in part, by the results of initial scoping field visits and “beta-testing” of outreach efforts. For example, the public’s responses to questions related to “meeting convenience” (e.g., what time of day/night, what day of the week/weekend, what location) and “project concerns” could shape strategies for subsequent outreach activities. When such responses are solicited, it can inform the basis for subsequent effectiveness assessments of the PIP (e.g., it can be used to demonstrate that concerns were heard and addressed as the project progressed through various decision-making stages). Many different practical approaches can be employed to implement an effective PIP: • Train community members. One practical approach for reaching EJ populations is mobi- lizing and training persons from the affected area to assist in performing various forms of public outreach. This may include distributing flyers, survey-taking, and/or holding or participating in a meeting, among other activities (Aimen and Morris, 2012). Once trained, a salary of $15.00 per hour may be appropriate (in many parts of the country) to provide sustainable compensation and reduce turnover during the subject outreach process phase. • Interpreters and translators. The cost of employing an interpreter and translator will vary by location, skills, training, and whether the person performing the task is volunteering their time, working on a personal services arrangement, or affiliated with a professional firm or organization. • Clothing and other paraphernalia. Often a project is branded with a color or a logo. Depend- ing on the number of shirts required, these can run $15.00 or more. On projects that last over several months, more than one shirt may be necessary for each individual. • Incentives. To increase the number of people willing to complete a survey, incentives may be necessary. The cost of these can vary widely from a drawing for an iPad or thumb drive, a children’s book or toy, a gift card, or groceries or a culturally appropriate meal. Who Has Used It Successfully? Several examples of practical approaches that have been employed on tolling projects to seek to establish meaningful involvement opportunities and full and fair participation have already been highlighted. Table 3 summarizes many of the practical approaches discussed here and how they can be used to support analysis of EJ in tolling. Further information on these approaches can be found by examining the references presented in the resource section.

4 Develop a social and economic profile • Identify populations • Prepare and implement PIP • Informs the development of the PIP • Statewide/Metropolitan Planning • Project Development/ National Environmental Policy Act (NEPA) 4 Define the project and study area • Identify populations • Prepare and implement PIP • Informs the development of the PIP • Project Development/NEPA 4 Conduct a community characteristics inventory • Identify populations • Prepare and implement PIP • Informs the development of the PIP • Identify community facilities for outreach events • Statewide/Metropolitan Planning • Project Development/NEPA • Maintenance and operations 4 Up-front site visits to establish the scope of the PIP • Prepare and implement PIP • Validates or invalidates data collected and can identify new information • Provides an opportunity to talk with residents about past history and present conditions • All stages 4 Develop and maintain a community contacts database • Prepare and implement PIP • Strengthens community’s network and expands practitioner’s knowledge of community organizations/leaders • Ensures information is being transmitted to as wide a range of community members as possible • All stages 4 Prepare a limited English proficiency (LEP) plan • Prepare and implement PIP • Identifies the size and location of LEP populations • Describes the most appropriate approaches to ensuring meaningful access • All stages Step Tools and Techniques Task Objective How it Supports Analysis of EJ in Tolling Stage of Decision-Making Table 3. Practical approaches for reaching low-income, minority, and other traditionally underserved populations— tools and techniques that support analysis of EJ in a tolling context. (continued on next page)

4, 5, 6, 8 Use "I Speak" cards to ensure communications with LEP populations • Prepare and implement PIP • Assists LEP populations in communicating their needs for interpretive and translation services 4, 5, 6, 8 Offer assistance for hearing and sight-impaired persons • Prepare and implement PIP • Provides opportunities to participate more fully in decisions • Improves awareness of how decisions can affect their everyday lives • Communicates information/awareness to others 4, 5, 6, 8 Offer assistance for low-literate persons • Prepare and implement PIP • Provides opportunities to participate more fully in decisions • Improves awareness of how decisions can affect their everyday lives • Communicates information/awareness to others 4, 5, 6, 8 Brand project through clothing and other paraphernalia • Prepare and implement PIP • Makes it easier for community members to see that outsiders have a purpose for being there • Makes outsiders more approachable and invites comments/questions from the public 4, 5, 6, 8 Offer refreshments • Prepare and implement PIP • Helps increase attendance at meetings • Fosters a more relaxed setting and puts people at ease 4, 5, 6, 8 Use videos to convey information • Provide information • Can be an engaging format that encourages participation • May be more easily digestible to a wider audience 4, 5, 6, 8 Distribute flyers • Provide information • Provides flexibility in information dissemination • Conveys information that is clearly visible. 4, 5, 6, 8 Advertise on billboards, marquees, and variable message signs • Provide information • Draws attention and communicates a simple message to a captive audience traveling through an area • All stages • All stages • All stages • All stages • All stages • All stages • All stages • All stages Step Tools and Techniques Task Objective How it Supports Analysis of EJ in Tolling Stage of Decision-Making Table 3. (Continued).

4, 5, 6, 8 Publicize through local and ethnic media outlets • Provide information • Reaches populations that would not necessarily be reached using mainstream media outlets 4, 5, 6, 8 Employ visualization techniques • Provide information • Increases understanding of the project, broadens awareness of how decisions affect their lives, and helps communicate this information and awareness to others 4, 5, 6, 8 Conduct outreach at non- traditional locations • Provide Information • Gather feedback • Effectively involves traditionally underserved populations in locations where they feel safe and comfortable • Increases likelihood of deeper interaction between agencies and the public 4, 5, 6, 8 Go to "their" meetings • Provide information • Gather feedback • Raises awareness of tolling project plans • Receives feedback on potential impacts or preferred mitigation solutions 4, 5, 6, 8 Go to the schools • Provide information • Gather feedback • Raises awareness of tolling project plans • Statewide/Metropolitan Planning • Project Development/NEPA • Maintenance and Operations 4, 5, 6, 8 Go to faith-based institutions • Provide information • Gather feedback • Raises awareness of tolling project plans • Receives feedback on potential impacts or preferred mitigation solutions 4, 5, 6, 8 Apply social and new media appropriately • Provide information • Gather feedback • Represents innovative approaches with accessible content 4, 5, Conduct market research • Gather • Helps practitioners develop a better understanding of • Statewide/Metropolitan 6, 8 interviews and focus groups feedback how various population segments access transportation services and travel Planning • Project Development/NEPA • All stages • All stages • All stages • All Stages • All Stages • All Stages (continued on next page)

4, 5, 6, 8 Undertake surveys to understand needs, preferences, and impacts • Gather feedback • Allows extreme versatility in ways to gather information • Can enable better understanding of potential impacts 4, 5, 6, 8 Use computer-assisted technologies to explore preferences • Gather feedback • Can be less daunting than public meetings, occur at more convenient times, and overcome physical and/or geographic barriers by using online access • Statewide/Metropolitan Planning • Project Development/NEPA 4, 5, 6, 8 Form advisory boards, committees, taskforces, and working groups • Build relationships • Enables transportation agencies to better understand what the community’s needs and issues are; how to reach out to these communities at large; and how to avoid, minimize, or mitigate impacts 4, 5, 6, 8 Foster understanding of communities through relationships with community organizations and other local experts • Build relationships • Facilitates input and feedback from members • Creates opportunities to conduct outreach to members • Strengthens relationships with local liaisons 4, 5, 6, 8 Recruit and mobilize community ambassadors, beacons, or trusted advocates • Provide information • Build relationships • Engages those possessing particular expertise in who and how to approach the members of the community • Utilizes existing relationships and networks • Statewide/Metropolitan Planning • Project Development/NEPA • Operations and maintenance 4, 5, 6, 8 Assess PIP effectiveness • Prepare and implement PIP • Mandates re-evaluation of successfulness in achieving goals and objectives as internal and external dynamics change • All Stages • All stages • All stages • All stages Step Tools and Techniques Task Objective How it Supports Analysis of EJ in Tolling Stage of Decision-Making Table 3. (Continued).

Preparing, Implementing, and Assessing a Public Involvement Plan 139 Resources Aimen, D. and Morris, A. 2012. NCHRP Report 710: Practical Approaches for Involving Traditionally Underserved Populations in Transportation Decisionmaking. Transportation Research Board of the National Academy of Sciences, Washington, DC. Buckeye, K. R. and Munnich, L. W., Jr. 2006. “Value Pricing Education and Outreach Model: I-394 MnPass Community Task Force.” Transportation Research Record: Journal of the Transportation Research Board, No. 1960, Transportation Research Board of the National Academy of Sciences, Washington, D.C., pp. 80–86. California Department of Transportation (Caltrans). 2010. The Interstate 110 (Harbor Freeway/Transitway) High Occupancy Toll Lanes Project Draft Environmental Impact Report/Environmental Assessment. FHWA. 2015. Federal Highway Administration Environmental Justice Reference Guide. Retrieved from https:// www.fhwa.dot.gov/environment/environmental_justice/resources/reference_guide_2015/fhwahep15035.pdf. International Association of Public Participation. 2007. IAP2 Spectrum of Public Participation. KYTC and Indiana DOT. 2014a. Appendix E3, Louisville-Southern Indiana Ohio River Bridges Project EJ Community Survey Populations. KYTC and Indiana DOT. 2014b. Appendix E4, Louisville-Southern Indiana Ohio River Bridges Project, EJ Community Leader Survey. KYTC and Indiana DOT. 2014c. Appendix E5, EJ Comparison of Surveys Conducted to Identify EJ Populations’ Perceptions of Tolling Options and Potential Mitigation for Disproportionate Adverse Effects of Tolling on Low-Income Commuters and Travelers in the Project Area. LaBorde, T. 2016. “E-Z Pass Transponders Popular at Westfield Senior Options 2016”. Masslive. Retrieved from http://www.masslive.com/news/index.ssf/2016/04/e-z_pass_transponders_popular.html. PBS&J. 2006. How to Engage Low-Literacy and Limited-English-Proficiency Populations in Transportation Decision- making. Federal Highway Administration, Washington, D.C. Prozzi, J., Victoria, I., Torres, G., Walton, C. M., and Prozzi, J. 2006. Guidebook for Identifying, Measuring and Mitigating Environmental Justice Impacts of Toll Roads. TxDOT Project 0-5208: Evaluation of Environmental Justice Aspects of the Tolling of Existing Non-Toll and Toll Roads. PRR. 2009. SR 520: I-5 to Medina Bridge Replacement and HOV Project Supplemental Draft EIS: Environmental Justice Discipline Report. Prepared for Washington State DOT and FHWA. Rash, J. 2015. I-405 Express Toll Lanes Part 2: A New Option for 450,000 People Stuck in Traffic. Retrieved from http://wsdotblog.blogspot.com/2015/03/i-405-express-toll-lanes-part-2-new.html. Resource Development Associates and Parsons Brinckerhoff. 2013. Regional Express Lanes Network—MTC Program: Final Focus Group and Intercept Survey Summary Report, Prepared for the Metropolitan Trans- portation Commission.

140 What Is It? Focus groups are a qualitative research method involving small group discussions led by a trained moderator (also referred to as a facilitator). The qualitative nature of focus groups comes from the data typically being collected in the form of the “why” behind people’s thoughts, feelings, and behaviors, rather than the quantitative number of people who think, feel, or behave in a particular way. As such, focus groups seek to get at an authentic and core values level of analysis. Because this level of analysis requires time for deep probing, focus groups typically have no more than six to ten participants and last from 90 to 120 minutes. One alterna- tive to the size of a focus group is to conduct a series of mini groups, each including three to four participants. Why Is It Effective in Environmental Justice Analysis? Assessment of impacts of tolling on low-income and minority populations as part of an EJ analysis requires not only valid information on the number of people affected, but also an understanding of the essence of those impacts and what meaningful mitigation would be for those affected. Consequently, focus groups are often used in conjunction with quantitative survey research. In some cases, focus groups may be conducted prior to a survey as a form of formative research to inform the content of the survey questions. In other cases, focus groups follow the survey and function to shed further light on the quantitative findings. Focus groups can be particularly effective in obtaining input from those with limited English proficiency. One may obtain deeper insights than from survey research, even if the survey is conducted in languages other than English. This is the case because conducting surveys in languages other than English has many challenges, including, but not lim- ited to, increased costs (e.g., for translations, phone number t o o l 5 Using Focus Groups in Assessing the Impact of Tolling on Environmental Justice Populations Framework Step • Scope Approach to Measure and Address Impacts • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Participant Recruitment • Participant Incentives • Partnering with Community Based Organizations • Facility Features • Moderator Guide • Moderator Style • Analysis and Reporting Affected Populations • Low-Income • Minority Examples Featured • Washington State DOT, I-90 Bridge Tolling • Washington State DOT, SR 520 Bridge Tolling • Metropolitan Transportation Commission, San Francisco, Regional Express Lane Network • Santa Clara Valley Transportation Authority, SR 85 Express Lanes • Elizabeth River Crossings and Virginia DOT, Elizabeth River Tunnels Tolling

Using Focus Groups in Assessing the Impact of tolling on Environmental Justice Populations 141 Articulation Question “And now, a quick question just for fun. If you could invite any person to dinner, living or deceased, who would it be and why?”[RECORD VERBATIM RESPONSE] • If respondent says they “do not know,” or gives a short response without elaborating, thank and terminate. • If respondent has difficulty hearing, is hard to understand, has poor language skills, or has an extremely heavy accent (for English language focus groups), or you have the slightest doubt as to his/her ability to communicate, thank and terminate. samples targeted by surname, programming in multiple languages) and small sample sizes (unless oversampling is involved). Increasingly, there is interest in online focus groups to reach individuals from a wider geo- graphic area, enabling interested sponsors to receive feedback from persons who might not otherwise be able to attend an in-person focus group. What Are Some Techniques for Implementing This Tool? Conducting focus groups will vary by the nature of the topics being explored, the composi- tion of the target audiences, the style of the focus group moderator, and the available budget. However, the following provides general guidelines that apply to conducting most focus groups, including those with EJ populations. Participant Recruitment. Obtaining input from EJ groups is critical but somewhat chal- lenging. Members of such groups are often more difficult to reach and may require working in languages other than English. • Lists, in which those from some minority groups are identified by surname or live in predomi- nantly minority group sections of a city, can be purchased through sample vendors. Many focus group facilities have their own lists (known as panels) of people who have volunteered to participate in focus groups and can be sorted by race/ethnicity and income. Other useful techniques for recruitment include partnering with community-based organizations that pro- vide services to specific minority and low-income groups, advertising in ethnic newspapers, and outreach through social media. • Recruiting from low-income households, especially those at some percentage of the poverty level, can be done by setting minimum income levels for households of particular sizes. For example, the U.S. Census considered a family of four with an income of less than $23,850 in 2014 to be in the poverty bracket. (However, this metric is applied uniformly across the United States regardless of the local cost of living, which can vary widely. In some higher cost metropolitan regions, one potential solution is to set the criteria for qualifying as “low- income” as some percentage above the poverty level, such as 200% of the poverty level). • A recruitment screener document is developed that contains criteria such as race, ethnicity, income, and any other criteria to ensure participants are qualified as members of EJ groups. The screener is then used for contacting potential participants. Recruitment screeners often include an articulation question to ensure participants can actively contribute to the group discussion (see text box, Articulation Question for an example).

142 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox • If written materials will be used in the focus group, it is important to include a question in the screener that confirms that the person can both speak and read in the language in which the group will be conducted. • It is best to over-recruit for each focus group since not all potential participants will show. For a group of 8 participants, it is standard to recruit 11 or 12. Incentives for Participants. Recruitment efforts will improve when an incentive is offered to compensate participants for their insights. Cash is often the best incentive, although gift cards can be also be used. When considering the amount of the incentive, it is reasonable to include the cost of travel to the focus group location, parking costs, and the potential need for participants to arrange for childcare. Incentives between $75 and $100 per person are standard, although higher amounts are used for particularly difficult-to-recruit minority groups, as well as for busi- ness owners. Number of Focus Groups to Conduct. At least two focus groups for each target audience are recommended in order to validate what was heard in the first group. For example, when conducting focus groups for the I-405 Express Toll Lanes in Washington State, four focus groups were conducted, two with single-occupancy users and two with carpool users. Location of Focus Groups. Conducting focus groups in formal focus group facilities offers advantages because these facilities routinely handle hosting (e.g., room set up, sign-in and distributing incentives, parking validation, providing food for participants), audio and video recording, and one-way mirrors for unobtrusive observation. However, many communities do not have formal focus group facilities within easy traveling distance for participants, and, even when they are available, other types of facilities may be more familiar and inviting for particu- lar EJ population segments. For example, even though there are numerous formal focus group facilities in Seattle, many researchers choose to conduct their Spanish language focus groups at El Centro de la Raza, a community center well known to the Hispanic community. Another option is to hold the focus groups in a local conference center or hotel conference room that is equipped to deliver a live video feed to an adjacent room so that the focus group can be observed on a monitor. This option can be fairly expensive since it requires hiring a video technician, and hotels typically require that any food served must be purchased from them. Conducting focus groups in locations without one-way mirror observation rooms or live video feeds means that those who want to observe must be seated (out of the way) in the same room as the participants. Although it may seem that the presence of observers in the same room might influence the discussion, many moderators find that participants quickly forget or ignore the observers. In those situations where the presence of observers in the same room would be too disruptive, one can use a high-quality “baby monitor” so those in another room can hear the conversation. Online focus groups present another “place” to hold focus groups, and these present some advantages and challenges. A big advantage is that recruitment can be easier because people can participate from the comfort of their homes. The focus groups also can include people from a broad geographic area because travel to a location is not involved. There is also the potential that people will be more willing to discuss sensitive topics more candidly when not in the same room as the other participants. Advances in online focus groups are happening quickly and moving in the direction of being more than online chat sessions where participants simply type in their ideas. There are companies that enable participants to upload documents needed for the discus- sion and then talk to and see each other. Finally, online focus groups are less expensive because there are no facility rental fees, no food costs, and participant incentives can be lower. Some disadvantages include that only those with computers equipped with cameras can participate,

Using Focus Groups in Assessing the Impact of tolling on Environmental Justice Populations 143 and such groups work best with fewer participants (four to five) because the technology works best with fewer people (at least for now). Developing Moderator Guides. The most important thing to keep in mind about the mod- erator guide is that it is a guide and not an exact script that needs to be read word for word. A good moderator will assess the group dynamics and adjust the wording of the discussion ques- tions, and even the order in which the questions are asked, to maintain a conversational tone to the discussion. Doing so means all of the topic areas from the moderator guide are still covered, but in a way that promotes deep insights. Other important guidelines to follow to reach deeper insights include: • Not trying to cover too much in the typically 1.5- to 2-hour period, • Building in additional probing questions relative to each main question, • Identifying how long each section of the discussion should take and writing down the time each section should start (this will help the moderator stay on time and avoid having to rush through later sections), and • Building in points in the moderator guide where the moderator stops to ask observers if there are any additional questions that they want asked. Moderating Focus Groups. Although each moderator will bring his/her own experience and style to the focus groups, there are characteristics that are common to good moderators: • Being curious about how people think, feel, and behave, while remaining objective or detached so as not to bias the discussion. This approach gets people to open up and talk and may gener- ate important “on-the-fly” probes. • Being an excellent listener while juggling multiple “moderator balls” in the air at the same time, such as keeping track of time available to ensure all topic areas are covered, reordering topic areas from the moderator guide in the moment to encourage discussion and sharing of ideas, and maintaining control of the conversation to make sure all topic areas are covered and discussion remains focused. • Being personable and putting people at ease by striking the right balance between being the “professional moderator” and “a nice person.” This may involve knowing when and how much to feign ignorance about the focus group topic as a means of drawing people out and obtaining more in-depth information. Ensuring No One Dominates the Group and that “Group Think” Does Not Happen. Experienced moderators are particularly aware of their responsibility to ensure that no one indi- vidual dominates the focus group and the diversity of voices and opinions are explored. Group discussion may be stifled when someone dominates the group, causing other participants to not share conflicting viewpoints and to be overly influenced (often unconsciously) by the viewpoints of the dominating group member. Nonetheless, focus group members are influenced by the ideas of other group members, and, in fact, one of the benefits of focus groups is the synergy of ideas as the discussion occurs. But it is also useful to capture focus group member’s ideas before they are influenced by the group discussion. This can be done though several techniques, including, but not limited to: • Pre-group online surveys to get participants’ initial impressions about key issues regard- ing the focus group topic. This step has the added benefit of giving the moderator a good idea of where the group stands before it convenes and can inform the development of the moderator guide. • Pre-group participant diaries in which participants record information (through text and images) about their relationship to the focus group topic. For example, focus groups comprising

144 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox those below the poverty line could ask participants to record what they think, feel, and do about the impact of tolls on their household budget in the week prior to the focus group. Participants may be more willing to share personal reflections in a diary format than sharing in the group setting. Such personal reflections also serve to crystallize a participant’s experience so, if they choose to share in the group setting, they can do so in a more concise and accurate way. • In-group exercises to obtain individual participant perspectives prior to group discussion, such as with the use of “Mind Maps” and “Word Bubbles” (see Figure 1). These exercises are detailed in the book Moderating to the Max: A Full-Tilt Guide to Creative, Insightful Focus Groups and Depth Interviews (Bystedt, Lynn, and Potts; 2003). Conducting in Languages Other than English. This is often necessary to ensure those with limited English proficiency have an opportunity to provide their input. Translation of the mod- erator guide (and other needed documents) is relatively easy, but to ensure accuracy, it should include a double-translation. This involves having one person translate (e.g., from English to Vietnamese). Then a different person translates the Vietnamese version back into English and the two English versions are then compared. More complicated and expensive is recruiting in languages other than English. Most importantly, it is essential that experienced moderators are used. Similar to English-language moderators, just because someone speaks the language does not mean they will be a good focus group moderator. What I’d be thinking: What I’d be feeling: What I’d say to another person: Figure 1. Word Bubbles can help participants organize their thoughts and provides pre-discussion opinions.

Using Focus Groups in Assessing the Impact of tolling on Environmental Justice Populations 145 Analyzing Focus Group Data. The primary task is to identify common themes that emerge from the group discussions. Since clients often need results quickly, the following steps can be useful for streamlining the analysis phase: • Conduct a “brain dump” immediately after each group. This can be done by the moderator and/or the note taker. The purpose here is to get initial impressions down in writing. Brain dumps also serve to identify any areas of the moderator guide that may need changing before the next focus group (see Figure 2). • Conduct a verbal debrief with the client the day after each focus group. The brain dump docu- ment can serve as the summary for the client debrief. • Review the audio/video recordings to flesh out the brain dump document and capture obser- vations and quotes that may enliven reports (see text box, Using Quotes to Bring the Focus Group Report to Life). Writing Focus Group Reports. Given the steps outlined above, full focus group reports are often anti-climactic. Nonetheless, they serve a useful function for those who did not observe the focus groups or attend the client debriefs. The report also allows for the presentation of similari- ties and differences across the focus groups. Finally, the report is the place where the findings can be used to present actionable recommendations. Digital copies of reports can include audio and/or video clips from the focus groups. What Are Some Techniques for Implementing This Tool? Understanding the impacts of tolling on EJ groups can be obtained through a variety of research methods. The following information showcases how five different projects used focus groups as part of their assessment process. Example 1: I-90 Bridge Tolling. Washington State DOT conducted focus groups as part of an EJ analysis on the effects of tolling I-90 users. More specifically, the focus groups were part of an effort to obtain in-depth information about the potential benefits and effects of tolling on I-90 users, including low-income users. Figure 2. Reviewing the video recording from the group will bring to light things that may have been missed in the initial “brain dump.”

146 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Using Quotes to Bring the Focus Group Report to Life To provide a sense of what happened at the focus groups, verbatim quotes from the events may be used to illustrate key findings. “It would be an extra $170 a month for me; that is too much.” “I can’t afford it, and I would have to find another job.” “My work hours are not flexible, it would affect me greatly.” “Taking the bus adds a terrible amount of time.” “Only if there is an Express bus (currently there is no direct service to where I need to go; I need to take more than one bus); also at the time I need to take it is too crowded.” “I just like the fact that I would be able to utilize it in dire need, getting some- where quickly. If I woke up late or something and need to get to work.” “When I need to get somewhere time is priceless to me.” “If you want to tax me some more and it’s going to go to the good of all roads for all the people that’s one thing. If you’re going to have this elitist solution, I think it’s terribly wrong.” “Putting the whole thing into play makes me wonder how are you going to edu- cate people so they’ll know how to do all this? I can just see mass confusion when you start this, and people are just stopped dead on the freeway going, can I get in, can I get out?” The recruitment screener was designed to identify appropriate participants for three differ- ent focus groups. For all the focus groups, participants were screened to confirm that they used the I-90 Bridge. One group was screened to qualify as low-income according to federal poverty guidelines, another group was conducted in Spanish, and the third group was conducted in English. Several steps were taken to encourage attendance, including: • The Spanish language group was held at a community center well known to the Hispanic community. • Each participant received a $75 stipend, plus $20 to cover any transportation and childcare costs. • A light meal was provided at each group, and parking was validated for groups held at the formal focus group facility. Example 2: SR 520 Bridge Tolling. Washington State DOT conducted research on the potential effects of tolling the SR 520 Bridge in Seattle on low-income and minority popula- tions. It developed a three-pronged approach that included a transit-intercept survey of people who used transit routes that cross the SR 520 Bridge, a telephone survey of SR 520 Bridge users, and three focus groups with SR 520 Bridge users (one with people who did not qualify as low- income or minority, one with low-income English speakers, and one in Spanish with low to moderate incomes).

Using Focus Groups in Assessing the Impact of tolling on Environmental Justice Populations 147 Participants were recruited from the pool of people who responded to the telephone survey, as well as from a purchased telephone list of low-income people who lived in King County. In addition, several social service agencies partnered with Washington State DOT and hung flyers at their sites inviting clients to participate. Only one of the nine people recruited for the Spanish-speaking focus group attended. As a contingency plan, Washington State DOT con- ducted six individual telephone interviews in Spanish with the people who did not show up for the focus group. The moderator guide was focused on learning about the following: • Attitudes toward bridge replacement and traffic congestion; • Attitudes toward tolling the SR 520 Bridge; • The impact of tolling on people’s current and future travel choices; • Whether or not the tolling would create a burden for SR 520 Bridge users, especially those who are low-income and have limited English proficiency; and • Ideas on what (if anything) would make tolling fair. Example 3: Regional Express Lane Network—Metropolitan Transportation Commission Program. Focus groups and intercept surveys were administered to conduct an EJ assessment of the proposed Regional Express Lanes Network Program on behalf of the MTC in the San Francisco Bay Area Region. Informed by demographic mapping of high concentration areas of low-income and minority populations, community-based organizations (CBOs) were identi- fied as potential partners for performing focus groups. Participating CBOs were selected based on their mission, their capabilities working with targeted EJ populations, and their ability to host a focus group discussion and recruit 12 to 15 participants within a required time frame. CBOs were asked to recruit focus group participants drawing on their existing contacts within their communities. In assembling the focus groups, CBOs were asked to screen for participants who were low-income and/or minority individuals who travel in the subject corridors by public transit, carpools, or as solo drivers. CBOs were provided MTC-approved background information that described the Regional Express Lanes Network Program, the purpose of the EJ engagement, and the objective of the focus groups. These materials were translated into Chinese and Spanish to facilitate access to limited English proficiency communities. CBOs were responsible for distributing the materi- als and encouraged to distribute flyers to promote the event. Participating organizations were given a $2,000 stipend for hosting the focus group, recruiting participants, providing child care when appropriate, and arranging for refreshments. Gift cards valued at $25 were given to the participants as an incentive once the focus group was completed. In total, 75 individuals par- ticipated in six focus groups held in community centers and other locales in Alameda, Contra Costa, and Solano counties focused on youth development, vocational assistance, workforce readiness, homeless transition, economic empowerment, and arts and culture, among other missions. Example 4: SR 85 Express Lanes. In 2008, the Santa Clara Valley Transportation Authority conducted a research, public outreach, and education program to gauge public sentiment about the adoption of express lanes on SR 85. As part of that research, a series of four focus groups were conducted with HOV lane users and single-occupancy drivers. Focus group participants were screened to reflect diversity in the ethnicity, income and education level, age, sex, and com- mute patterns of the general population in Santa Clara County. Through the use of a mock news article, one exercise sought to introduce and explore initial attitudes toward the HOT lane and the potential facility layout (see text box, Excerpt from Discussion Guide: SR 85 HOT Lane and Facility Layout).

148 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Findings from the focus groups included: • Although concerns about a “Lexus Lane” initially divided survey respondents evenly, when focus group participants were provided more information and project benefits were explained, participants were more likely to view the project favorably. • The use of toll revenues for other improvements in the corridor, including public transit improvements, was identified as the number one benefit. • Participants reported that they could see how everyone could benefit from express lanes, whether through public transit improvements, better air quality, or improved quality of life from less congestion. Example 5: Elizabeth River Tunnels Tolling. This project provides an example of using focus groups to test public education/marketing approaches about tolling and how different population segments may react differently. Elizabeth River Crossings (ERC) conducted two focus groups, one with those living on the Norfolk side of the Elizabeth River (a generally higher- income area), and one with those living on the Portsmouth side (a generally lower-income area). The purpose of the focus groups was to gather opinions and preferences of campaign ad con- cepts. The ads being tested were divided into two major categories—those for the “education” campaign and those for the “go live” campaign. In general, the campaigns were intended to edu- cate the population about the new tolls and provide instructions about how to obtain an E-ZPass for use on the tunnels between Norfolk and Portsmouth, Virginia. Excerpt from Discussion Guide: SR 85 HOT Lane and Facility Layout Discussion: 30 minutes Introduce SR 85 High-Occupancy Toll (HOT) Lane: • Hand out mock newspaper article about project and get reaction. • Show video of similar project and get reaction. • Show map of facility layout with entrance, exit points, and interchanges. – Answer questions raised by above. – Get Initial reactions. (Probe for general positives and negatives of project: How do they feel about project? How do they think other commuters and community members (who do not commute) will feel about the project? How do they feel about access points—is it local trip-friendly?) Probe: Benefits: provide an option not currently available, convenience, choice, less congestion, less pollution, better use of underutilized high-occupancy vehicle (HOV) lanes, reduced congestion in non-HOV lanes, revenue source for construction and maintenance of HOT lane corridor including public transit, sense of safety, value to time, etc. Negative: elitist, in operation 24/7—no longer unlimited access (during non-car pool hours), cost, what about cheaters, safety, etc., doesn’t encourage carpooling, carpoolers must now travel farther to get into HOT lane— not continue access. Answer questions/clarify project as necessary. Discuss 101 HOT Lane. Where should HOT lanes be built first? Why? Test attitudes toward possibility of two HOT lanes on 101 and 85—Carpoolers could use the second lane to get by slower carpoolers or those trying to merge to General Purpose lane that is congested. Source: SA Opinion Research, 2008.

Using Focus Groups in Assessing the Impact of tolling on Environmental Justice Populations 149 Focus group participants were recruited from two sources: • A list of those who, during the baseline survey, had indicated an interest in participating in a focus group and • Those who were part of an earlier Issues and Answers focus group facility panel. In particular, ERC sought to get a good mix of those who use the tunnels at different frequen- cies, times of the day, and travel alone or with others. Each participant received a $75 stipend to compensate them for their opinions, time, and travel costs. A light meal was also provided at each session. The moderator guide was designed to present to participants each campaign ad concept in print, radio, and TV. This approach was successful in identifying ways to make some of the ad concepts more effective, but it also identified one ad concept that needed to be abandoned because it met with strong negative reactions, especially among those from lower-income house- holds (see text box, Was “Matt’s Spare Change” Worth It?). Was “Matt’s Spare Change” Worth It? Focus group participants did not embrace all of the ads presented. The few par- ticipants that preferred the “Matt’s Spare Change” ad campaign did so because it was witty and humorous. However, others failed to see its humor, or how it would be funny to folks who have to use the tunnels and pay a toll. The ad was seen as out of touch because having “spare change” in our current economy is not so common, and although the humorous approach can be attention-getting, the message about tolling can get lost. What Are Its Limitations? The major limitation of focus groups is that the results are not intended to be extrapolated to the larger population represented in the focus groups because of the relatively small number of participants and because those who agree to participate in a focus group may be different from those who would not, or could not participate. In this limitation, focus groups are quite different in intent than a statistically valid survey. Rather, they are intended to provide insights that would be unavailable with a questionnaire. Focus groups also have the benefit of providing additional insights not available with in-depth interviews, since the synergy that occurs in a group discus- sion has the potential for totally new ideas to emerge. Another limitation of focus groups is that they rely on participants’ memories, which can be faulty. This limitation, not unique to focus groups, is why we see the increased use of “in-the- moment” research, which, through the use of mobile phones and online technology, allows participants to provide feedback at the moment of the participant’s experience (such as for consumer products). What Types of Resources and Costs Are Required? In general, conducting focus groups entails a number of resources and cost items, which, when used with EJ populations, tend to be somewhat more expensive. Below are a list of items and typical costs to conduct one focus group. In general, a focus group could cost between $7,500 and $10,000, although the more groups conducted, the lower the cost per group due to economies of scale (see Table 1).

150 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Who Has Used It Successfully? Several projects assessing the impact of tolling on EJ groups are described in the five examples presented above. More information on the use of focus groups and their use when assessing the impacts of tolling can be found by examining the references presented in the resource section. Resources Bystedt, J., Lynn, S., and Potts, D. 2003. Moderating to the Max: A Full-Tilt Guide to Creative, Insightful Focus Groups and Depth Interviews. Paramount Market Publishing, Ithaca, New York. California Department of Transportation and Santa Clara Valley Transportation Authority. 2012. Community Impact Assessment: State Route 85 Express Lanes Project, Santa Clara County, California. Retrieved from http://dot.ca.gov/dist4/documents/85ExpressLanesProject/ea_4a7900_sr_85_el_community_impact_ assessment.pdf. FHWA. 2008. Income-Based Equity Impacts of Congestion Pricing: A Primer. Washington, D.C. Retrieved from http://ops.fhwa.dot.gov/publications/fhwahop08040/fhwahop08040.pdf. Resource Development Associates and Parsons Brinckerhoff. 2013. Regional Express Lanes Network—MTC Program: Final Focus Group and Intercept Survey Summary Report, Prepared for the Metropolitan Transportation Commission. SA Opinion Research. 2008. Focus Group Report: Attitudes Toward the State Route 85 HOT Lane. Conducted for Santa Clara Valley Transportation Authority. Retrieved from http://www.vta.org/sfc/servlet.shepherd/ document/download/069A0000001Fw72IAC. U.S. DOT. 2008. Equity Concerns of Congestion Pricing Initiatives in the U.S. Draft Report, Washington, D.C. Retrieved from http://www.browardmpo.org/userfiles/files/06-%20Equity%20Brochure-v6.pdf. Weinstein, A., and Sciara, G. C. 2004. Assessing the Equity Implications of HOT Lanes. Santa Clara Valley Transpor- tation Authority, San Jose, California. Washington State DOT. 2007. SR 167 HOT Lanes: Social, Economic and Environmental Justice Report. Retrieved from http://www.wsdot.wa.gov/NR/rdonlyres/9B439F74-7CB2-46BB-A669-7684D31BDCA6/0/ EJ_HotLanesPilotProjectfinal.pdf. Category Cost Focus group facility rental (including hosting and audio/video recording) $700 to $1,200 Recruitment (more for non-English and business participants) $90 to $120 per participant Food for participants $100 to $150 Incentives for participants (higher for non-English and business participants) $75 to $125 Development of the participant recruitment screener $100 Development of the moderator guide $1,500 Translation of the recruitment screener and moderator guide $1,200 to $1,500 per language Moderating the focus group (higher for non-English languages) $1,000 Analysis and reporting $5,000-$7,000 Table 1. Example items and costs for a focus group.

151 What Is It? This tool examines the design and implementation of surveys to assess attitudes toward tolling facilities, travel behavior, and the willingness to pay for the use of managed lanes and tolling facilities. Perceptions of fairness or equity are an important factor in the acceptance of trans­ portation projects that involve the use of pricing. Properly designed surveys can be used to examine how tolling solutions are perceived in terms of fairness and how they may affect low­income and minority travelers compared to other populations. This tool inventories topics and questions used to conduct these surveys and survey collection methods and findings resulting from them. Why Is It Effective in Environmental Justice Analysis? To be responsive to the protected population focus of EJ, surveys should be designed and analyzed to explore similarities and differences in attitudes and travel behavior from the perspective of different travel user groups and segmented by categories of income and race. With proper design and sampling, a survey can directly assess the benefits and burdens of travel as borne by low­income and minority populations compared to other population groups. What Are Some Techniques for Implementing This Tool? This tool can be referenced to identify the types of questions that have been asked and the survey distribution methods that have been typically employed to assess attitudes and measure the demand of toll­ ing facilities and managed lanes in order to support an assessment of the socioeconomic distributional impacts of toll implementation. The tool was developed after inventorying and preparing a content review analysis of toll surveys. The content review analysis examined findings from prior surveys on how attitudes and behavior have been t o o l 6 Designing and Executing Surveys to Assess Attitudes and Travel Behavior for Environmental Justice Analyses and to Monitor Implementation Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Sampling Approaches • Survey Design Questions • Inventory of Survey Methods and Findings Affected Populations • Low-income • Minority Examples Featured • Travel Behavior and Opinion Surveys • FHWA, Pre- and Post-Implementation Surveys, Urban Partnership Agreements/Congestion Reduction Demonstration

152 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox reported by user groups, particularly by low­income or minority users in comparison to other populations. A table summarizing the content review analysis was prepared and can serve as a resource for considering whether and to what extent prior toll surveys have focused on dis­ tributional considerations that can support environmental justice assessments. The summary table is presented in NCHRP Web-Only Document 237: Environmental Justice Analyses When Considering Toll Implementation or Rate Changes—Final Report, which documents the activities undertaken in preparation of the Guidebook and the Toolbox. The content review seeks to expand on earlier publications that have collected evidence from multiple surveys about how tolling impacts equity considerations (e.g., income, modal, geo­ graphic). The 2008 FHWA report Income-Based Equity Impacts of Congestion Pricing: A Primer, provided a high­level summary of key findings from surveys reported up to the time of its pub­ lication (see text box, Income-Based Equity Impacts of Congestion Pricing–Observations Distilled Income-Based Equity Impacts of Congestion Pricing—Observations Distilled from Prior Surveys The Federal Highway Administration (FHWA) report Income-Based Equity Impacts of Congestion Pricing: A Primer makes reference to several examples from early high-occupancy toll (HOT) lane implementation or “partial pricing” studies. Distilling from surveys, polling, and focus groups, the primer makes an argument for congestion pricing and suggests that income-equity concerns may be overstated. “Project experience has shown, particularly for the most common projects funded under the early phases of the program (e.g., HOT lanes), that the percep- tion of unfairness may be exaggerated.” Also, “overall, the perception that congestion pricing is an inequitable way of responding to the problem of traffic congestion does not appear to be borne out by experience” (FHWA, 2008). Examples of acceptance from San Diego, Denver, Minnesota, Houston, and Orange County are given, including the following: • A travel survey of San Diego I-15 found that “users of San Diego’s I-15 HOT lanes were more likely to have higher incomes than were drivers in regular lanes, but lower-income drivers sometimes did use the HOT lanes.” • On plans and implementation of Denver’s I-25/US 36, “public outreach leading to implementation of HOT lanes did not uncover critical concerns regarding equity or other social impacts, nor have such concerns arisen since implementation.” • On I-394 in Minneapolis-St. Paul, “Patterson and Levinson (2008) stated that ‘the [HOT] lanes are Lexus Lanes in the sense that increased income predicts increases in three of the four metrics used to measure direct benefit . . . Individuals with higher incomes receive more direct benefits from the lane than those with lower incomes.’ However, according to the University of Minnesota and NuStats (2005), HOT-lane usage with MnPass was reported across all income levels, including by 79 percent of high-income respondents, 70 percent of middle-income respondents, and 55 percent of low-income respondents.” • For the I-10 and US-290 HOT lanes in Houston, Texas, “focus groups held during project planning did not find concerns about social equity among either corridor users or the public at large. The general reaction was that all would benefit if congestion were reduced. There also have been no equity concerns raised during operations. Burris et al. (2007) found that even in the lowest income group, over two-thirds of respondents were interested in paying to use the HOT lanes.” However, in a closing section, the FHWA Primer concedes that congestion-priced facilities will be used more by those with higher incomes. But since all users will use priced lanes from time-to-time, “income-related equity concerns may not be entirely warranted.” The facilities meet drivers’ needs when they require a reliable trip to reach their destination on time (e.g., to avoid a late charge to pick up a child at a day care center). Moreover, there is little variation by income-level when polling is conducted (e.g., 60 to 80 percent range). All income groups would appear to value the “insurance” of a reliable trip time when they absolutely need it. Source: FHWA, 2008

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 153 from Prior Surveys). More recently, a technical memorandum published for the San Francisco region’s Metropolitan Transportation Commission summarized survey findings related to how people of different income, race, and ethnicity groups responded to tolling projects (HDR Engi­ neering, 2013) and is also included as a reference table in the separate research report. Research undertaken in support of this tool includes an inventory and review of more recent surveys, an examination of survey design methods (e.g., sampling approach, question wording), and considers whether these surveys were designed to carry out an EJ analysis and present equity­ related findings (e.g., income, mode, geography). Most of the surveys in the content review analysis compare travel behavior and opinions for low­income and minority travelers with findings for other groups. To be included in the analysis, a survey had to collect data on respondents’ income or race/ ethnicity and present some analysis of the results by at least one of those demographic categories. Overview of the Surveys Examined. A total of 24 surveys, covering a period from 2003 to 2015, were included in the content review analysis (see text box, Travel Behavior and Opinion Survey Content Review Attributes). Table 1 lists the surveys and includes the facility name, survey sponsor­ ing agency, year of data collection, sampling frame, survey mode, and number of responses ana­ lyzed. Table 2 provides additional information about sociodemographic data collected and whether the survey asked questions about attitudes, behavior, and/or transponder ownership barriers. Travel Behavior and Opinion Survey Content Review Attributes A content review analysis was performed for the following attributes. (1) Survey sponsor: Entity that commissioned the survey (2) Participating organizations: Firms that implemented the survey and/or published project reports docu- menting the survey (3) Project name and location: Name of the facility or corridor under study, and the region of the country (4) Type of tolling project: Managed lane, tolled bridge, etc. (5) Who was surveyed: The type of person selected for surveying (i.e., anyone who made a peak-hour week- day trip within the recent past on a specific facility) (6) Data collection period: The months/years in which respondents completed a survey (7) Survey mode: Phone, online, etc. (8) Survey objective: A description of what the project sponsors wanted to learn from the survey results, as explained in a project report (9) Number of responses analyzed: The number of responses analyzed in the documentation about the survey. In some cases, this is fewer than the number of people who answered the survey questions, such as when incomplete responses were eliminated from the analysis. (10) Decision-making stage: When the survey was completed (11) Were survey materials available in languages OTHER than English: List of any alternative languages in which the survey materials were available (12) Survey question language: Information about the questions asked that are most directly relevant to equity analysis by income or race/ethnicity. Where possible, the exact language used in the survey instrument is reproduced. If the instrument was not provided, then a description of the questions is given. In surveys with complex survey skip patterns, such as the stated preference questions, a representative sample of the questions asked is presented, rather than reproducing the entire relevant set of questions. (13) Survey findings reported by race and/or ethnicity: This section reproduces content in the project reports that describes the survey findings for different race/ethnicity groups. (14) Survey findings reported by income: This section reproduces content in the project reports that describes the survey findings for different income groups.

Facility Type of Tolling Project Performing Organizations Survey Sponsor Data Collection Year(s) Sampling Frame Survey Mode # of Responses Analyzed CALIFORNIA—LOS ANGELES REGION I-110 HOT lanes Redhill Group (Survey Sampling, Inc.) Los Angeles County Metro 2008 General public consisted of residents and persons who lived near the I-110 corridor that used the freeway at least once a week. EJ population included non-Caucasians, 64 years of age or older, or persons with an FHWA Act-defined income bracket of low. Computer- assisted telephone interviewing GP: 160 EJ: 50 I-10, I-210 HOT lanes Redhill Group (Survey Sampling, Inc.) Los Angeles County Metro 2008 General public consisted of Los Angeles County residents who lived near the I-10 and I-210 corridor and used the freeway at least once a week. The EJ population included non-Caucasians, 64 years of age or older, or persons with an FHWA Act- defined income bracket of low. Computer- assisted telephone interviewing GP: 650 EJ: 200 I-10, I-210 HOT lanes Redhill Group (Survey Sampling, Inc.) Los Angeles County Metro 2008 General public consisted of San Gabriel Valley residents and persons who lived near the I-10 and I-210 corridor and used the freeway at least once a week. The EJ population included non-Caucasians, 64 years of age or older, or persons with an FHWA Act-defined income bracket of low. Computer- assisted telephone interviewing GP: 160 EJ: 50 I-10, I-110 HOV lanes/ express lanes Not listed Los Angeles County Metro 2009 Motorist using the general purpose and HOV lanes during peak and off-peak periods. License plate and mailing 1,075 Table 1. Surveys reviewed: organizations involved and methodological details.

I-10, I-110 HOV lanes/ express lanes Redhill Group (Survey Sampling, Inc.) Los Angeles County Metro 2012 (pre- implementation) Response targets were set for each of the eight categories. User category responses were split evenly between I-10 and I-110 freeways. Of the targeted responses from each corridor, half were split between the GP lane and HOV lane. For the HOV and GP lanes, a three-quarters/one- quarter split was sought between those who were on the freeway during peak periods versus those who were using the freeway during the off-peak period. License plate and mailing Total sample: 700 I-10: 350 I-110: 350 For each lane, peak period: 131 Off-peak period: 44 I-10, I-110, I- 210 (Los Angeles County Express Lanes) HOV lanes/ express lanes None listed Los Angeles County Metro 2013 Residents of Los Angeles County who own an ExpressLanes FasTrak account and are enrolled in the Equity Plan (Low-Income Assistance Plan). Online 580 I-10, I-110 HOV lanes/ express lanes Not listed Los Angeles County Metro, Caltrans 2014 (post- implementation) Samples drivers on the I-10 and I-110 in the HOV and GP lanes. Equal percentages of responses from drivers on peak HOV and GP lanes, off-peak HOV and GP lanes on both corridors are retained. License plate and mailing Total sample: 452 I-10: 236 I-110: 216 I-110, I-10 HOT lanes/ express lanes Noble Insights Los Angeles County Metro 2015 Targeted low-income travelers, mostly non-users of Metro ExpressLanes and FasTrak, but no users of these programs were excluded. In-person intercept Sample: 450 Blacks: 100 Asian: 100 White: 50 Spanish speakers: 200 (continued on next page)

Facility Type of Tolling Project Performing Organizations Survey Sponsor Data Collection Year(s) Sampling Frame Survey Mode # of Responses Analyzed COLORADO – DENVER REGION I-25 Managed lanes UrbanTrans, Florida State University Marketing Institute Colorado DOT 2003 Area residents who commute along the I- 25 corridor north of Denver Phone 326 US-36 Managed lanes Wilbur Smith, RSG Colorado DOT 2010 Automobile travelers who recently made a trip in the US 36 corridor (Denver/Boulder) Online 5,340 GEORGIA – ATLANTA REGION I-75 Managed lanes NuStats George State Road & Tollway Authority 2005 Individuals 18 years of age or older, residing within the I-75/575 area who travel on the target segment at least once a week and have a vehicle available Computer- assisted telephone interviewing 1,501 I-20, I-75, I-95, I-285 Managed lanes HNTB, RSG Georgia DOT 2007 People who made a recent weekday trip on a study route Online and in-person intercept 4,173 I-75 South Managed lanes HNTB, NuStats Georgia State Road & Tollway Authority 2007 Individuals 18 years of age or older residing within Clayton, Coweta, Dekalb, Fayette, Fulton, Gwinnett, Henry, Newton, Rockdale and/or Spalding Counties that have a telephone in their home and travel on a target road segment at least once a week. Computer- assisted telephone 1,210 I-85 Express Lanes Managed lanes Volpe, RSG FHWA 2011, 2012 Peak-hour corridor travelers and adult members of their households (drivers, transit riders, organized vanpool members) Online and phone panel study 3,126 Table 1. (Continued).

ILLINOIS – CHICAGO REGION KENTUCKY/ INDIANA – LOUISVILLE REGION MINNESOTA – MINNEAPOLIS REGION OREGON – PORTLAND REGION Chicago Metro Region Managed lanes or congestion pricing on tollways RSG, Wilbur Smith Illinois Tollway 2008 People who within the last month traveled during a weekday at peak period on one of 14 tollways or free expressways in the region including: Jane Adams Memorial, Ronald Reagan, Tri-State, Veterans Memorial Tollways, Dan Ryan, Edens, Eisenhower, Elgin-O’Hare, Kennedy and Stevenson Expressways, IL-53, I-57, I-80, and Bishop Ford Expressway Online 1,976 Louisville- Southern Indiana Ohio River Bridges Project Tolled bridge IQS Research Kentucky Trans- portation Cabinet, Indiana DOT 2014 Racial minorities and/or low-income persons who are members of EJ populations In-person intercept 287 I-394 MnPASS Express Lane Managed lanes NuStats, Humphrey Institute of Public Affairs Minnesota DOT 2004, 2005, 2006 Users and potential users of the express lane (I-394 and I-35W travel sheds) Phone panel study 1,228 (Wave 3) Columbia River Crossing Tolled bridge Stantec, RSG Columbia River Crossing Team 2009 Automobile travelers in the Portland– Vancouver region Online 1,744 (continued on next page)

Facility Type of Tolling Project Performing Organizations Survey Sponsor Data Collection Year(s) Sampling Frame Survey Mode # of Responses Analyzed TEXAS – HOUSTON & DALLAS–FORT WORTH REGIONS WASHINGTON – SEATTLE REGION Katy Freeway, US 290* 2-person carpools pay toll to use HOV lane Texas A&M Transportation Institute Texas DOT 2003 Former and current QuickRide enrollees Mail 525 Katy Freeway, US 290* 2-person carpools pay toll to use Texas A&M Transportation Institute Texas DOT 2003 Travelers in the QuickRide corridors Mail and online 4,005 HOV lane Houston and Dallas regions Managed lanes Texas A&M Transportation Institute Texas DOT 2006 Houston and Dallas residents Online, paper, and interviewer -assisted 4,635 Katy Freeway Managed lanes Texas A&M Transportation Institute FHWA 2008 Travelers who use the Katy Freeway regularly or have at least used it in the past week Online 3,871 I-30 Freeway (Tom Landry Freeway) Use of express lanes with loyalty reward incentives Texas A&M Transportation Institute North Central Texas Council of Govern- ments and FHWA 2014 Travelers on the I-30 Freeway between Arlington and Dallas Online 898 SR-520 Bridge Tolled bridge Volpe, RSG FHWA 2010, 2012 Peak-hour and shoulder-peak corridor travelers and adult members of their households (drivers, transit riders, organized vanpool members) Online and phone 3,698 *Note: Report combines data from two surveys and analyzes all the combined data as one data set. Table 1. (Continued).

Facility Data Collection Year(s) Survey Languages Other Than English Topics Surveyed Demographic Info Collected Results Published By: Transponder Usage Opinions or Attitudes Actual or Predicted Toll Facility Use Income Race/ Ethnicity Income Race/ Ethnicity CALIFORNIA – LOS ANGELES REGION COLORADO – DENVER REGION GEORGIA – ATLANTA REGION I-110 2008 Spanish I-10, I-210 2008 Spanish I-10, I-210 2008 Spanish I-10,I-110 2009 Spanish I-10, I-110 2012 Spanish I-10, I-110, I-210 (Los Angeles County) 2013 None* N/A** I-10, I-110 2014 Spanish I-10, I-110 2015 Spanish or Mandarin I-25 2003 Spanish US-36 2010 None* I-75 2005 None* I-20, I-75, I-95, I-285 2007 None* I-75 South 2007 None I-85 Express Lanes 2011, 2012 Spanish Table 2. Surveys reviewed: survey design and analysis features relevant to analyzing EJ populations. (continued on next page)

ILLINOIS – CHICAGO REGION KENTUCKY/INDIANA – LOUISVILLE REGION MINNESOTA – MINNEAPOLIS REGION OREGON – PORTLAND REGION TEXAS – HOUSTON AND DALLAS–FORT WORTH REGIONS Facility Data Collection Year(s) Survey Languages Other Than English Topics Surveyed Demographic Info Collected Results Published By: Transponder Usage Opinions or Attitudes Actual or Predicted Toll Facility Use Income Race/ Ethnicity Income Race/ Ethnicity Chicago Region 2008 None* Ohio River Bridges 2014 Spanish I-394 MnPASS Express Lane 2004, 2005, 2006 None* Columbia River Crossing 2009 None* N/A*** Katy Freeway, US 290 2003 None* N/A** Katy Freeway, US 290 2003 None* Houston and Dallas regions 2006 Spanish Katy Freeway 2008 Spanish I-30 Freeway (Tom Landry Freeway) 2014 None * Table 2. (Continued).

SR-520 Bridge 2010, 2012 None *       *Information is not specified in the documentation about the survey, but it appears likely that this is the correct information. **All eligible survey respondents had a transponder. *** At the time survey was conducted, regional drivers did not use transponders. Sources: 1 Los Angeles County Metropolitan Transportation Authority. 2009. Express Lanes Congestion Reduction Demonstration Program License Plate Survey Report. 2 Redhill Group, Inc. 2012. ExpressLanes Public Education and Market Research Support: 2012 Pre-Implementation Survey License Plate Study. Prepared for Los Angeles County Metropolitan Transportation Authority. 3 Los Angeles County Metropolitan Transportation Authority. 2013. Equity Plan Survey Analysis. 4 Los Angeles County Metropolitan Transportation Authority. 2014. Post-Deployment License Plate Survey. 5 Noble Insight, Inc. 2015. Metro ExpressLanes Low Income Field Surveys. Prepared for Los Angeles County Metropolitan Transportation Authority. 6 Los Angeles County Metropolitan Transportation Authority. 2008. I-110 Corridor General Public and Environmental Justice Survey. 7 Los Angeles County Metropolitan Transportation Authority. 2008. Los Angeles County General Public and Environmental Justice Survey. 8 Los Angeles County Metropolitan Transportation Authority. 2008. San Gabriel Valley General Public and Environmental Justice Survey. 9 Resource Systems Group (RSG), Inc. 2010. Appendix 1: Denver–Boulder Stated Preference Survey Report of Investment Grade Traffic and Revenue Study U.S.36 Managed Lanes. Prepared for Wilbur Smith Associates and Colorado Department of Transportation. Retrieved from https://www.codot.gov/library/studies/us- 36-managed-lanes-investment-grade-traffic-and-revenue-study/WS%20T-R%20Final%20Appendices1.pdf 10 Ungemah, D., Swisher, M., Tighe, C. 2005. Discussing High-Occupancy Toll Lanes with the Denver, Colorado, Public. Transportation Research Record: Journal of the Transportation Research Board, No. 1932, 129–136. UrbanTrans. 2004. I-25 HOT Lanes Public Outreach: Summary Report: Stated Preference Telephone Survey. Prepared for Colorado Department of Transportation. WASHINGTON – SEATTLE REGION (continued on next page)

11 Peirce, S., Petrella, M., Puckett, S., Minnice, P., Lappin, J. 2014. Urban Partnership Agreement and Congestion Reduction Demonstration Programs: Lessons Learned on Congestion Pricing from the Seattle and Atlanta Household Travel Behavior Surveys. Volpe National Transportation Systems Center. Prepared for U.S. Federal Highway Administration. Retrieved from http://ntl.bts.gov/lib/54000/54000/54065/UPA- CRD_Panel_Survey_Lessons_Learned_Final_Report_Volpe.pdf Peirce, S., Petrella, M., and Green, E. 2014. 2010–2012 Longitudinal Household Travel Diary Study: Seattle & Atlanta. Poster. Retrieved from http://static.tti.tamu.edu/conferences/tss12/posters/14.pdf Petrella, M., Puckett, S., Peirce, S., Minnice, P., Lappin, J., Volpe National Transportation Systems Center. 2014. Effects of an HOV-2 to HOT-3 Conversion on Traveler Behavior: Evidence from a Panel Study of I-85 Corridor in Atlanta (Final Report). Prepared for the U.S. Federal Highway Administration. Retrieved from http://ntl.bts.gov/lib/54000/54000/54062/CRD_Panel_Survey_Atlanta_Final_Report_Volpe.pdf Ray, R., Petrella, M., Peirce, S., Minnice, P., Puckett, S., Lappin, J., Volpe National Transportation Systems Center. 2014. Exploring the Equity Impacts of Two Road Pricing Implementations Using a Traveler Behavior Panel Survey: Full Facility Pricing on SR 520 in Seattle and the I-85 HOT-2 to HOT-3 Conversion in Atlanta (Final Report). Prepared for the U.S. Federal Highway Administration. Retrieved from http://ntl.bts.gov/lib/54000/54000/54064/UPA- CRD_Panel_Survey_Equity_Final_Report_Volpe.pdf Zimmerman, C., Gopalakrishna, D., Pessaro, B., Goodin, G., Saunoi-Sangren, E. 2011. Atlanta Congestion Reduction Demonstration: National Evaluation: Surveys and Interviews Test Plan. U.S. Department of Transportation. Retrieved from http://ntl.bts.gov/lib/51000/51600/51687/11-104.pdf 12 HNTB Corporation. 2008. Study of Potential Managed Lanes on I-75 South Corridor: Final. Prepared for Georgia State Road and Tollway Authority. Retrieved from http://www.georgiatolls.com/assets/docs/I-75_VPPP_Final_Report.pdf NuStats. 2007. I-75 South Stated Preference Survey: Final Report. Prepared for Georgia State Road and Tollway Authority. 13 Hess, S., et al. 2008. Managed-Lanes Stated Preference Survey in Atlanta, Georgia: Measuring Effects of Different Experimental Designs and Survey Administration Methods. Transportation Research Record: Journal of the Transportation Research Board, No. 2049, 144–152. HNTB Corporation. 2010. Atlanta Regional Managed Lane System Plan: Stated Preferences Survey. Prepared for Georgia Department of Transportation. Resource Systems Group (RSG), Inc. 2010. Atlanta Regional Managed Lane System Plan: Technical Memorandum 1B: Greater Atlanta Stated Preference Survey Documentation. Prepared for Georgia Department of Transportation. 14 NuStats. 2005. I-75 Stated Preference Survey: Final Report. Prepared for Georgia State Road and Tollway Authority. 15 Resource Systems Group (RSG), Inc. 2008. Documentation for Chicago Travel Options Study. Prepared for Wilbur Smith Associates and Illinois Tollway Authority. 16 Kentucky Transportation Cabinet and Indiana Department of Transportation. 2014. Appendix E3, Louisville-Southern Indiana Ohio River Bridges Project EJ Community Survey Populations. Table 2. (Continued).

17 NuStats. 2005. I-394 MnPASS Project Evaluation: Attitudinal Panel Survey: Wave 1: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. NuStats. 2006. MnPASS Evaluation: Attitudinal Panel Survey: Wave 2: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. NuStats. 2006. MnPASS Evaluation: Attitudinal Panel Survey: Wave 3: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. 18 Resource Systems Group (RSG), Inc. 2009. Columbia River Crossing Stated Preference Travel Study. Prepared for Stantec. Retrieved from http://www.wsdot.wa.gov/NR/rdonlyres/0DDDCE1C-68F0-4F10-A860-CE92852A0168/0/2012_CRC_ExB.pdf 19 Burris, M., Patil, S., Texas A&M Transportation Institute. 2009. Estimating the Benefits of Managed Lanes. 20 Burris, M., Sadabadi, K.F., Mattingly, S.P., Mahlawat, M., Li, J., Rasmidatta, I., and Saroosh, A. 2007. Reaction to the Managed Lane Concept by Various Groups of Travelers. Transportation Research Record: Journal of the Transportation Research Board, No. 1996, 74–82. 21 Burris, M., Han, N., Geiselbrecht, T., Wood, N., Texas A&M Transportation Institute. 2015. I-30 Express Lanes Survey Report. Prepared for North Central Texas Council of Governments and the Federal Highway Administration. 22 Burris, M., Appiah, J., Texas A&M Transportation Institute. 2003. An Examination of Houston’s QuickRide Participants by Frequency of QuickRide Usage. Prepared for the Texas Department of Transportation. Burris, M., Figueroa, C. 2006. Analysis of Traveler Characteristics by Mode Choice in HOT Corridors. Journal of the Transportation Research Forum, 45 (2), 103–117. 23 Burris, M., Appiah, J., Texas A&M Transportation Institute. 2003. An Examination of Houston’s QuickRide Participants by Frequency of QuickRide Usage. Prepared for the Texas Department of Transportation. 24 Batelle Memorial Institute. 2009. Seattle-Lake Washington Corridor Urban Partnership Agreement National Evaluation Plan. Prepared for U.S. Department of Transportation. Retrieved from http://www.ops.fhwa.dot.gov/congestionpricing/assets/docs/fhwajpo10017/seattleupa.pdf Peirce, S. et al. Volpe National Transportation Systems Center. 2014. Urban Partnership Agreement and Congestion Reduction Demonstration Programs: Lessons Learned on Congestion Pricing from the Seattle and Atlanta Household Travel Behavior Surveys. Prepared for the Federal Highway Administration. Retrieved from http://ntl.bts.gov/lib/54000/54000/54065/UPA-CRD_Panel_Survey_Lessons_Learned_Final_Report_Volpe.pdf Peirce, S. et al. Volpe National Transportation Systems Center. 2014. Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the SR-520 Corridor in Seattle. Prepared for the Federal Highway Administration. Peirce, S. et al. 2014. 2010-2012 Longitudinal Household Travel Diary Study: Seattle & Atlanta. Poster. Retrieved from http://static.tti.tamu.edu/conferences/tss12/posters/14.pdf Ray, R. et al. Volpe National Transportation Systems Center. 2014. Exploring the Equity Impacts of Two Road Pricing Implementations Using a Traveler Behavior Panel Survey: Full Facility Pricing on SR 520 in Seattle and the I-85 HOT-2 to HOT-3 Conversion in Atlanta (Final Report). Prepared for the Federal Highway Administration. Retrieved from http://ntl.bts.gov/lib/54000/54000/54064/UPA-CRD_Panel_Survey_Equity_Final_Report_Volpe.pdf

164 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Tables 3 through 9 present examples of the types of questions used in various surveys of atti­ tudes, opinions, and behavior toward tolling facilities and managed lanes, variable pricing, tran­ sponder usage, and demographic self­identification, among other topics. The set of tables are presented at the end of this write­up. Key findings from these surveys, describing how tolling affects the travel behavior of respondents and/or their opinions about the fairness of toll proj­ ects, can be found in the separate research report. Statement/Question Answer Choices Thoughts on managed HOT lanes a. Good idea b. Bad idea c. Don’t know Why do you feel this way (regarding HOT lanes)? a. Defeats the purpose b. Only people in carpool lanes should be rewarded c. Don’t think it’s fair d. Will not help because it will be the same amount of cars e. Will help reduce the flow of traffic f. Gives people a better option to shorten trip g. The state can raise money h. Oppose tolls/already taxes/no direct answer i. Other I will use a toll route if the tolls are reasonable and I will save time Likert scale I am willing to pay higher tolls if they are used to reduce air pollution and carbon emissions Likert scale Single-occupancy vehicles should be allowed to use the HOV lanes if they pay a toll and speed can remain at least 45 miles per hour (MPH) Likert scale Highway tolls are unfair for travelers with limited incomes Likert scale HOT lanes benefit all travelers because the toll revenue is used to improve local transit which provides a low-cost travel alternative to everyone Likert scale Even if I don’t want to pay to use HOT lanes on a regular basis, it is good to have it as an option when I need to get someplace fast Likert scale Because it is free for carpools, HOT lanes are fair for everyone Likert scale Changing the carpool lanes to HOT lanes is a good idea if it will reduce congestion in the carpool lanes on the I-110 freeway Likert scale Changing carpool lanes to HOT lanes will increase congestion on surface streets around the freeways Likert scale Which of the following best describes your reason for opposing the option to allow those who drive alone to use the HOV lanes in exchange for paying a toll? a. Unfair to lower-income individuals b. Not fair to those who carpool or ride the bus c. Concerned that the lanes will become congested d. Not an appropriate concept for Denver e. Other Express lanes have improved my travel a. Agree b. Disagree Table 3. Examples of attitude or opinion questions for tolled and managed lanes.

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 165 Question Answer Choices The MnPASS program permits single drivers on I-394 to pay a fee to use the MnPASS lanes. Drivers who pay the fee can use the carpool lanes without being in a carpool. The fee varies based on how congested the roadway is. What do you think of allowing single drivers to use the carpool lanes by paying a toll? a. Good idea b. Bad idea c. Don’t know Why do you feel this way (regarding variable pricing)? a. Tolls should be a flat fee/need to anticipate cost b. Should not charge too much when traffic is light c. May create more traffic on those lanes d. Will help bring in more money e. Gives a good option to either carpool or pay toll f. Oppose tolls/already taxes/no direct answer g. Other Why do you feel this way? a. Saves time for busy people b. Users pay, not everyone c. Time is money for some people d. Better use of carpool lanes e. Adds capacity to roadway f. Unfair, specify g. Delays roadway improvement for all h. Levels of service worse in carpool lane i. Increases bureaucracy j. Will not work k. Inefficient l. Only benefits the rich m. Bad for the environment n. Too confusing for people o. Gives too much money to the road agency p. Other, please specify q. Carpool lanes should be free to all r. Don’t know Table 4. Examples of opinion questions for variable pricing. Question Answer Choices Which of the following factors was the most important reason that you use the MnPASS lane? a. To reduce overall travel time b. To reduce the amount of time you spend in heavy traffic c. To increase reliability of your travel time d. To increase personal safety while driving in traffic e. Or something else, please specify Why didn’t you use the MnPASS lane? a. I am not an MnPASS subscriber b. Traffic levels were lighter than usual c. Price was too high d. MnPASS lanes were not available in my direction of travel e. Or some other reason, please specify f. Unsure [Asking about a specific past trip] Have you changed your typical departure time for this trip because of MnPASS? a. Yes b. No c. Refused Table 5. Examples of usage questions for tolling or managed lane facilities. (continued on next page)

166 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Table 5. (Continued). Question Answer Choices In November 2010, you indicated that you usually drove when traveling across or around Lake Washington. Why do you now use public transportation MOST OFTEN? a. To avoid paying the toll on SR 520 b. Price of gasoline c. Environmental reasons d. Bus service has improved e. Travel times are better than driving f. I can be more productive while traveling g. It is less stressful than driving h. It is safer to take the bus than drive i. It is more convenient for the trips I make j. Other, please specify Compared to November 2010, how often do you use each of the following to travel across or around Lake Washington? Drive on SR 520 Drive on I-90 Drive on SR 522 Take public transportation a. Much less often b. Less often c. No change d. More often e. Much more often Question Answer Choices Have you considered getting a transponder? a. Yes b. No c. Unsure Why have you not considered using a transponder? a. Transponder is too expensive to lease b. Don’t want to pay to use MnPASS c. Traffic is not that bad d. Generally don’t drive the I-394 route e. I use carpools f. I use transit g. Unaware of MnPASS h. Wouldn’t use MnPASS lane enough to justify leasing transponder i. Unlikely to use it [specify] j. Other k. Unsure What are the reasons why you do not have a Peach Pass account? a. I do not use GA 400 or the I-85 tolled Express Lanes often enough b. Tolls are too expensive c. I’m against tolling in general d. I’m concerned about my privacy e. I don’t want to have to manage another account f. I don’t want my account to be charged automatically g. I have not yet had a chance to set up an account h. Other To what extent does the $2.00 toll factor into your decision to use QuickRide?* a. Very significant b. Somewhat significant c. No impact d. Somewhat insignificant e. Very significant *Houston’s QuickRide allows two-person carpools to travel in the HOV lane for this toll amount in peak morning and evening periods. Table 6. Examples of transponder ownership questions.

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 167 Question Answer Choices Please identify your race/ethnicity a. White or Caucasian b. Hispanic or Latino c. African American d. Asian American e. Other f. Don’t know/refuse Your race: a. African American or Black b. American Indian or Alaskan Native c. Asian d. White or Caucasian e. Other Are you of Hispanic or Latino origin? a. Yes b. No Please identify your age a. 18–24 b. 25–34 c. 35–44 d. 45–54 e. 55–64 f. 65+ g. Don’t know/refuse Please state your annual household income a. $25,000 or less b. $25,000–$49,999 c. $50,000–$74,999 d. $75,000–$99,999 e. $100,000–$149,999 f. $150,000–$199,999 g. $200,000–$249,999 h. $250,000 or more Please identify your gender a. Male b. Female Please state your employment status a. Employed full-time b. Employed part-time c. Self-employed d. Student e. Student and employed f. Retired g. Homemaker h. Not currently employed Table 7. Examples of demographic self-identification questions.

168 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Statement/ Question Answer Choices In the next section, you will compare the trip you just described with two alternative ways of making the same trip along an improved I-75 and I-30 east of I-75. The options are… a. Drive alone and use the existing lanes with no toll b. Drive alone and use the new managed lanes with a toll c. Carpool and use the new managed lanes, most times with a toll In the next section, you will see 8 questions asking you to compare the trip you just described with 3 alternative ways of making your trip. a. Your current trip (the most frequently traveled highway) b. Use a managed lane on (the most frequently used highway) with a toll paid using I-PASS c. Travel using city streets or local roads only d. Travel by your preferred form of transit What is the primary reason you did not choose the managed lane option in the previous section? a. Time savings is not worth the toll cost b. Toll too high c. Time savings not great enough d. Opposed to paying an additional managed lane fee e. Do not want to set up an I-PASS account (only if don’t have I-PASS) f. Do not want to use electronic toll collection (only if don’t have I-PASS) g. Other If you were to use the carpool lane on this segment of I-75 as a single driver, you would pay [$] and your trip would take [TT] (travel time in minutes). If you were to use the general traffic lanes, your trip would take TT+[#], [#] minutes longer than in the toll lane, but it would be free. You could also choose to carpool with someone to use the carpool lane for free. Now under these conditions, would you choose to: Question 1: a. Use the general lane for free b. Use the carpool lane, pay [$3] and save [5] minutes c. Carpool with someone to use the carpool lane for free Question 2: If toll route is not selected/if toll route is selected a. Use the general lane for free b. Use the carpool lane, pay [$2] and save [5] minutes/use the carpool lane, pay [$5] and save [5] minutes c. Carpool with someone to use the carpool lane for free NCTCOG is exploring how different incentives could change the habits of drivers, carpoolers, and transit riders. How likely is it that you would change your travel if the following benefits were offered? The following six incentives were presented in a random order in the questionnaire: (1) Gifts such as cash, gift card s, or gas cards to local retailers and entertainment venues if you telecommute, travel off-peak, or travel in the express lanes; (2) Reduced transit fare during peak hours; (3) Free items and discounts to local retailers and entertainment venues if you travel off-peak or in the express lanes; (4) For every 10 trips on the express lanes you earn a free trip; (5) Regular transit riders can earn credit toward reduced bus fares or reduced express lane tolls; (6) An express bus service to Downtown from park-and-ride lots on the express lanes Respondent fills in Likert-type scale from: (1) I would change my trips (3) I might change some of my trips (5) I would likely change a lot of my trips Table 8. Examples of stated preference questions.

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 169 Several attributes of the toll surveys that were referenced for the content review analysis and to inform the tool development are briefly summarized below: • Sponsors and regions. The survey sponsors included county, state, and federal departments of transportation. Surveys were sampled across the U.S. in several metropolitan areas such as Los Angeles, Houston, Dallas–Fort Worth, Denver, Atlanta, Chicago, Minneapolis, Portland, and Seattle. Some of the surveys were prepared for the Urban Partnership Agreement (UPA) or Congestion Reduction Demonstration (CRD) Programs. • Language. About half of the surveys were conducted in Spanish in addition to English, while the other half appear to have been conducted only in English. • Mode of distribution. Most surveys were conducted online or via phone with the exception of a couple of mail­back surveys and an intercept survey (see text box, Mode of Distribution). Question Answer Choices Why would you say that you are using the I-85 Express Lanes more often? a. The tolled express lane is faster/less congested b. Road conditions are safer now in the express lanes c. I ride the bus on I-85 more often now d. I can use the express lanes for free (motorcycle, alternative fuel vehicle, and/or 3+ carpool) e. Due to changes in my personal/work situation, I use I-85 more often f. I can drive alone in the express lanes now if I pay a toll g. Other, please specify For your trips in the I-85 corridor northeast Atlanta, how often have you done each of the following in the last month as a result of tolling on the I-85 Express Lanes? Please specify Carpooled/vanpooled on the I-85 instead of driving alone a. Never b. Rarely c. Sometimes d. Often e. Not applicable Rode a public bus instead of driving a. Never b. Rarely c. Sometimes d. Often e. Not applicable Took a different route/road to avoid using the I-85 a. Never b. Rarely c. Sometimes d. Often e. Not applicable Switched to the I-85 Express Lanes instead of using another road a. Never b. Rarely c. Sometimes d. Often e. Not applicable If you usually drive alone to work, what is preventing you from using a commute alternative such as ridesharing or transit? a. Transit service is not adequate b. Difficult to find others to rideshare c. Work late or irregular hours d. Cannot get home in an emergency e. Use my car on the job f. Prefer to drive my own car Table 9. Examples of revealed preference questions.

Mode of Distribution Computer Assisted Telephone Interviewing (CATI). This method typically involves the use of a Random Digit Dialing Process (RDD). The RDD sample has six digits out of a standard 10-digit telephone number based on the region’s area code and randomly generates a set of three numbers. The number set excludes cell phone numbers. With the CATI system, the questions are populated into the system and the interviewer asks the ques- tions to the respondents similar to any telephone survey. The answers are recorded by the computer and the interviewer does not have to manually populate the responses. CATI surveys are typically used to sample popu- lations that live near the corridors of interest or in neighboring counties and populations that have used the corridors within a specified time frame, typically the last week. CATI surveys can target specific demographics such as low-income and minority travelers by using the RDD to provide telephone numbers for households in a neighborhood with this demographic near the corridors of interest. License Plate Mail In and Back. A camera is trained on the freeway in both the general purpose and managed or tolled lanes to record the license plates of drivers. The local Department of Motor Vehicles is contacted to retrieve postal addresses that match the license plates that were recorded. The household of the license plate holder is mailed the survey to complete and requested to mail it back to the sponsoring agency’s representa- tive. Surveys may also use postal addresses to send households a postcard with a link to an online version sur- vey. The license plate mail in and back survey approach is used when the agency is trying to sample populations that currently use the corridor; the approach can target drivers who used the corridor’s general purpose or managed lane specifically during peak or non-peak hours so that the sample can be representative of various user segments. Surveys also used this method to sample drivers that owned transponders. Online. A unique website URL for the survey is created and distributed to targeted populations. Using email addresses of transponder account holders, an email may be sent explaining the purpose of the study, the link to the survey, and requesting participation. Businesses and organizations in the region may be contacted to pro- vide the survey to their employees. An online market research firm may be contracted to develop a panel of par- ticipants to complete the survey. Drivers that paid cash at toll plazas may be given an invitation to complete the survey via a postcard by the toll monitor. The online survey method is used to sample a variety of different popu- lations such as transponder owners, drivers that live near the corridors of interest or in neighboring counties, or drivers that recently made a trip on the corridor of interest; surveys are typically restricted for weekday trips. In-person Intercept. Interviewers are stationed at frequently trafficked shopping areas, public offices, universi- ties, and other institutions to stop travelers or pedestrians and ask them to complete a survey. The in-person intercept survey may be administered with the assistance of a laptop, hand-held tablet, or clipboard and may or may not be completed in a discussion with the interviewer. The sampling frame is not random but based on convenience, although a representative mix of respondents along several dimensions may be sought (e.g., peak periods, off-peak periods, managed lanes, general purpose lanes, transit, and persons of various demograph- ics). In-person intercept interviews can also be used to target specific affected populations such as low-income and minority travelers or local residents of an affected area. Interviewers can be stationed at places that low- income and minority travelers would frequent such as shopping centers within a low-income or minority neighborhood (see case example, “Targeting Local Grocery Stores to Administer Community Surveys, Louisville– Southern Indiana Ohio River Bridges Project”). Panel Study. The panel study method is used for before and after implementation surveys because it enables the same households to be surveyed in each round. The panel study method targets a variety of sample popu- lations. Drivers on the corridors of interest are selected via a camera that records license plates, and the agency uses the postal address connected to the license plate to contact households. Households near the corridors of interest or in the neighboring counties are contacted through random digit dialing based on the area codes in the region and neighborhood population density. Drivers owning a transponder are contacted by the agency via telephone numbers linked to their accounts. Interviewers intercept transit riders at local stations and col- lect their information. Transit riders are then sent a link to the survey. Members of organized carpools are con- tacted via email. For surveys that are implemented during different waves, a reminder postcard is sent to the households before the next wave for surveying ends.

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 171 The intercept survey followed a convenience method focused on reaching low­income and minority residents. • Topics covered. All surveys explored the respondent’s attitudes and actual or predicted use of the tolled facility. • Socioeconomic data. While survey tools did ask questions that could enable an analysis broken out by income, many surveys reported by user groups (e.g., peak/off­peak, HOV/non­HOV, travel mode), but only a few findings cross­tabulated by income. The surveys reported user findings by income group more often than by racial group. • Timing of surveys. Most of the surveys were conducted during the pre­implementation stage for toll pricing and questions centered on respondent’s attitudes, stated preference to use the toll facility, and/or willingness to pay under certain conditions, as well as likelihood to have a toll account and transponder. Post­implementation surveys also examined attitudes toward tolling, actual use, transponder ownership, and awareness of programs. Survey Questions. Many of the surveys asked respondents about attitudes toward tolling or managed lanes, willingness to purchase a transponder, willingness to pay tolls, value of time (VOT), and predicted use of the facility. Depending on the stage of planning and implementa­ tion, surveys used stated preference and revealed preference questions to assess respondents’ predicted and actual use of the facility. • Stated preference survey. Stated preference questions were used in the planning and project development stages to assess respondents’ likelihood of using the tolled facility or managed lane. Prior to asking stated preference questions, the surveys are designed to capture the respondents’ typical travel behavior. Questions prompt the respondent with a typical commute and ask which of the listed travel options he/she would be most likely to use. Respondents are presented with a series of experimental tests and asked in each test to choose one among several travel options each reflecting a randomized level of time savings and a randomized toll price. With each new question respondents are presented with a different choice set reflecting perhaps a higher toll price, if they selected the toll route in the previous question, or a lower toll price, if they did not select the toll route in the previous question. This successive question method is used until the respondent ceases to select the tolled route and the toll price is identified as the respondent’s willingness to pay. The agency can collect respondents’ willingness to pay and the sample’s value of time savings. Stated preference questions have also recently been applied to gauge preferences for various incentive programs to increase interest in tolled facilities or managed lanes (see text box, Stated Preference Question Example). Stated Preference Question Example If you were to use the general traffic lanes on this segment of I-75, your trip would take [#] minutes of travel time (TT) and be free. If you used the new carpool lane as a single driver, you would pay [$] and your trip would take TT, saving [#] minutes. You could also choose to carpool with someone and use the carpool lane for free. Now under these conditions, would you: a) Use the carpool lane, pay [$] and save [#] minutes b) Carpool with someone to use the carpool lane for free c) Use the general lane for free d) Don’t know

172 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox • Revealed preference survey. These surveys were used in the post­implementation stage to mea­ sure respondent’s actual use of the tolled facility or managed lane. In some cases, the survey reported findings for both the stated and revealed preference surveys in order to explore deviations between the ways that respondents believed they would use the facility and how they actually used the facility. A good example of this method is used in Effects of an HOV-2 to HOT-3 Conversion on Traveler Behavior: Evidence from a Panel Study of the I-85 Corridor in Atlanta (see text box, Revealed Preference Question Example). The example highlights a ques­ tion intended particularly for former HOV­2 users that were required to pay for HOT lane usage after implementation. Revealed Preference Question Example Why would you say you are using the I-85 Express Lanes LESS OFTEN? Please select all that apply: 1) The regular lanes on I-85 are less congested now. 2) I’d rather not pay a toll in the I-85 Express Lanes. 3) I carpool with one other person and we no longer can use the express lane for free. 4) I use a different route now to avoid I-85. 5) Entering/exiting the express lanes is difficult/inconvenient. 6) The express lanes are less safe. 7) Due to changes in my personal/work situation, I use I-85 less often. 8) Other, please specify. • Opinion or Attitudes. Surveys were used to assess respondents’ attitudes toward the tolled facility or managed lane project. These questions seek to enable the conducting agency to understand what aspects of the project design are liked or disliked to guide the planning and operations of the facility. Questions that assess attitudes usually have a set of statements around a theme and ask respondents to use a Likert scale to rate their agreement with the statement. Respondents’ attitudes are assessed by themes such as support for funding addi­ tional transit with revenue from tolls, incentive options, equity of facility, environmental consciousness, price of tolls, and more. • Toll Account Policies and Transponder Usage. Surveys have been developed to explore respon­ dents’ perspectives on various features of toll policies, such as up­front deposit, monthly maintenance fees, minimum balance requirements, automatic replenishment charges, and requirements to use a credit card or debit card account with the toll or transponder account. These and other policies may function as barriers to having a toll account or transponder for some low­income and unbanked populations. These types of questions have been posed at early planning and post­implementation phases to assess factors affecting the awareness of tolling policies and the use of transponders based on income, race, or geography. Under­ standing respondents’ choices can inform a sponsoring agency’s subsequent policies and practices, leading to adjustments in marketing strategies; deposit, maintenance, and credit policies; or incentives. Corrective strategies may differ if respondents are unlikely to use a transponder because they have insufficient credit or are unable to pay, than if respondents are unlikely to own a transponder because they do not travel often or need to get places fast (see text box, Toll Account and Transponder Ownership Example).

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 173 • Demographic Self-Identification. Many of the surveys have explored various demographic features influencing travel demand such as gender, age, employment status, household size, vehicle ownership, and household income. Only some of the surveys also ask respondents to provide other relevant racial/ethnic group information to support a comprehensive consid­ eration of EJ. What Are the Limitations? Several of the reviewed surveys were prepared to estimate values of the toll sensitivity or VOT of travelers to identify current travel preferences and behaviors and explore the users’ potential will­ ingness to pay to use tolling facilities such as managed lanes under various pricing and travel time options. Many of these surveys were either not specifically designed or analyzed to support detailed equity and EJ analyses. A few patterns were observed in the preparation of this content review. • Many travel demand related surveys were not designed to ask about race/ethnicity. • Many survey instruments were available only in English. • Many sampling plans for surveys were done with user group populations that included very small numbers of low­income or minority respondents, resulting in too few people in the protected populations of concern to compare responses, with confidence, to responses of the rest of the population. • In some cases, surveys combined low­income and minority populations into one group and middle­to­high­income and non­minorities into one group, making it difficult to deter­ mine whether the differences in responses are related to differences in income or race/ ethnic demographics. • Few of the reviewed surveys specifically targeted low­income and minority groups as catego­ ries for the sample population. Sampling plan design could be more proactive in addressing this issue with pre­planning and appropriate budgeting. With greater attention to overweight­ ing or requiring a minimum sample size, low­income and minority travelers’ perspectives and travel behaviors could be rigorously considered for time segments (e.g., peak­hour, off­peak). Although this level of effort is not common practice for travel user survey implementation, oversight agencies could require evidence of this effort in their regional planning certification, during environmental review approval phases for specific project regions or corridors, or in the conditions of grant commitments. • Making a commitment to preparing sampling plans that seek greater representation of low­ income and minority populations could be advanced in several ways. For example, the license Toll Account and Transponder Ownership Example What are the reasons why you do not have a Peach Pass account? 1) I don’t use GA 400 or the I-85 Express Lanes (with tolls) often enough. 2) Tolls are too expensive. 3) I’m against tolling in general. 4) I’m concerned about privacy. 5) I don’t want to have to manage another account. 6) I don’t want my account to be charged automatically. 7) I prefer not to have to pay a deposit in advance. 8) I have not yet had a chance to set up an account. 9) Other, please specify.

174 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox plate mail in and back method should direct surveys to addresses within low­income and minority census blocks. Greater attention to surnames in random digit dialing phone lists would increase Hispanic household participation. The online survey method could send links to companies within industries whose workers typically are working class or low­income and to businesses and companies with many minority employees. Both the license plate mail in and back and online survey method should allow adequate time for participants to complete the survey to ensure sample size. The in­person intercept survey method should place inter­ viewers in strategic locations in low­income and minority communities, for example, at shop­ ping locations that accept Electronic Benefits Transfer (EBT) cards. Even surveys that have the raw data needed for EJ analyses (i.e., asked the right questions in the survey instrument) receive very little specific analysis relating to income or race/ethnicity. The review of surveys found the following: • Survey reports typically do not systematically present findings for every question by income or race/ethnicity. Even the reports with the most comprehensive coverage only discuss some of the questions asked by income and/or race. In some cases, response patterns for the EJ populations were presented in separate reports or chapters and did not facilitate the types of comparison of benefits and burdens that would be anticipated in an EJ analysis. • Race/ethnicity is less better covered than income. Far more reports provide some findings by income group than by race/ethnicity. And even the ones that do report some findings by race/ ethnicity usually report far less than they do by income. • Many reports present no analysis at all by race/ethnicity or income except for looking at income to determine VOT/willingness­to­pay. Improved documentation and reporting expectations for the topic of EJ could advance the state­of­the­practice and support greater knowledge sharing and transparency. Many reports refer to appendices with details such as the survey questionnaire, but do not include these materials when the reports are posted. Some reports reference other documents that contain more detailed survey reports or analyses, but these reference documents are unavailable. Fund­ ing and sponsoring agencies should expect complete reports to include supporting techni­ cal appendices related to survey methods and questionnaire design, among other materials. In the context of EJ, the affected public and interested stakeholders should be afforded access to the completed datasets for research and monitoring. Sharing these materials through a central repository of surveys or high­quality surveys at a clearinghouse website (e.g., AASHTO or FHWA Congestion Pricing website) could also improve the state­of­the­practice. What Types of Resources and Costs are Required? Resources and costs vary greatly depending on the sample size, mode of survey distribution, and other protocols for survey administration. Sampling plans must set sample sizes large enough to support travel demand analysis for various user groups (e.g., peak/off­peak, HOV/non­HOV, mode) and support low­income and minority segmentation to assess EJ considerations with con­ fidence. Regardless of mode of distribution, it is also important to have surveys translated and available in multiple languages to reach limited English proficiency populations in the region or corridor to avoid non­response bias and exclusion of these populations from the sample. Before and after surveys involving the recruitment of a panel through multiple waves of a large sample survey with a complex questionnaire are expensive and time­consuming. Project sponsors that are committed to the recruitment of a panel study approach will need to budget for the work and begin the planning process at least a year or more before the tolling project opens. Panel study surveys can be particularly expensive because of the need to attract and retain panel

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 175 members over time, requiring the research team to track changes in the work status, residential location, and family status of the participants, among other issues; maintain regular communi­ cations; and provide participation incentives. The license plate survey method is appropriate for capturing travelers on specific corridors. Time and costs must be anticipated for videotaping, coordinating with the local Department of Motor Vehicles (DMV) for addresses, and postage to mail and return surveys and/or send postcards with links to an online survey. Survey materials can only be sent to the addresses of the registered owners of the vehicles that are photographed, which can have the effect of excluding travelers in the corridor who are passengers in a vehicle owned by someone else or the driver of the vehicle if not the owner. Because the owner’s name may not be disclosed by DMV, the targeted respondent can only be impersonally referenced (e.g., “Dear Motorist”). The combined effect of these factors may lead to returned survey response rates of only 8 to 15 percent. Vanpool vehicles also cannot be matched to the lead driver’s address or to the addresses of pas­ sengers. Computer­assisted telephone interview surveys tend to be costly because they require mobilizing a team of interviewers to conduct the phone interviews. The in­person intercept survey may be the least expensive and useful for soliciting general opinions about tolling in the general population or target EJ populations. Who Has Used It Successfully? Prior policy research studies have found only a limited amount of research focusing on the before and after evaluations of the effect of pricing projects on travel behavior. One strategy for addressing this gap in knowledge would be the use of an equity­type audit tool after the implementation of road pricing projects. Periodic monitoring of project effects can examine the success of mitigation measures and whether differences in equity are being effectively addressed. Where persistent inequities are revealed in the monitoring phase, modifications to road pricing and mitigation can be subsequently made (Ecola and Light, 2009). Consistent with this observation, the FHWA UPA and CRD program grants have employed “Before­After” survey protocols in recent years in Minnesota, Los Angeles, Seattle, and Atlanta. The longitudinal panel design was particularly effective for identifying opinions, travel behavior, and willingness to pay information from users and potential users of the tolling facilities such as managed lanes. Initially stated attitudes and behavior were compared with subsequent opinions and travel decisions on such themes as equity, acceptance of tolling solutions, and perceptions of effectiveness in congestion management, as well as changes in travel behavior, mode choice, route choice, and willingness to pay for the priced lane before and after project implementation. Additional information regarding the survey performed in Atlanta in association with FHWA’s evaluation plan for the UPA and CRD program grants conducted by the Volpe National Transpor­ tation Systems Center is provided in a case example, “Conducting Pre­and Post­Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I­85 Corridor.” The discussion high­ lights some of the challenges of instituting longitudinal surveys reported in the Volpe study, as well as the benefits of designing surveys and obtaining a sample size sufficient to draw conclusions with confidence about specific protected populations of concern for EJ along multiple themes (e.g., transponder ownership, use of tolled express lanes, carpooling, and attitudes). As toll pricing solutions spread to manage congestion, interest in the role of toll credits and other incentives for changing the behavior of drivers, carpoolers, and transit riders may increase. Recent survey research in the Dallas–Fort Worth region (see Tables 1 and 2), in association with the NCTCOG, explored the role of loyalty programs and other incentives, including credit rewards to transit users and regular users of express lanes as a means to optimize throughput

176 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox in the managed lanes. Credit and other loyalty­type programs may hold promise as a demand management tool. While they may also function as a form of mitigation for low­income travelers, their fairness and relevance to EJ analysis must be carefully assessed. For this evaluation to occur, sampling plans must be sufficiently robust, and analysis and reporting must be designed to comprehensively assess how the benefits and burdens of these initiatives may be perceived and borne by low­income and minority populations in comparison to the broader general population (i.e., the non­EJ populations). Resources Battelle Memorial Institute. 2009. Seattle­Lake Washington Corridor Urban Partnership Agreement National Evaluation Plan. Prepared for the U.S. Department of Transportation. Retrieved from http://www.ops.fhwa. dot.gov/congestionpricing/assets/docs/fhwajpo10017/seattleupa.pdf. Burris, M., Figueroa, C. 2006. “Analysis of Traveler Characteristics by Mode Choice in HOT Corridors.” Journal of the Transportation Research Forum, Vol. 45, No. 2, pp. 103–117. Burris, M., Sadabadi, K. F., Mattingly, S. P., Mahlawat, M., Li, J., Rasmidatta, L., and Saroosh, A. 2007. “Reaction to the Managed Lane Concept by Various Groups of Travelers.” Transportation Research Record: Journal of the Transportation Research Board, No. 1996, Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 74–82. Ecola, L., and Light, T. 2009. Equity and Congestion Pricing: A Review of the Evidence. Rand Corporation, Santa Monica, California. FHWA. 2008. Income-Based Equity Impacts of Congestion Pricing—A Primer. Retrieved from http://www.ops. fhwa.dot.gov/publications/fhwahop08040/fhwahop08040.pdf. HDR Engineering. 2013. MTC Regional Express Lanes Interstate 680 Corridor: Environmental Justice Technical Memorandum. Prepared for the Metropolitan Transportation Commission. Hess, S., Smith, C., Falzarano, S., and Stubits, J. 2008. “Managed­Lanes Stated Preference Survey in Atlanta, Georgia: Measuring Effects of Different Experimental Designs and Survey Administration Methods.” Trans- portation Research Record: Journal of the Transportation Research Board, No. 2049, Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 144–152. HNTB Corporation. 2008. Study of Potential Managed Lanes on I­75 South Corridor: Final. Prepared for the Georgia State Road and Tollway Authority. KYTC and Indiana DOT. 2014. Appendix E3, Louisville-Southern Indiana Ohio River Bridges Project EJ Community Survey Populations. Noble Insight. 2015. Metro ExpressLanes Low Income Field Surveys. Prepared for the Los Angeles County Metropolitan Transportation Authority. NuStats. 2005. I­75 Stated Preference Survey: Final Report. Prepared for Georgia State Road and Tollway Authority. NuStats. 2005. I­394 MnPASS Project Evaluation: Attitudinal Panel Survey: Wave 1: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. NuStats. 2006. MnPASS Evaluation: Attitudinal Panel Survey: Wave 2: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. NuStats. 2006. MnPASS Evaluation: Attitudinal Panel Survey: Wave 3: Final Report. Prepared for the Humphrey Institute of Public Affairs, University of Minnesota. NuStats. 2007. I­75 South Stated Preference Survey: Final Report. Prepared for the Georgia State Road and Tollway Authority. Peirce, S., Puckett, S., Petrella, M., Minnice, P., and Lappin, J. 2014. Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the SF-520 Corridor in Seattle. Volpe National Transportation Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54063/UPA_Panel_Survey_Seattle_ Final_Report_Volpe.pdf. Peirce, S., Petrella, M., Puckett, S., Minnice, P., and Lappin, J. 2014. Urban Partnership Agreement and Conges- tion Reduction Demonstration Programs: Lessons Learned on Congestion Pricing from the Seattle and Atlanta Household Travel Behavior Surveys. Volpe National Transportation Systems Center. Retrieved from http:// ntl.bts.gov/lib/54000/54000/54065/UPA­CRD_Panel_Survey_Lessons_Learned_Final_Report_Volpe.pdf. Peirce, S., Petrella, M., and Green, E. 2014. 2010–2012 Longitudinal Household Travel Diary Study: Seattle & Atlanta. Poster. Retrieved from http://static.tti.tamu.edu/conferences/tss12/posters/14.pdf. Petrella, M., Puckett, S., Peirce, S., Minnice, P., and Lappin, J. 2014. Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the I-85 Corridor in Atlanta. Volpe National Transportation Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54062/CRD_Panel_Survey_Atlanta_ Final_Report_Volpe.pdf.

Designing and Executing Surveys to Assess Attitudes and travel Behavior for Environmental Justice Analyses and to Monitor Implementation 177 Ray, R., Petrella, M., Peirce, S., Minnice, P., Puckett, S., and Lappin, J. 2014. Exploring the Equity Impacts of Two Road Pricing Implementations Using a Traveler Behavior Panel Survey: Full Facility Pricing on SR 520 in Seattle and the I-85 HOT-2 to HOT-3 Conversion in Atlanta. Volpe National Transportation Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54064/UPA­CRD_Panel_Survey_Equity_Final_Report_Volpe.pdf. Redhill Group. 2012. ExpressLanes Public Education and Market Research Support: 2012 Pre­Implementation Survey License Plate Study. Prepared for Los Angeles County Metropolitan Transportation Authority. Resource Systems Group. 2008. Documentation for Chicago Travel Options Study. Prepared for Wilbur Smith Associates and the Illinois Tollway Authority. Resource Systems Group. 2009. Columbia River Crossing Stated Preference Travel Study. Prepared for Stantec. Retrieved from http://www.wsdot.wa.gov/NR/rdonlyres/0DDDCE1C­68F0­4F10­A860­CE92852A0168/0/ 2012_CRC_ExB.pdf. Resource Systems Group. 2010. Appendix 1: Denver­Boulder Stated Preference Survey Report of Investment Grade Traffic and Revenue Study U.S. 36 Managed Lanes. Prepared for Wilbur Smith Associates and the Colorado Department of Transportation. Retrieved from https://www.codot.gov/library/studies/us­36­managed­ lanes­investment­grade­traffic­and­revenue­study/WS%20T­R%20Final%20Appendices1.pdf. Resource Systems Group. 2010. Atlanta Regional Managed Lane System Plan: Technical Memorandum 1B: Greater Atlanta Stated Preference Survey Documentation. Prepared for the Georgia DOT. Ungemah, D., Swisher, M., Tighe, C. 2005. Discussing High­Occupancy Toll Lanes with the Denver, Colorado, Public. Transportation Research Record: Journal of the Transportation Research Board, No. 1932, Transporta­ tion Research Board of the National Academy of Sciences, Washington, D.C. pp. 129–136. UrbanTrans. 2004. I­25 HOT Lanes Public Outreach: Summary Report: Stated Preference Telephone Survey. Prepared for the Colorado DOT. Zimmerman, C., Gopalakrishna, D., Pessaro, B., Goodin, G., Saunoi­Sangren, E. 2011. Atlanta Congestion Reduction Demonstration: National Evaluation: Surveys and Interviews Test Plan. Retrieved from http://ntl.bts.gov/lib/ 51000/51600/51687/11­104.pdf.

178 What Is It? Travel demand modeling, generally, is the modeling of regional or local traffic for the purpose of assessing current and future traf- fic performance. Depending on the sophistication of the modeling effort, TDMs are used to forecast and simulate vehicle and per- son trips on the highway and multimodal transportation systems. TDMs can and have been used to estimate the quality of the level of flow (travel time, speed, levels of service for regional transpor- tation systems, and individual links in the network). They have been used to identify link locations where traffic volumes exceed roadway capacity as a step toward finding needed improvements. TDMs are typically maintained and used by MPOs. Every urbanized area with a population of greater than 50,000 as desig- nated by the U.S. Census Bureau has a TDM. State DOTs as well as toll authorities also have developed and used TDMs. These models vary but typically rely on large and diverse datasets, includ- ing socioeconomic data, land use data, current road network traf- fic, surveys, socioeconomic forecasts, projected land use, and future road network data. In a tolling context, TDMs are used to assess the impact of toll- ing projects to capture the systemwide effects of travelers switch- ing to other routes to avoid tolls and to forecast usage of the toll facility under different growth and revenue scenarios. Revenue forecasts depend on travel demand forecasts and their underlying assumptions. Generally speaking, most models applied for highway pricing projects represent a modification of the existing regional travel network model. Many of the limitations and deficiencies of the regional model may be echoed in the pricing study. For example, the current state-of-the practice suggests that in most cases, only route itinerary and binary route type choice models have been employed to evaluate the impacts of pricing alternatives. This approach falls short of examining the effect that the toll itself may play in altering mode choice, time-of-day choice, trip distribu- tion, or trip generation. Although the traditional 4-step model- ing approach can address these questions, they have not generally Using Travel Demand Models for Environmental Justice Assessments t o o l 7 Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Policy and Planning • Project Design and the National Environ- mental Policy Act • Implementation Tools and Techniques • Trip-based modeling • Activity-based modeling • Dynamic Traffic Assignment • Microsimulation Affected Populations • Low-Income • Minority Examples Featured • North Central Texas Council of Govern- ments, Dallas–Fort Worth, Regional Tolling Analysis • Kentucky Transportation Cabinet and Indiana DOT, Ohio River Bridges • San Francisco County Transportation Authority, Mobility, Access, and Pricing Study

Using travel Demand Models for Environmental Justice Assessments 179 been implemented to address such concerns. However, innovations in travel demand modeling now underway, such as activity-based models (ABMs) and dynamic traffic assignment (DTA), appear to be slowly advancing the state-of-the-practice. In time, many agencies and practitio- ners, particularly in the larger urban metropolitan regions, will likely see refinements in the tools available for assessing the effects of toll policies and investments on various types of travelers, including low-income travelers. For toll projects, TDMs are employed during four decision-making stages that can lead to modeling refinements representing travel behavioral considerations and addressing issues relevant to the public, regulatory agencies, funders, and other decisionmakers. 1. Exploratory. During the exploratory stage, the very first “go/no-go” decision point, the regional model will be applied with toll pricing converted into travel time equivalents for highway assignments and skimming. 2. Preliminary Feasibility. During the preliminary feasibility study, the model may be improved to include more segmentation (including mode choice, auto occupancy) and a differentiation of model coefficients relative to VOT. 3. Project Development and Environmental Studies. During this stage, an Environmental Impact Statement (EIS) will be prepared, that typically incorporates additional sub-models and introduces a toll diversion model to reflect the travel behavior effects from the toll. 4. Investment Grade Study. For an investment grade study, a forecast is prepared in anticipation of a new toll project as the basis for credit ratings, finance approval, and the sale of capital market debt. Investment grade studies typically involve some form of microsimulation of individual drivers to represent travel demand choices. Using stated or revealed preference sur- veys, the likely travel market’s VOT will be established in these models and will include users’ willingness to pay over multiple model runs in order to optimize toll values. Ratings agencies typically place investment grade studies under a high level of scrutiny, with a well-calibrated model that integrates network simulation. Why Is It Effective in Environmental Justice Analysis? Travel demand modeling is an important tool to better understand how specific groups of travelers—organized by income, minority status, or geographic area—respond to changes in trans- portation networks. With regard to tolls, TDMs generally model traveler responses to pricing, which can shed light on how various populations may be affected by the pricing change. An understanding of EJ considerations should inform methods taken by travel demand mod- elers at many of the decision-making stages associated with pricing projects. TDMs may be improved by devoting greater attention to the attitudes and travel behavior of low-income and minority groups, including estimates of VOT or willingness to pay. As projects move through the project design and environmental review stage of decision-making, EJ considerations are placed under particular scrutiny. Several benefits and burden measures can be usefully quantified by using a TDM. With proper TDM design and use of supportive surveys, it should be possible to examine mobility and access impacts to low-income users in the following categories: • Vehicle trips by time of day; • Vehicle miles and hours traveled; • Mode share for work, non-work, and all other trips; • Accessibility to transit service and frequency; • Accessibility to key destinations including jobs, health care, and education; • Average trip length (considering both time and distance);

180 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox • Average roadway speeds and delay; and • Reliability of travel. What Are Some Techniques for Implementing This Tool? Several types of TDMs are in use to assess the effects of tolling. The basic types of models are described below with reference to how they may be applied in a tolling study, how they may be used to address EJ, and their limitations. Some key definitions and considerations in the applica- tion of travel demand modeling for toll purposes are also described. Trip-Based, 4-step Model. A trip-based, or 4-step model is an aggregate TDM that uses trips made between pairs of geographic locations. TAZs are typically the base unit of analysis. Each trip is modeled as an individual unit and is associated with a trip purpose (work, school, socio- recreation, shopping, non-work). The trip-based model is the most commonly used model in the United States and is frequently applied to assess traffic performance and forecast future traffic demand. The trip-based model forecasts travel demand in four basic steps: 1. Trip Generation. In the first step, the magnitude of daily travel in the model system is evalu- ated. This includes evaluating annual average daily travel in a region and combining this infor- mation with land use, population, and economic forecasts to estimate the number of trips that will be made to and from each zone. Trip ends are modeled as productions or attractions and converted into origins and destinations during various times of the day. Trips can be modeled at the zonal, household, or personal level. Household cross-classification matrix models are relatively common for trip productions and zonal-level models are common for trip attractions. 2. Trip Distribution. In the second step, a trip matrix or trip table is developed by means of a trip distribution model (frequently a gravity model). In this stage, each zone’s origins are proportionately distributed to all other zones typically based on travel time or impedance and the level of attractiveness for each zone. 3. Mode Choice. By factoring (logit and nested logit models) the trip tables produced in the second step, transport mode-specific trip tables are produced to reflect the choice probability of individual trip-makers. This is typically where the use of a toll will be introduced, as part of a binary (toll or no toll) route type choice model. In more recent applications, the determina- tion whether to use a toll road is made in step 4. 4. Traffic Assignment or Route Choice—During this step, trips are assigned to specific routes and links in the network. User equilibrium (equal impedance) is applied to the assignment by iterating the assignment process. Modal origin-destination trip tables are loaded on the modal networks. Generally speaking, the first three steps of the trip-based model constitute the transportation demand-side, while the fourth step, traffic assignment, represents the supply side. In the fourth step, demand is loaded onto the transportation network to determine route choice. In many applications of trip-based models, there are no feedback loops to inform travelers on how they might change route choice based on congestion, mode availability, or time-of-day variations. However, the 4-step process is capable of reiterating through the entire process beginning with land use forecasting, trip generation, and trip distribution. Many models, at a minimum, will iterate back to distribution to account for congestion, mode availability, or time-of-day impacts. Each step makes important assumptions that can be critically examined from an EJ per- spective. In the trip generation step, households are distinguished according to several char- acteristics, including household size, level of income, car ownership level and, less often, race/ ethnicity. Wealthier households are more likely to own cars and make more trips and will likely

Using travel Demand Models for Environmental Justice Assessments 181 generate more of the trips observed during the trip generation stage. For each household, the average number of trips is estimated based on the current travel patterns of comparable house- holds, which may change in the future. Also, because income determines, in part, where people live, trips will be proportionally distributed to some extent toward high-income areas because higher-income households have a higher vehicle trip generation rate. Depending on the spatial distribution of the income groups, there are often large differences in the average number of trips per household in low-income versus high-income areas, which will be reflected in the model by indicating more or fewer trips for specific income groups. This approach is problematic, in part, because it relies on current travel patterns in forecast- ing. Low-income travel patterns are subject to constraints of budget and availability of transport. By using current patterns, TDMs may often carry forward current travel constraints into the future and ignore latent demand. However, many models will forecast future economic growth and future land use forecasts that reflect changes in socioeconomic patterns and capture poten- tial changes in travel patterns. The criteria used to identify projects for improvement can be influenced by these embedded modeling assumptions. For example, many projects are selected on the basis of improvements to level-of-service (although many other performance metrics are also used). If a route does not provide the desired level-of-service, it will be identified as a potential improvement project to address its deficient level-of-service. Because the models typically identify high trip rates among groups with higher-income levels or high rates of car ownership and then forecast even higher future trip rates among those groups, infrastructure improvements that will improve the level- of-service for high-income groups may receive greater attention and higher priority. However, the disparate modeling impact results may also be the result of inadequate land use forecast- ing rather than an implicit weakness in the model. If the land use forecast considers potential changes in demographic and economic patterns in all areas of the region, the model results may better reflect the travel demand of low-income areas. Within the field of travel demand modeling, the trip-based model is the most commonly used model. The primary advantage of the trip-based model, in fact, is its ubiquity and relatively lim- ited data requirements. However, one of the most significant criticisms of its use is its failure to treat travel as a derived demand. The trip-based model assumes that trip origins and destinations determine travel patterns but does not account for the specific activities that influence travelers’ behavior, such as scheduling constraints, response to cost or congestion, or varying levels of infor- mation that may inform each decision. Although many 4-step models may not reflect schedul- ing constraints, cost, congestion, or information provided to the traveler, there have been many improvements to these types of models so that these issues can be considered in the modeling process (see text box, Some Key Definitions Relevant to the Trip-Based, 4-Step Model Discussion). Traditional 4-step models typically assume that all travelers have the same VOT or segment the VOT into only a handful (typically three) of income groups. However, a growing body of research has shown that even within the same income level, travelers will have very different values of time. Ignoring these distinctions leads to aggregation error in which a small number of travelers may actually be willing to pay for a proposed tolling project. However, proprietary software has developed the capability of using multiclass assignments and allowing for the use of VOT distributions, which eliminates the problem of aggregation error. Activity-Based Travel Demand Model. An ABM is a “tour-based” model that treats travel as a derived demand—specifically, that individuals and households choose to participate in activi- ties that are spatially dispersed, which necessitates travel to different locations. In recognizing travel as a derived demand, ABMs address a variation of the traditional trip-based model. ABMs consider the linkages among all the activities and travel made by an individual over the course of

182 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox a day as well as, in most cases, the activities and travel of other household members (Cambridge Systematics, 2015). In a tour-based structure, a tour is represented as a sequence of connected trips that begin or end at a person’s home or workplace. The ABM can be used to model the tradeoff between making a stop at the grocery store on the way home from work versus making an additional tour for the sole purpose of going to the grocery store. Non-home-based travel can be linked to home-based travel by this spectrum of travel dimensions (see Figure 1). There is consistency across trips comprising each tour in terms of destinations, timing, and modal choice. When people make multiple tours in a day, there is also consistency in terms of purposes and timing as Some Key Definitions Relevant to the Trip-Based, 4-Step Model Discussion Trip. A unit of travel connecting any two locations. In trip-based models, a typical trip classification would be home-based work, home-based other, or non-home-based. Most existing models will have at least five to seven trip purposes. Impedances. During trip assignment, trips are allocated to routes proportional to the attraction of the desti- nation zone, but also inversely proportional to impedances (also referred to as “costs” in terms of time, tolls, fares, operating costs). The primary impedances for toll projects include distance, speed, capacity, and toll price. Speed, distance, and capacity are used to determine travel time. Skimming. A process of summing impedances along paths identified as the route in the travel network with the lowest cost for the traveler. Depending on how the model is structured, the cost may be the actual dollar value of the toll but may also include the monetized value of time. Typically, skimming uses travel time and cost which is converted to time. Gravity Model. The gravity model is the primary model used for trip distribution and represents trips as a prod- uct of attractions and productions, divided by an exponential function of travel costs. Typically these costs are measured only by travel times, but they may also include other factors (including the cost of a toll). The gravity model assumes that the number of trips between two zones is (1) directly proportional to the trips produced and attracted to both zones and (2) inversely proportional to the travel time between the zones. 1 T P A F K A F K ij i j ij ij j ij ij j n ∑( ) = = Where: i and j are travel zones T = Number of trips between zones P = Number of trip productions in zone A = Number of trip attractions in zone F = Friction factor (spatial separation between zones) K = Fudge factor and socioeconomic factors User Equilibrium. A condition at which all users within a transportation network have chosen routes that result in the lowest possible network travel time. Static user equilibrium is established over the entire analysis period; dynamic user equilibrium is established for each departure time. Calculating equilibrium is an appropriate approach for optimal system performance; it may be less appropriate when calculating individuals’ immediate response to a short-term incident or evacuation.

Using travel Demand Models for Environmental Justice Assessments 183 well as incorporation of joint activities or intra-household interactions (see text box, Some Key Definitions Relevant to the Activity-Based Demand Model Discussion). ABMs are gradually being adopted by several of the nation’s larger and medium-sized MPOs in support of their long-range transportation plans and studies. In a recent survey, six MPOs were reported to be currently using an ABM and 10 others were in the process of developing an ABM (Cambridge Systematics, 2015). These large-scale TDMs are increasingly being developed in regions that are undergoing or have considered pricing-related plans and projects. An ABM can be distinguished from the 4-step model in its use of a disaggregate approach to modeling travel demand. In ABMs, the activities and travel of every individual in the modeling region are simulated using a “synthetic population” approach. This approach makes it possible to report and analyze travel demand by various population segments (e.g., income, car avail- ability, worker status, age, gender). This differs from the 4-step model in which trips by various segments of travelers or households are modeled together with aggregate trips subsequently split by various choices (e.g., trips by mode or trips by time period). Thus, demographic information associated with household characteristics is preserved in an ABM. The U.S. Census PUMS data are used in ABMs that have been developed or are being developed in the United States to operationalize disaggregate person-level or household-level analysis. These models also typically rely on travel survey data to complete the analysis and require data to be properly sequenced to reflect representative chains of travel. ABMs are implemented in three main steps: 1. Population Synthesis. In the first step, disaggregate-level sociodemographic inputs are created by synthesizing the population for the study level. Representative households from PUMS were selected to correspond with the aggregate characteristics of the geography (often the TAZ Figure 1. Trips vs. tours. Trips are highlighted in red, tours are highlighted in yellow. (Adapted from NCHRP Report 406)

184 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Some Key Definitions Relevant to the Activity-Based Demand Model Discussion Tour. The sequence of trips that begin or end at the traveler’s home. For each tour, a primary purpose is assigned based on the most important activity on the tour (e.g., work). Demand Simulation. TDMs are built on a core set of probabilistic choice models. In a demand simulation, or microsimulation, or pseudo-random sample enumeration, a “Monte Carlo” sampling approach is used to simu- late a behavior rather than applying a fractional probability in a more deterministic way as would be generally done through trip-based models (Donnelly et al. 2010). While there are computational storage efficiencies for large problem sizes, it is possible to track the choices of individual agents across a series of upstream models to downstream choices. This use of the term “microsimulation” should not be confused with traffic microsimulation, where individual vehicles are simulated along roadways. PUMS Data for Synthetic Population Microsimulation. Researchers have used PUMS to support travel demand modeling, travel survey sampling validation and weighting, and the development of synthetic population micro- simulations (Tierney, 2012). In this context, microsimulation examines a sample of households or individuals and uses statistical weighting methods to represent a larger population group. The synthetic sample consists of a hypothetical set of households with a certain set of characteristics that match those of the larger population group. This complex modeling method is used to identify the travel impacts across many population groups, including multiple income levels and minority and ethnic status, among other categories. Household Travel Survey. These surveys typically provide socioeconomic data on households and the associated individuals (including income), travel-related activity information (including activity type and location), and the number of household vehicles. The surveys collect a different set of information depending on the model being used. ABMs may require additional detail, including the duration of the activity, timing, departure time, arrival time, and whether the activity was performed jointly or singly. Best practices for many elements of a TDM require the use of a household travel survey, including mode choice and trip generation. Because these surveys are very expensive, they are not frequently done as part of a TDM update. Stated Preference Survey. This is a household travel survey used to estimate travelers’ VOT for proposed tolls. A stated preference survey can be used to present a series of different managed lane options, including different toll rates versus time savings to respondents who then choose between the various options and associated prices. Revealed Preference Estimation. This is an estimation of travelers’ VOT based on actual travel patterns or by experimentation. In regions with existing toll and managed lanes facilities, revealed preference data may be used to estimate price effects. Some studies suggest that the VOT, when estimated by revealed preference, may be more than double the VOT when calculated using a stated preference survey. Discrete Choice Model. This is a model applied to traveler behavior to represent decisions regarding destina- tion, mode, and route. Many of the choices made are discrete, meaning that the choice of one prevents the traveler from selecting a different one. Choices are mutually exclusive and collectively exhaustive. The most common discrete choice model used is a logit function called a multinomial logit formulation. This model estimates the probability that an alternative is chosen given the utility of that alternative. Another common discrete choice model is a nested logit model that pools together similar alternatives. though other levels may be used) that are to serve as control totals for sampling households. In the making of a synthetic population, the end result is a table with a record for each household and person in the region, with attributes that are representative of the region’s actual popula- tion. The design of the synthetic population should reflect the policy analysis needs of the region or study. If a study requires an analysis of EJ populations (i.e., race, ethnicity, income) or other demographic considerations (e.g., age, gender), these attributes should be reflected in

Using travel Demand Models for Environmental Justice Assessments 185 the synthetic population. The objective in an ABM is to support the level of disaggregation in population characteristics needed to support microsimulation ABM methods. 2. Long-Term Choice Models. Using data from household travel surveys, a series of utility- maximizing discrete choice models is built to predict travelers’ activity-related travel decisions. The methods and sequence of the models vary significantly but all models generally include at least day-level activity pattern choice models (to capture all of the activities completed over the course of a typical day), tour-level choice models (to capture the combination of activities captured within a tour), and trip/stop-level choice models (to capture the stop location, choice of mode, and time of day). These models will also typically capture (and give precedence to) whether trips are completed jointly or singly. 3. Activity-Based Travel Model. Combining transportation network data similar in approach to that of the early stages of a trip-based model to assign traffic. ABM proponents have offered several reasons why the approach may have advantages over the trip-based methods: (1) time is treated as a continuum; (2) sequences and patterns of activi- ties and travel (i.e., tours) rather than individual trips are the focus; (3) linkages among various activity-travel decisions are recognized; (4) intra-household interactions, inter-personal and intra-personal consistency measures are considered; (5) space-time constraints on activities and travel are considered; and (6) individual level travel patterns are emphasized (Lin et al., 2008). An advanced model, such as an ABM, can be calibrated and validated along a broader range of criteria (Donnelly et al., 2010). Specifically, a “good-practice” ABM could have as many as 12 core models that can be individually calibrated, including: population synthesis, usual work- place location, auto ownership, tour generation, joint travel, tour destination, tour time-of-day, tour mode, stop location, trip time-of-day, trip mode choice, and assignment. Thus, the disaggre- gate framework of the ABM can facilitate calibration and validation in ways not available with a more aggregate 4-step model (Donnelly et al., 2010). However, having so many core models to calibrate makes the calibration process more complex. For its proponents, an ABM offers a framework to explore the sensitivity effects of policy changes, bringing more capability to examine complex questions, including the effects of pricing. This includes various forms of congestion pricing, dynamic real-time pricing, daily area pricing, and other innovative policies that cannot be modeled more precisely than the trip-based, 4-step model. The explicit modeling of joint travel in ABM, for example, was introduced to refine the modeling of high-occupancy vehicle/high-occupancy tolling facilities. As the time or cost of travel options change, the travelers in an ABM can be observed to respond by changing routes, mode, time-of-day, destinations, frequency of travel, or auto ownership. The traditional 4-step model, by comparison, is more limited and can consider changes to route, mode, or destination and, often, route and destination would be sensitive to highway travel time, but not changes to cost (Donnelly et al., 2010). However, recent applications of the 4-step model to toll facilities have applied changes to cost as part of the toll scenario analysis. The complexity of potential responses to pricing can be favorably represented within an ABM. Traditional models may assume that all travelers have the same VOT or will segment the VOT into only a limited number of income groups, although some recent applications have used VOT distributions with a greater number of income groups. Stated preference and revealed preference surveys of travelers indicate that VOT varies significantly between differ- ent groups of travelers and is usually higher when considered in real rather than hypotheti- cal scenarios. In an ABM framework, it is possible to assign VOT to individual travelers and simulate a continuous distribution to support a detailed accounting of how costs are borne by different travelers by their location, income, occupation, or how they respond to peak period congestion (Donnelly et al., 2010). The 4-step process with multiclass assignments also has the same capability.

186 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox An ABM offers a more detailed analysis of outputs that can be very informative for assessing EJ. In a trip-based model, demand model results report trip tables segmented by purpose and mode while in an ABM, the decisions of individual travelers are simulated so that the model reports its results as a list of individual households, persons, tours, and trips (Donnelly et al., 2010). This detailed output provides a platform for assessing transport mode and demographic dimensions such as income, age, gender, or any other category included in the synthetic population (Donnelly et al., 2010). This complex differentiation of the population in the reporting platform enables consideration of pricing policy differences such as toll exemptions or discounts for low-income households. The use of ABMs is not without challenges according to the literature researched and the sur- veys. In a recent survey of MPOs, several concerns or perceived barriers were noted, including greater complexity of modeling components and more complex formulations than conventional models, greater expense to develop the tool, difficulties in debugging or tracing errors and in obtaining calibration or validation data, long run-times in operating models, potentially greater hardware requirements, the need for custom software, and few demonstrated examples of model transferability between various regions (Cambridge Systematics, 2015). Nonetheless, the perceived benefits of ABM, along with the concerted effort to build ABMs, have sparked an increasing interest in microsimulation-based modeling systems and several micro simulation platforms have been developed based on the activity-based paradigm of transportation demand forecasting (Lin et al., 2008). Traffic Assignment Modeling. Most ABMs, similar to trip-based models, require an assign- ment of vehicle trips to the roadway network. Nearly all regional TDMs use static traffic assignment methods that are founded on a deterministic user equilibrium methodology to as sign traffic demand to network links. Static user equilibrium methods are able to predict volumes and travel times on congested networks and this has generally sufficed for the evalu- ation of capacity-enhancing projects. The static assignment method has generally recognized benefits, not least of which, are its well-understood methods and assumptions supported by years of research, its reliability in returning the same results each time the model is run with the same inputs, its computational efficiency, and its prevalence in available transportation mod- eling packages (Cambridge Systematics, 2015). However, as transportation networks become more heavily congested, policymakers and researchers are seeking to understand the effects of various regional and corridor-specific solutions on both the supply- and demand-side beyond lane capacity enhancements, such as toll pricing, signal optimization, synchronization, and real-time information systems. These types of operational changes and scenarios can reveal some of the limitations of the deterministic methods. The static assignment models have been faulted for their inability to capture the effects of upstream or downstream congestion or the effects of bottlenecks, intersection geometry and delay, transit vehicle interactions, and queueing (Cambridge Systematics, 2015). Vehicles can be assigned to links beyond their ultimate capacity. Static assignment methods also cannot represent time-varying flows and congestion or the impacts on travel times and costs with detail necessary for some analyses (Cambridge Systematics, 2015). This can result in significant discrepancies between the VOT as used in the demand model and in the network simulation. DTA models, by comparison, have the ability to represent time-varying network times and costs. DTAs can offer more information about network performance by time-of-day—a topic of increasing interest in setting transportation policies, and which can be used as input into TDM components. The DTA model’s treatment of congestion and spill-back effects is argued to result in more realistic travel time measures than the congested travel times reporting from static traffic assignment models (Cambridge Systematics, 2015). DTA relies on a large-scale simulation

Using travel Demand Models for Environmental Justice Assessments 187 of all the drivers and their interactions with traveler information and the roadway network. One intriguing feature of the DTA models is its ability to draw on the simulated individual values of time from the ABM and incorporate it into its driver simulation, thereby preserving the disaggregate nature of both tools. This differs from the static model in which individual traveler information is aggregated into average values at the TAZ level for purposes of traffic assign- ment (see text box, Some Key Definitions Relevant to the Traffic Assignment, Microsimulation, and Mesoscopic Discussions). Some Key Definitions Relevant to the Traffic Assignment, Microsimulation, and Mesoscopic Discussions Static Network Assignment. This is a method of determining travel time and cost in which variables of interest are time-invariant. These models typically perform analysis using relatively broad time-of-day divisions. In these models, travel time increases as traffic volume increases; volume may increase indefinitely and exceed the physical capacity (vehicles per hour) of the infrastructure. Dynamic Traffic Assignment Techniques. DTA techniques are developed for accommodating temporal dynamics of demand and are well-suited for integra- tion with ABMs. DTA is a time-varying process that describes changes to traveler’s trips based on congestion and other factors and can model changing start times and changing routes mid-trip. DTA uses iterative procedures that capture the relationship between traffic on a route with overall network assignment. DTA models may be built on a variety of behavior assumptions and model formu- lations (e.g., equilibrium or system-optimal, or non-equilibrium or reactive). DTA can be used either with aggregate or disaggregate models and is typically contrasted with static network assignment. The main characteristics of DTA include: (1) vehicles that depart at different times are assigned different routes; (2) vehicles departing at the same time, with the same origin and destination but different routes, should have the same experienced travel time; and (3) experi- enced travel time can only be realized at the end of the trip (not at departure, as is typical with other models). The network and demand specifications of traffic simulation models are used but operate with greater temporal detail. They typi- cally use link-based simulation models that enable more robust estimates of link travel times and costs. DTA Model Output. Aggregate measures include time-dependent static measures, such as flows, travel times, vehicle miles traveled (VMT), and vehicle hours traveled (VHT), as well as queues, densities, bottleneck locations, length, and dissipation time. Disaggregate model output includes vehicle trajectories describing the position of each vehicle in each time. These detailed trajectories can also inform subsequent operational performance analyses such as control delay through the Highway Capacity Manual or support reliability performance measures at the planning level (Cambridge Systematics, 2015). Microscopic Scale Traffic Simulation Models. These models simulate the movement of individual vehicles based on car-following, lane-changing, and gap-acceptance theories and are used for the purpose of conducting detailed analysis of geometric design configurations and

188 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox operational plans. Microsimulation models are used to model individual vehicle movements on a second or subsecond basis, assessing and illustrating the traffic performance of highway and street systems, transit, and pedestrians. Microsimulation is particularly relevant for modeling complex systems under congested conditions, illustrating through visualization tools the inter- action of vehicles and the effects that operational difficulties in one segment can have on the upstream and downstream flows. For tolling projects, microsimulation is particularly useful in capturing temporal shifts of traffic among different roadway facilities and identifying capacity constraints that may develop from toll implementation. In application, microsimulation modeling requires detailed atten- tion to geometric representation, lane configurations and coding of traffic control systems, and signalization. Because of their data requirements, such models typically have focused on smaller areas, but larger areas such as freeway corridors and city-wide traffic microsimulations such as in Phoenix, Arizona (Morgan, 2013), have also been modeled. These efforts have been advanced by technological efficiencies in GIS, remote sensing, online traffic, signal optimization programs, and graphic user interfaces (GUI) linked to macroscopic traffic assignment models (Donnelly et al., 2010). Nonetheless, for an urban region, where a microscope scale model may be applied, the zonal structure of a TDM is too coarse to support the level of detail required to conduct the analysis. Traffic count data and spatial adjustments to origin-destination matrices to identify sinks and sources within a traffic zone and temporal adjustments to encapsulate smaller units of time, for example, are required (Donnelly et al., 2010). Newer microscopic models are route-based (i.e., vehicles select a route at departure and follow that route but may update that route choice over the course of the journey as part of the simula- tion). A one-shot, non-iterative assignment-simulation is commonly used in many microsimu- lation models in which vehicles departing at different times are assigned a route that is updated based on current travel times. With this approach, travelers seem to strictly adhere to guidance received prior to departure with no update over the course of travel. However, some microsimu- lation models will allow vehicles to change route based on updated information on the shortest route at a later time. The most sophisticated microsimulation efforts fully integrate DTA and combine the movement of traffic (in a traffic simulation model) with the movement of travel (network demand) in an iterative process. Mesoscopic Models. These models typically share certain features of microscopic models, but lose some of the details associated with those models. In mesoscopic models, individual vehicles are grouped into a “cell” treated as the simulation entity. Certain dynamic states are simulated but detailed inter-vehicle interaction may not be included (including lane-changing or gap-acceptance). DTA is strongly associated with mesoscopic models. However, mesoscopic simulation offers much faster simulation speed and can be applied to a much larger network. The primary disadvantage of microsimulation is the time, cost, and level of expertise required. The models are widely seen as data-hungry to calibrate the model and, as a result, more suscep- tible to small errors either in data input or model parameters. DTA models, in particular, can take a long time to build and calibrate compared to static models. If there is excess demand or network errors, the model can present oversaturation or gridlock conditions that can be hard to detect (Cambridge Systematics, 2015). Good convergence of equilibrium is not guaranteed in model development, and it can take multiple network corrections and a series of time-consuming calibration steps to achieve. Given the model’s complexity, it can also be time-consuming to run the necessary iterations and requires attention to hardware and processing requirements. How- ever, the simulations are able to capture socioeconomic characteristics of travelers and model different behavior assumptions.

Using travel Demand Models for Environmental Justice Assessments 189 Modeling a Toll Corridor. Toll corridor studies require a refined level of detail that involves a subarea analysis. There are three main approaches in use for subarea analysis, including subarea focusing, windowing, and multiresolution modeling. • Subarea focusing uses an existing TDM, but adds substantial detail to the TAZs in the study area. • Windowing is often used when the regional TDM would be cumbersome to use. Windowing is used to isolate a small area from the rest of the traffic network and is typically used when count stations make it possible to prepare origin-destination tables for traffic moving into the area. • Multiresolution modeling (MRM) is becoming increasingly common in toll studies because it makes it possible to more accurately reflect vehicle movement, queue formation, queue dis- sipation, bottlenecks, lane-changing behavior, and other factors critical to toll studies. MRM typically involves incorporating a macroscopic regional travel forecast with microsimulation, though there is no established method for creating and applying MRM models. What Are Some Examples of Implementing This Tool? Understanding the impacts of tolling on EJ groups can be obtained through a variety of travel demand modeling approaches. The following discussion showcases how three different projects used travel demand methods and models to assess pricing effects on EJ populations and travelers. Example 1: Regional Tolling Analysis, Mobility 2035, Dallas–Fort Worth Region, North Central Texas Council of Government. NCTCOG applied an Environmental Justice Index method to identify “communities of concern” at its travel survey zone (TSZ) level as either low-income or minority (see text box, North Central Texas Council of Government’s Dallas– Fort Worth Model, and case example, “Using an EJ Index to Identify Affected Populations, Dallas–Fort Worth Metro Region”). This method was used to support the preparation of EJ analyses of the Regional Transportation Plan and to conduct a separate Regional Tolling Analysis (RTA) described further below. For the RTA study, regional traffic was modeled under three transportation network conditions: 1. Existing network (using then-current demographics), 2. Proposed future 2035 build network, and 3. Proposed future build network without priced facilities. The Dallas–Fort Worth TDM was used to evaluate transportation system performance mea- sures, including average vehicle trip time (in minutes), average vehicle trip length (in miles), and average vehicle trip speed (in miles per hour) for EJ and non-EJ zones across the three modeled transportation network conditions. For transit trips, utilizing a similar framework, the RTA examined the number of transit trips, average transit trip lengths, and highest average speeds in the morning peak by EJ status. Daily congestion levels were also comparatively measured by categorizing the TSZs in terms of their level of congestion—no, light, moderate, and severe—and their EJ status. In all three network conditions, EJ TSZs were found to have fewer no congestion and severe congestion conditions, but more light to moderate congestion conditions. Where a large difference between EJ and non-EJ TSZs was found for “no congestion” zones, NCTCOG found the difference attrib- utable to the fact that EJ communities were less prevalent in the region’s rural locations. Layering in the three transportation network conditions, NCTCOG’s RTA report examined the regional origin-destinations of the morning peak period travelers. Efforts included an analy- sis of the use and distribution of trips on priced facilities and compared this distribution in terms

190 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox of the EJ status of the originating TSZ. Table 1 shows how this information was organized to succinctly encapsulate the differences in travel patterns according to these several dimensions. The analysis illustrates, for example, that 5 percent of the trips from EJ zones would use tolled facilities compared to 6.6 percent for non-EJ zones in the base case network in 2013. NCTCOG attributed this lower percentage of usage to the fact that the geographic location of existing toll roads tended to be closer to the non-EJ communities. While vehicle trips grow from EJ zones in the future build condition, they still represent only 8.5 percent of all trips. Example 2: Louisville–Southern Indiana Ohio River Bridges Project. Over the course of the Louisville–Southern Indiana Ohio River Bridges project, the regional trip-based model was updated to give more focused attention to the tolling and traffic diversion issues of the project (see text box, Louisville–Southern Indiana Ohio River Bridges Time of Day Model). Among the adjustments, the Time of Day (TOD) model disaggregated internal trip purposes based on the income of the household making the trip. The disaggregation of trip purposes is carried through mode choice and traffic assignment. The consultants responsible for developing the Louisville–Southern Indiana Ohio River Bridges TOD model gave the following reasons for carrying an income dimension to the trip purpose throughout the steps of the modeling: • By including income in the home-based work trip purpose, high-income employment can be linked to high-income households, thus improving the trip distribution patterns in the model. North Central Texas Council of Government’s Dallas–Fort Worth Model NCTCOG’s Dallas–Fort Worth model was used to conduct a traffic analysis performance report, travel time comparisons, and origin-destination studies. The Dallas–Fort Worth model, an application of software components using the TransCAD platform, is a 4-step trip-based TDM covering North Central Texas. User surveys and traffic counts were used to validate the model (i.e., ensure that roadway volume, transit usage, peak and off-peak conditions, and roadway speeds were reproduced in the model’s outputs). In describing the model, NCTCOG acknowledged several data and modeling limi- tations, including: • Future year demographic projections assume the same distribution of income, race, and ethnicity. Potential shifts in population types across the region are not reflected in the modeling approach. • No data about race, ethnicity, and economic status of the users of priced facili- ties in the model were available, although the agency reported plans to con- duct such a survey within the next year. • The model cannot identify trips based on the race or ethnicity of individual users. • Income quartiles are only used in the assignment of home-based work trips, which account for only 25 percent of trips; all other vehicle trips are not assigned based on income. • All vehicle trips of the same type are treated identically during the trip distri- bution, model choice, and traffic assignment steps. The model cannot produce outputs that differentiate vehicle trips based on the economic characteristics of transportation system users.

Data of Interest All Municipal Planning Area TSZs Environmental Justice Status Environmental Justice TSZ Type Non-EJ Status TSZ EJ TSZ Low-Income Alone Minority Alone Both Low- Income and Minority TSZs in Region 5,252 2,978 (56.7%) 2,274 (43.3%) 111 (2.1%) 1,942 (37.0%) 221 (4.2%) 2013 Population 6.778,201 3,802,580 (57.0%) 2,915,621 (43.0%) 70,191 (1.0%) 2,613,464 (38.6%) 231,996 (3.4%) 2035 Population 9,833,378 5,977, 328 (60.8%) 3,856,050 (39.2%) 98,372 (1.0%) 3,462,607 (35.2%) 295,071 (3.0%) TSZ Utilizing Priced Facilities (at least once per day) 2013 Network 4,923 (93.7%) 2,728 (55.4%) 2,195 (44.6%) 87 (1.8%) 1,896 (38.5%) 212 (4.3%) 2035 Priced Facility No-Build Network 5,096 (97.0%) 2,843 (55.8%) 2,253 (44.2%) 105 (2.1%) 1,930 (37.9%) 218 (4.3%) 2035 Build Network 5,184 (98.7%) 2,921 (56.3%) 2,263 (43.7%) 106 (2.0%) 1,937 (37.4%) 220 (4.2%) Vehicle Trips Utilizing Priced Facilities from TSZs with Any Priced Facility Trips 2013 Network 193,257 125,322 (64.8%) 67,935 (35.2%) 1,678 (0.9%) 63,473 (32.8%) 2,784 (1.4%) 2035 Priced Facility No-Build Network 303,587 211,302 (69.6%) 92,285 (30.4%) 3,809 (1.3%) 84,872 (28.0%) 3,604 (1.2%) 2035 Build Network 476,640 319,316 (67.0%) 157,324 (33.0%) 5,303 (1.1%) 144,394 (30.3%) 7,627 (1.6%) Vehicle Trips On Entire Transportation Network from TSZs with any Priced Facility Trips 2013 Network 3,273,568 1,902,424 (58.1%) 1,371,144 (41.9%) 41,079 (1.3%) 1,222,605 (37.3%) 107,461 (3.3%) 2035 Priced Facility No-Build Network 4,816,738 2,974,797 (61.8%) 1,841,941 (38.2%) 62,529 (1.3%) 1,639,361 (34.0%) 140,051 (2.9%) 2035 Build Network 4,823,072 3,013,182 (62.0%) 1,949,169 (38.0%) 64,105 (1.3%) 1,644,695 (33.8%) 140,051 (2.9%) Percentage of Vehicle Trips (from TSZs with any Priced Facility Trips) Utilizing Priced Facilities 2013 Network 5.9% 6.6% 5.0% 4.1% 5.2% 2.6% 2035 Priced Facility No-Build Network 6.3% 7.1% 5.0% 6.1% 5.2% 2.6% 2035 Build Network 9.8% 10.6% 8.5% 8.3% 8.8% 5.4% Source: NCTCOG, 2014 Table 1. Morning peak period origin-destination by EJ status over three network conditions.

192 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox • Mode choice segments the market for transit by trip purpose by income. Income can be used to approximate auto availability and the willingness to use transit for choice riders. • The traffic assignment model uses a generalized cost approach that includes operating cost and tolls in the future. Segmenting trips by income allows for VOT testing that is based on income groups. PUMS proved instrumental in developing the model’s capacity to consider income disaggrega- tion by trip purpose. Three income groups were defined using PUMS from the U.S. Census in 2000: $0 to $40,000, $40,000 to $60,000, and more than $60,000. Through PUMS data, households were also disaggregated by their dwelling type (i.e., single family and multi-family), household size, and vehicle ownership. Ultimately, household estimates by TAZ for some 40 distinct categories were defined as the basis for deriving trip productions. The trip generation model estimated a set of trip productions by purpose for each income group in the model. For home-based work, the model used this information to link the household income to appropriate job types in the region. Because PUMS data reports household income distribution by employment type [North American Industry Classification System (NAICS) employment sectors], it was used to understand trip attractions by income and improve the trip distribution patterns of the model by income. In the model’s interim development phase, an important objective was to ensure consistency with work performed by the toll and finance consultant. To support consideration of tolling effects, the toll cost was translated into travel time terms as a penalty for traversing specific links in the network. Thus, the toll cost was converted to a time using a VOT factor. The resulting travel time was integrated into the trip distribution step of the model and the path choice in traf- fic assignment. By including the toll time as a penalty in trip distribution, the effect of the toll was to limit trip-making between zone pairs that require the use of a toll link. Louisville–Southern Indiana Ohio River Bridges Time of Day Model The Kentucky Transportation Cabinet (KYTC) and Indiana DOT commissioned the development of a new Time of Day (TOD) model. The Louisville–Southern Indiana Ohio River Bridges TOD model is based on the regional TDM used by Kentuckiana Regional Planning & Development Agency, the region’s MPO for metropolitan transportation planning. The TOD model was developed to facilitate more detailed analysis of tolling effects on traffic patterns. The Louisville–Southern Indiana Ohio River Bridges enhanced the regional model in several ways by: • Providing income-stratified trip purposes in the trip generation model; • Creating time-of-day structure (morning peak, midday, evening peak, night); • Updating the mode choice model to reflect the new TOD structure; • Updating the socioeconomic data to more currently available information; • Incorporating the latest available traffic volumes, including counts collected for this purpose; • Adding a truck model component; and • Improving the traffic assignment methodology The TOD structure of the model presented traffic for eight separate periods. This switch eliminated the need to apply factors for design hour and directional movements to daily volumes as was required for the former 24-hour model.

Using travel Demand Models for Environmental Justice Assessments 193 EJ populations were identified at the block group level using U.S. Census data; block groups were evaluated based on total population, total minority population, and the total population below the poverty level. Similar to the Dallas–Fort Worth example, this approach considered trips that originated in an EJ block group, but could include trips taken by travelers that are not, in fact, low-income or minority populations (i.e., EJ communities). The model was used to calculate the average user cost per trip, including all vehicle user costs (i.e., operating costs, total cost of time spent in a passenger car, and toll costs) for all trips within the planning area. The results suggest when all vehicle user costs are considered, the average cost per cross-river trip is lower for cars containing EJ communities than for cars for the non-EJ com- munities, although the financial burden is proportionately greater for low-income persons. While the cost of the tolls would not be predominantly borne by EJ populations, the increase in average user costs would be relatively greater for EJ travelers. The consequences of toll implementation were ultimately found to be appreciably more severe or greater in magnitude for EJ populations, and this finding was reflected in the FHWA’s Revised Record of Decision. Further study of mitiga- tion implementation steps was subsequently undertaken to address concerns expressed by FHWA (see tool, “Assessing User Costs and Household Burden Effects”). The traffic forecasts in the Supplemental Final EIS were prepared using the TOD model devel- oped for the project, and these forecasts informed the findings of the EJ assessment. After pub- lication of the Revised Record of Decision, an investment grade traffic and revenue study was prepared for the Kentucky Transportation Cabinet (KYTC) to inform feasibility assessments and attract potential financing. The subsequent traffic and revenue study adopted a more conservative estimate of the cross-river trips on the new and rehabilitated bridges, adjusting socioeconomic growth and travel demand levels in the region. The traffic and revenue study also differed from the TOD model in that it only forecasted traffic volumes for a “Build” scenario. As such, it could not be used to compare the likely changes in traffic or in average user costs between a “Build” and a “No-Build” scenario. Example 3: Mobility, Access, and Pricing Study, San Francisco County Transportation Authority. The San Francisco County Transportation Authority (SFCTA) examined the effects of various pricing strategies to manage travel in the urban core of San Francisco as part of its Mobility, Access, and Pricing Study (MAPS). This congestion pricing feasibility study was funded through an FHWA Value Pricing Pilot Program grant in 2006. For this project, SFCTA refined its existing ABM to complement the economic and policy analyses and public outreach activities needed to thoroughly address the pricing effect on travel, including how it may be borne by the region’s low-income and other traditionally disadvantaged populations. SFCTA was particularly well-positioned to carry out the demand model refinements for the Regional Pricing Study, having been an early user of ABMs (see text box, San Francisco County Transportation Authority Regional Pricing Model). Earlier, in planning for its Doyle Drive study in 2001, SFCTA had determined that the then-current Metropolitan Transportation Commis- sion (MTC) regional model for the nine-county Bay Area Region would not adequately support the evaluation of the types of alternatives under consideration. SFCTA was interested in better understanding the travel patterns and congestion effects by TOD, but the MTC model at the time was only a peak period/24-hour model and could not assign trips by TOD or offer the level of geographic detail required. This led SFCTA toward the development of its initial model, San Francisco County Chained Activity Modeling Process (SF-CHAMP), one of the first ABMs used in practice. With the FHWA Value Pricing Pilot Program grant funding, SFCTA was able to develop its Regional Pricing Model (RPM-9) to assess the impact of peak period tolls for entering an area surrounding downtown San Francisco, with the goal of inducing travelers to shift out of the

194 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox San Francisco County Transportation Authority Regional Pricing Model While the San Francisco County Chained Activity Modeling Process (SF-CHAMP) has been continually used for a variety of transportation projects, the MAPS study team needed greater geographic breadth, toll, and TOD sensitivity to evaluate the congestion pricing strategies integral to the MAPS study. The Nine-County Regional Pricing Model (RPM-9) contains several refinements and enhancements to the existing model. • Added Feedback Loops. The SF-CHAMP model did not include a feedback loop from assignment to travel demand. This limited the responsiveness to travel demand to auto levels of service. • Expanded Geography. The revised model predicts the daily activity patterns and tours of all Bay Area residents, moving away from the prior hybrid model approach that relied on the MTC model for non-SF resident travel. A sampling routine was established to handle data processing memory issues. • Toll-Choice. The SF CHAMP model treated toll costs as a generalized cost of a particular link in highway assignment. Each vehicle in a user class had the same VOT. Moreover, in the auto nest of the tour choice model, it was possible to be a “driver” or “passenger” but the possibility that a driver of a three- person shared ride vehicle could split costs lowering the total cost of a toll was ignored. RPM-9 differs in that it included separate nests under the auto nest to account for ride sharing (i.e., drive alone, shared ride 2, shared ride 3+). Importantly, under each of these choices is a set of choices—value toll, no value toll, or already-paid value toll—to address the distinct complexities of area and cordon pricing. • Accessibility. Mode choice and destination choice logsums represent accessibility in tour choice and auto ownership models. This approach allows different popu- lation segments to value level of service attributes in accordance with the mode choice and destination models. • Time-of-Day Choice. The study sought to understand if pricing could be used as incentive to induce travelers to move travel to outside the peak periods. The RPM-9 model contains two TOD models. • Values of Time. Acknowledging the heterogeneity of VOT is important for pric- ing. For mandatory tour purposes, the mean VOT is determined as a function of income and for non-mandatory tour purposes at two-thirds of the mandatory purpose. The distribution of VOTs was derived from a stated preference survey and a maximum likelihood estimation for mixed logit models. • Highway and Transit Assignment. With its broader geographic expansion, the RPM-9 model required intensive refinements of network coding and final high- way assignments, enabling a maximum of 99 iterations toward convergence to ensure consistency in scenario comparisons. • Tolling Discount Policies. Policies for tolling discounts needed to be coded into the RPM-9 model, including for “pass-thru” trips and a bridge toll “feebate” which gives discounts to cars that have already paid a bridge toll. Distinct logical treatments for area and cordon pricing were also required. Source: Sall et al., 2010; Perez et al., 2012

Using travel Demand Models for Environmental Justice Assessments 195 peak periods or to switch to transit. The study team focused on two basic pricing approaches: “area pricing,” where drivers are assessed a fee for any vehicle movement inside the boundary of a designated zone, including entry, exit, and travel entirely within the zone and “cordon pricing,” where drivers are assessed a fee only for movements that cross the boundary line and internal movements are not charged. TOD models had to be sensitive to cost and the choice of destinations also had to consider cost to identify how many people would likely shop or recreate elsewhere, rather than switch to transit or travel in the off-peak period. Because it linked trips into tours and simulated individual travelers, the RPM-9 could also model area pricing scenarios in which travelers would pay a fee for driving anywhere within a specified zone. The TOD model was informed by a stated preference survey of travelers driving to downtown San Francisco. The stated preference survey, administered in 2007, was designed to help under- stand a traveler’s response to a potential entry fee into the downtown area. In this case, a total of 663 respondents were asked to trade off cost, shifts in their trip time, and mode shifts in a series of experiments (Perez et al., 2012). The resulting stated preference data were used to estimate the VOT distributions for use throughout the modeling. Using a mixed logit model, a joint mode and departure time choice model was estimated. Mixed logit was applied because it allows the user to estimate a distribu- tion on a coefficient, rather than just the mean value (Sall et al., 2010; Perez et al., 2012). In this case, a distribution was estimated on the travel time variable, asserting a lognormal form. The cost coefficients were also estimated as standard coefficients segmented by income. The derived mean and median VOT are shown in Table 2. When the estimated VOT distributions are plot- ted as lognormal functions, the curves exhibit the shapes shown in Figure 2. These distributions make clear the differences in VOT by income and they were used in the RPM-9 model (Perez et al., 2012). The RPM-9 model’s outputs include various performance measures of the various congestion pricing scenarios used as evaluation criteria (see Table 3). In the context of the MAPS study, SFCTA focused on making travel options available for all communities, seeking to minimize vehicle hours of delay and finding strategies for improved frequency and reliability of transit. The overall effect on mobility was measured by the total number of trips made (Sall et al., 2010). However, as a disaggregate model, the RPM-9’s microsimulation stores daily travel pat- terns of every individual, allowing for comparisons of the mobility of low-income households with that of the total population (Sall et al., 2010). Thus, the study team could discern that the proportion of low-income drivers in the peak was relatively low (approximately 5%) with the Income Groups VOT Median VOT Mean $0–$30,000 $4.12 $6.01 $30,000–$60,000 $6.03 $8.81 $60,000–$100,000 $7.15 $10.44 $100,000 $8.80 $12.86 Source: Perez et al., 2012 Table 2. Mean and median value of time estimates by income group, from the stated preference survey and the mixed logit estimation in the San Francisco region.

Figure 2. Estimated value of time distributions from mixed logit estimation (2005 Dollars). Source: Sall et al., 2010 Table 3. Select evaluation criteria and measures used for San Francisco County Transportation Authority cordon pricing study. Evaluation Criteria Categories Measures Applied Use of Measure for Study Mobility Vehicle hours of delay Time spent in congestion Peak/off-peak travel time ratios Auto and transit modes measured between key origins and destinations to identify where mobility could be improved by time of day Transit system performance/loads Corridors with high transit potential and corridors with overcrowding that require greater service frequency; potential target transit routes for revenue recycling investments Traffic volume changes Maps in response to toll pricing that indicate diversions around the tolled cordon. Accessibility Overall trip-making and mode share Regional Pricing Model (RPM-9) reported daily person trips, peak period mode share, trips by mode by income group Equity Trips by mode by income group RPM-9 model reported trips by mode by income group, including trips made by lower-income households (earning less than $50,000), zero-car households, and residents within or adjacent to a priced zone Health Emissions Emissions are calculated based on vehicle miles traveled by speed and facility type Collisions Collisions are calculated based on vehicle miles traveled by facility type Revenue Generation Number of toll transactions program Financial consultants developed a separate, detailed financial model to estimate the amount of revenue that would be generated and to test financing approaches for implementation as well as program costs for toll collection Amount and number of discounts Various policies were explored for individual vehicles Source: Sall et al., 2010; San Francisco County Transportation Authority, 2010

Using travel Demand Models for Environmental Justice Assessments 197 majority reaching the focus area core by other transport modes. Moreover, under the SFCTA’s Preferred Cordon Scenario, the daily trips to the subject focus area were expected to be reduced by 38 percent, but the trips made by low-income households would only decrease by 1.5 percent, leading to the finding that pricing would have a positive effect on the mobility of low-income households relative to the general population. With stakeholder input and feedback, the SFCTA study team evaluated a range of potential discount policies. The technical analysis considered the impact of discounts on system perfor- mance and financial feasibility, which helped shape recommendations. Transit vehicles were to be exempt from the fee, as were taxis, which were viewed as an extension of the transit and paratransit fleet in San Francisco. Program discounts of 50 percent of the cordon charge were contemplated for low-income drivers and disabled motorists primarily to address income equity considerations and a similar discount eligible to residents within the cordon zone to address a geographic equity issue. A “feebate” program for Bay Bridge toll payers, shaving $1 off the cordon toll, was also proposed. The disaggregated nature of the model helped support revenue impact estimates attributable to particular forms of toll discounting. The disaggregated approach made it possible to evaluate impacts on low-income users across the entire area of study with additional information about household size and income. In the case of the RPM-9, individual and household characteristics were derived from the PUMS data. The approach taken seeks to overcome the “aggregation bias” inherent in EJ analysis approaches that rely on zone-based EJ mapping. In such cases, by comparison, affected communities of concern are identified at the TAZ level when a certain percentage of low-income or minority households are present. This kind of characterization masks forms of racial or economic diver- sity through the use of the “threshold” method of defining zones, including pockets of low- income and minority residents who may live within a “non-EJ” zone. By evaluating travelers at an individual level, the model is able to specifically target improvements that benefit low-income or minority groups. In the case of San Francisco, low-income households were evaluated under a baseline and a build alternative. What Are Its Limitations? Several limitations of the various models and approaches have been presented. Within the constraints of the models currently in use, a number of minor improvements can be imple- mented to improve the output of the models with respect to analyzing EJ in toll implementation and pricing projects. First, and probably most relevant for EJ assessment, there are considerable advantages in better segmentation of travel markets when considering the various uses and needs of travelers. Income, age, and gender all influence travelers’ VOT. Employed travelers, particularly when traveling for work, also are more likely to have a higher VOT. Larger households and lower- income households are more likely to make joint trips, carpool, and take advantage of the managed lanes because of the opportunities to share costs. Second, many TDMs for toll projects treat road users’ willingness to pay for modeling as a single value. This can be overcome by applying a probabilistic distribution of VOT, instead of using a single deterministic value. VOT distribution is currently being used for most new toll studies. A probabilistic distribution of VOT also allows policymakers to understand the relative likelihood of low-income travelers choosing to use managed lanes. Third, the influence of different technology types needs to be better explored. The use of electronic toll collection technology may accompany a change in travelers’ willingness to pay.

198 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox What Types of Resources and Costs Are Required? Existing TDMs may be used to evaluate impacts in EJ populations based on available resources. Expertise in modeling software is necessary to obtain information specific to the EJ population. As noted earlier, plans for the development of activity-based, tour-based models are under way in several of the nation’s larger MPOs but will likely require significant investment of resources and likely take several years to develop. For example, the average time reported by six MPOs currently using ABMs was about 4.5 years, with a range of 3 to 6 years (Cambridge Systematics, 2015). However, data on low-income and minority populations, combined with revealed and stated preference type survey information obtained through an investment grade study, may make it possible to produce a defensible detailed analysis. Who Has Used It Successfully? Various approaches for employing TDMs have been presented and referenced. The discus- sions have highlighted some major advantages and limitations of the various analytical tools currently being used or developed. Much more information can be found by exploring the references presented in the resources section below. Resources Bhat, C., Srinivasan, S., Guo, J., and Sivakumar, A. 2003. Activity-Based Modeling of Travel Demand. Handbook of Transportation Science. Retrieved from http://www.caee.utexas.edu/prof/bhat/ABSTRACTS/TSHANDBK.pdf. Bujanda, A., Aldrete, R., and Navarro, U. 2014. “Estimating Transponder Trips Integrating Binational Communi- ties as Single Transportation System.” Transportation Research Record: Journal of the Transportation Research Board, No. 2302. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 51–64. Cambridge Systematics, Inc. 2015. Status of Activity-Based Models and Dynamic Traffic Assignment at Peer MPOs. Metropolitan Washington Council of Governments. Washington, D.C. Retrieved from https://www.mwcog. org/uploads/committee-documents/aVxfXVhW20150827091119.pdf. Castiglione, J., Bradley, M., and Gliebe, J. 2015. Activity-Based Travel Demand Models: A Primer. SHRP 2 Report S2-C46-RR-1. Retrieved from http://onlinepubs.trb.org/onlinepubs/shrp2/SHRP2_S2-C46-RR-1.pdf. Castiglione, J., Hiatt, R., Chang, T., and Charlton, B. 2006. “Application of Travel Demand Microsimulation Model for Equity Analysis.” Transportation Research Record: Journal of the Transportation Research Board, No. 1977. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 35–42. Chiu, Y., Bottom, J., Mahut, M., Paz, A., Balakrishna, R., Waller, T., and Hicks, J. 2011. Dynamic Traffic Assignment: A Primer. Transportation Research Board Circular E-C153. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Retrieved from http://onlinepubs.trb.org/onlinepubs/circulars/ ec153.pdf. Chiu, Y., Nava, E., Sloboden, J., Lewis, J., and Alexiadis, V. 2012. Traffic Analysis Toolbox Volume XIV: Guidebook on the Utilization of Dynamic Traffic Assignment in Modeling. FHWA, Washington, D.C. Retrieved from http:// ops.fhwa.dot.gov/publications/fhwahop13015/fhwahop13015.pdf. Donnelly, R., Erhardt, G., Moeckel, R., and Davidson, W. 2010. NCHRP Synthesis 406: Advanced Practices in Travel Forecasting. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Duthrie, J., and Waller, S. 2007. Incorporating Environmental Justice Measures into Equilibrium-Based Transporta- tion Network Design Models. Texas Transportation Institute. College Station, Texas. Retrieved from http:// d2dtl5nnlpfr0r.cloudfront.net/swutc.tamu.edu/publications/technicalreports/167265-1.pdf. FHWA. 2011. TAZ Delineation Business Rules. Retrieved from https://www.fhwa.dot.gov/planning/census_issues/ ctpp/data_products/tazddbrules.cfm. FHWA. 2014. Time-of-Day Modeling Procedures: State-of-the-Practice, State-of-the-Art. Retrieved from https:// www.fhwa.dot.gov/planning/tmip/publications/other_reports/tod_modeling_procedures/ch06.cfm. Franklin, J. 2007. “Decomposing the Distributional Effects of Roadway Tolls.” Presented at the 86th Annual Meeting of the Transportation Research Board, Washington, D.C.

Using travel Demand Models for Environmental Justice Assessments 199 Horowitz, A., Creasey, T., Pendyala, R., and Chen, M. 2014. NCHRP Report 765: Analytical Travel Forecasting Approaches for Project-Level Planning and Design. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Huang, G., Hu, H., Cheng, H., and Aleksandr, B. 2013. “SCAG Activity-Based Model Development Workshop.” Retrieved from http://www.scag.ca.gov/Documents/MTF_-_SCAG_Activity-based_Model_Development_ Final.pdf. KYTC and Indiana DOT. 2015. Assessment of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations. Retrieved from http://kyinbridges.com/wp-content/uploads/ 2015-04-24_EJ-Assessment-Plan.pdf. Klein, N. 2013. “Spatial Methodology for Assessing Distribution of Transportation Project Impacts with Environmental Justice Framework.” Transportation Research Record: Journal of the Transportation Research Board, No. 2013. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 46–53. Kriger, D. 2006. NCHRP Synthesis 364: Estimating Toll Road Demand and Revenue. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Lin, D. Y., Eluru, N., Waller, S. and Bhat, C. 2008. “Integration of Activity-Based Modeling and Dynamic Traffic Assignment.” Transportation Research Record: Journal of the Transportation Research Board, No. 2076, Trans- portation Research Board of the National Academy of Sciences, Washington, D.C. pp. 52–61. Martens, K., and Hurvitz, E. 2009. “Distributive impacts of demand-based modeling.” Transportmetrica. Retrieved from http://www.tandfonline.com/doi/abs/10.1080/18128600903322333. McNally, M. 2007. The Four Step Model. Institute of Transportation Studies. Retrieved from http://www.its.uci. edu/its/publications/papers/CASA/UCI-ITS-AS-WP-07-2.pdf. McNally, M., and Rindt, C. 2007. “The Activity-Based Approach. Institute of Transportation Studies.” Retrieved from http://krypton.its.uci.edu/its/publications/papers/CASA/UCI-ITS-AS-WP-07-1.pdf. Metropolitan Transportation Commission. 2001. The 2001 Regional Transportation Plan Equity Analysis and Envi- ronmental Justice Report. San Francisco, California. Retrieved from http://www.mtc.ca.gov/library/2001_rtp/ downloads/EJ/EquityReport.pdf. Mid-Ohio Regional Planning Commission. 2012. 2012-2035 Metropolitan Transportation Plan—Appendix D: Environmental Justice Technical Analysis. Retrieved from http://morpc.org/pdf/2012MTPAppendixD_ EJAnalysis_May.pdf. NCTCOG. 2014. Regional Tolling Analysis for the Dallas-Fort Worth Metropolitan Planning Area Based on Mobility 2035–2013 Update. Retrieved from http://www.nctcog.org/trans/ej/documents/RTAJan2014.pdf. Parsons Brinckerhoff, Northwestern University, Mark Bradley Research & Consulting, University of California at Irvine, Resource Systems Group, University of Texas at Austin, Koppelman, F., and GEOSTATS. 2013. Improving Our Understanding of How Highway Congestion and Pricing Affect Travel Demand. SHRP Report S2-C04-RW-1. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Perez, B. G., Batac, T., and Vovsha, P. S. 2012. NCHRP Report 722: Assessing Highway Tolling and Pricing Options and Impacts, Volume 1: Decision-Making Framework. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Perez, B. G., Batac, T. and Vovsha, P. S. 2012. NCHRP Report 722: Assessing Highway Tolling and Pricing Options and Impacts, Volume 2: Travel Demand Forecasting Tools. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Rogoff, P. 2014. Revised Departmental Guidance on Valuation of Travel Time in Economic Analysis. U.S. Depart- ment of Transportation, Office of the Secretary of Transportation. Washington, D.C. Retrieved from https:// www.transportation.gov/sites/dot.gov/files/docs/USDOT%20VOT%20Guidance%202014.pdf. Sall, E., Bent, E., Koehler, J., Charlton, B., and Erhardt, G. D. 2010. “Evaluating Regional Pricing Strategies in San Francisco—Application of the SFCTA Activity-Based Regional Pricing Model.” Presented at the 89th Annual Meeting of the Transportation Research Board, Washington, D.C. San Francisco County Transportation Authority. 2010. San Francisco Mobility, Access, and Pricing Study. Retrieved from http://www.sfcta.org/sites/default/files/content/Planning/CongestionPricingFeasibility Study/PDFs/MAPS_study_final_lo_res.pdf. Shaw, T. NCHRP Synthesis 311: Performance Measures of Operational Effectiveness for Highway Segments and Systems. Transportation Research Board of the National Academy of Sciences, Washington, D.C. Southern California Association of Governments. 2012. Regional Transportation Plan 2012–2035. Retrieved from http://rtpscs.scag.ca.gov/Documents/2012/final/SR/2012fRTP_EnvironmentalJustice.pdf. Steer Davies Gleave. 2014. Louisville-Southern Indiana Ohio River Bridges, Traffic and Revenue Study: Final Report. Kentucky Public Transportation Infrastructure Authority. Retrieved from http://kyinbridges.com/ wp-content/uploads/Traffic-Revenue-Study-8-30-13.pdf. Tierney, K. 2012. NCHRP Synthesis 434: Use of the U.S. Census Bureau’s Public Use Microdata Sample (PUMS) by State Departments of Transportation and Metropolitan Planning Organizations. Transportation Research Board of the National Academy of Sciences, Washington, D.C.

200 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox U.S. Census Bureau. 2009. A Compass for Understanding and Using American Community Survey Data: What Researchers Need to Know. U.S. Government Printing Office. Washington, D.C. Retrieved from http://www.census.gov/content/dam/Census/library/publications/2009/acs/ACSResearch.pdf. U.S. Census Bureau. 2015. American Community Survey: About PUMS. Retrieved from http://www.census.gov/ programs-surveys/acs/technical-documentation/pums/about.html. U.S. DOT. June 2012. Revised Record of Decision. Louisville-Southern Indiana Ohio River Bridges. Jefferson County, Kentucky and Clark County, Indiana. FHWA-KY-SEIS-12-01-F. Retrieved from http://updates. kyinbridges.com/wp-content/uploads/2012/06/LSIORB-Revised-ROD-w-signature-6-20-2012.pdf.

201 What Is It? A select link analysis is a travel demand procedure that shows where trips that traverse a selected link start and end. The beginning and end points of the trips are displayed in various formats, including volumes and percentages. The displays can be shown by bandwidth, color, or numerically. A typical select link analysis follows the steps as listed below: 1. Open up a previous model run. 2. Open a loaded network (with traffic assignments). 3. Click on Select Link Procedure (see technical documentation). 4. Select the link or nodes of interest for a select link analysis. 5. Run the traffic assignment model with the selected link or nodes. 6. Select the output format for the select link analysis. Why Is It Effective in Environmental Justice Analysis? Select link analysis is effective for an EJ analysis because it allows the user to apply available output of TDMs to assess and compare the impacts that toll facilities or similar types of transportation improve- ments will have on travel time and accessibility measures. Once the zones that contain high concentrations of low-income or minority populations are identified, select link analyses can be set up and quickly analyzed to compare scenarios with and without toll facili- ties in the network. What Are Some Techniques for Implementing This Tool? The select link procedure is available through the application of travel demand modeling software packages, including TransCAD, Cube/Voyage, EMME2, VISUM, or similar software packages. The specific steps for applying a select link analysis are well documented in their respective technical manuals and documentation. Every urbanized area with a population of more than 50,000 has an MPO with an ongoing planning process, which will typically include a TDM. Application of the select link analysis T O O L 8 Applying a Select Link Analysis to Assess Trip Patterns Framework Step • Conduct Impact Analysis and Measurement • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Travel demand modeling • Statistical analysis Affected Populations • Low-Income • Minority Examples Featured • Georgia DOT, Atlanta Region, Northwest Corridor Project • Alamo Area Metropolitan Planning Organization, Regional Toll and Managed Lane Analysis

202 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox can be accomplished through the use of the existing TDMs that have been set up for the urbanized areas. Many states have also set up statewide TDMs. Depending on the location and extent of the toll facility being evaluated, a statewide TDM may be more suitable than a model that is limited to an urbanized area. A variation of the select analysis can be used for an EJ analysis. One variation would be to track trips through a node (i.e., pinpointing a location of interest) rather than through a link (i.e., focusing on the travel path). Trip movements are tracked through a node in a similar fashion as they are in a select link analysis. A second variation would be to track trips from zonal centroids with low-income and minor- ity populations. Travel times from the designated zones could be calculated to the destination zones, and skim trees would be created. Skim tree travel times for networks with and without the toll facility could be compared to determine the effect of the toll roads on specific zones with high low-income and minority populations. What Are Its Limitations? There are several limitations to the select link analysis procedure. In its application for deter- mining the effect of toll facilities on EJ, the select link analysis is based on aggregate impact. Select link determines how many trips go from Zone A to Zone B, and individual trips are not identified on an individual basis according to their socioeconomic characteristics. Rather the trips are treated as an aggregate movement, and the individuals that are making the trips are identified by the aggregate characteristic of the zones they are originating from and their zones of destination. For EJ analysis, the select link analysis may identify those zones with characteristics of high low-income and minority populations. Although not exclusively made up of populations that have EJ characteristics, the evaluation of these zones may capture the main impact on the low- income or minority members of the community. A second major limitation is the complexity of running a select link analysis. Knowledge of travel demand modeling and the microcomputer travel demand programs that are used in the planning process are essential for the use of a select link analysis. What Types of Resources and Costs Are Required? In order to use select link analysis, the urbanized area or region in which the toll facility is located must have an existing microcomputer TDM. However, most urbanized areas and many states have regional demand models. For the application of the select link procedure to be practical, existing microcomputer models must be used. To initiate a regional microcomputer TDM “from scratch” would require a large investment of time and cost. However, most major toll projects use demand models to forecast both traffic and revenue. Pivoting off of these models, a select link analysis would be a practical application of the models used for traffic and revenue forecasting. If a select link analysis is used in coordination with the existing model, a modest investment of time and cost would be needed. However, the user must identify the zones where EJ populations are concentrated and must use experienced modelers who know how to use the microcomputer

Applying a Select link Analysis to Assess trip Patterns 203 demand modeling software packages that are available and used in the area where the toll project is located. Who Has Used It Successfully? Select link analysis has existed in the planning community and demand modeling software packages for many years and is frequently used for scenario analysis procedures, including cali- bration of the models. Agencies that have employed select link analysis for tolling evaluations include the Metropolitan Transportation Commission (MTC) in San Francisco (I-680 Corri- dor Project, 2013, cited in Step 4); Georgia DOT, Northwest Corridor Project (2013) and I-75 Express Lanes Project (2013); and the Alamo Area MPO, Regional Toll Analysis (2011) and Regional Toll and Managed Lanes Analysis (2014). Two examples of its use in the evaluation of toll roads for their impact on EJ are briefly described below. Example 1: Northwest Corridor Project, Atlanta Region. A series of transportation improvements to I-75 and I-575 were proposed in the Atlanta metropolitan area. The proposed improvements included managed lanes on the facilities. The proposed study area and managed lanes are shown in Figure 1, which provides an example of the 2018 morning peak-hour trip time from designated origins and destinations. The first step in the Atlanta process was to determine the distribution of low-income popu- lations. Household income information is available from the Atlanta TDM on a zonal basis. The model network also included the managed lanes in the highway system. For the second step, a select link analysis was run for the managed lanes. A report and maps were produced that display the distribution of originating zones of managed lane users. The map of low- income populations was used in combination with the select link analysis to show the pro- pensity of EJ populations to use the tolled lanes. By applying the select link analysis, it was determined that one-third of the trips on the managed lanes came from low-income areas in 2018 as shown in Table 1. Figure 2 shows the percentage of households in 2018 with incomes of $20,000 or lower (far left), the percentage of households with incomes of $50,000 or lower (center), and the number of trips on the corridor highlighted for select link analysis (far right). However, the planning team notes that the TDM that was used does not separate trips by income level, and individual trips are not associated with a particular income level. As a result, although the model may be used to estimate whether trips originate or terminate in areas with a high percentage of low-income households, it is not certain that those trips were exclusively made by low-income drivers, and may require further analysis. In an effort to understand the relationship between income and trips in the managed lanes, the planning team conducted additional statistical analysis of the travel demand modeling data. This analysis included a comparison of the trip-making characteristics of these zones with their low-income profiles. Figure 3 presents a plot of each zone in the subject two-county subarea, with the percentage of households in each zone with annual incomes below $20,000 on the x-axis, and the number of trips generated in each zone that use the proposed managed lanes in a typical weekday under the baseline 2018 scenario on the y-axis. The planning team sought to identify whether areas with a larger percentage of low-income households (i.e., zones defined as having an annual income below $20,000) were less likely to use the managed lanes. If the two were negatively correlated, the points could be expected to cluster around a trend line similar to the downward sloping red line included in Figure 3. However, no such relationship was found using the travel demand modeling data, suggesting that areas with a

204 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Figure 1. Morning peak-hour trip time, Northwest Corridor project, 2018. Source: Georgia DOT, 2013.

Applying a Select link Analysis to Assess trip Patterns 205 Year Total Subarea TAZs Total TAZ Trips Trips from Low- income TAZs Trips from Low/Mid- income TAZs Percentage of Total Trips from Low-income TAZs Percentage of Total Trips from Low/Mid- income TAZs 2018 321 14,061 103 141 4,670 7,254 33% 52% 2035 342 35,166 112 136 13,947 17,634 40% 50% Source: Georgia DOT, 2013 Note: K = $1,000. Total Low- income TAZs (<$20,000) Total Low/Mid -income TAZs (<$50,000) Table 1. Trip comparison from select link analysis for managed lanes. Figure 2. Income distributions and daily traffic analysis zone volumes, 2018. Source: Georgia DOT, 2013. higher percentage of low-income households are not less likely to use the managed lanes. In fact, the analysis found a very low R2 value (0.0035), indicating that there was not a close fit between the independent and dependent variables on the x and y axes. Moreover, with a correlation coefficient of 0.0588 for these variables, there was little indication of closeness in fit for the data (values of 1 or –1 indicate perfect correlation). These results confirmed for the planning team that there does not appear to be a strong relationship between the percentage of low-income households and managed lane usage in the subject corridor. The planning team also examined a few other formulations of the data to see if they gave any indication of a relationship between the percentage of low-income households and managed lane usage. To control different TAZ sizes, the total number of trips generated in each TAZ over a typical weekday was examined in terms of the ratio of trips generated in the zone that use the managed lanes divided by total trips generated in the zone. As shown in Figure 4, the resulting values (shown in the y-axis here) also did not find a strong relationship between the percentage

206 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox of low-income households and managed lane usage as a percentage of the total trips (R2 value is 0.0065, and correlation coefficient is 0.0807) nor followed the hypothesized downward sloping trend line, as shown in Figure 3. The planning team also performed an analysis of two subgroups of the 348 TAZs in the 2-county area—those households in the 90th percentile and the 10th percentile of households with annual incomes below $20,000. The break points for the 90th and 10th percentiles of the data are 24.3 percent and 2.3 percent of TAZ households with annual incomes below $20,000, respec- tively. Again, within these two subgroups (each containing 37 households), there was no strong relationship between the percentage of low-income households and trips in the managed lanes as a percentage of total trips from the zones (see Figures 5 and 6). However, when compared to each other, the 90th and 10th percentile data exhibited some differences in the average number of man- aged lane trips as a fraction of the total trips from the zones. The 90th percentile subgroup (see Fig- ure 5) shows a higher average number of managed lane trips as a fraction of the total trips than does the 10th percentile subgroup (see Figure 6). This relationship was not evident at percentile levels across the spectrum (i.e., 80th percentile, 70th percentile), according to the planning team, bolster- ing their finding that there was no statistically significant relationship between the percentage of low-income households in a TAZ and managed lane usage, using the travel demand modeling data. Figure 3. Comparison of low-income profile and trip-making characteristics of traffic analysis zones. Source: Georgia DOT, 2013.

Applying a Select link Analysis to Assess trip Patterns 207 Figure 4. Alternative comparison of low-income profile and trip-making characteristics of TAZs. Source: Georgia DOT, 2013. In presenting these statistical analysis findings, the report makes no mention of estimating the potential effect that distance between the managed lane corridor access points and the zone to see what influence, if any, this variable might also have on managed lane usage. Example 2: Alamo Area Regional Toll and Managed Lane Analysis. The Alamo Area MPO, formerly the San Antonio–Bexar County MPO, undertook a study of toll facilities on several facilities included in the transportation network in 2014. The MPO currently does not have any toll or managed lane facilities within the study area, but anticipates that tolls and managed lanes will become a significant percentage of the non-local vehicle miles traveled by 2035. The MPO evaluated the use of tolls and managed lanes on a number of existing facilities, as shown in Figure 7. The Alamo Area MPO’s first step was to identify the location of both low-income and minor- ity populations to be included in the analysis of EJ populations. The populations were mapped at the TAZ into “EJ zones” and “non-EJ zones.” The second step in the process was to run a select link analysis to evaluate travel times for EJ populations using tolled and managed lanes. “Candidate trips” were identified as trips where the toll path offered a shorter travel time for households traveling from EJ zones. The analysis was conducted to determine whether the tolled and managed lane facilities would have a disproportionately adverse effect on the EJ populations, and whether the EJ populations could be expected to experience longer travel times as a result of the implementation of toll facilities. The analysis showed that travel times decreased for EJ zones when using tolled facili- ties in comparison to the free path lane, but would increase more if the tolled facilities were not implemented. Table 2 shows how the data was organized to compare EJ and non-EJ zones and indicates the number of trips that would save time as a result of using a new toll facility.

208 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Figure 5. Comparison of 90th percentile income and trips. Source: Georgia DOT, 2013. Figure 6. Comparison of 10th percentile income and trips. Source: Georgia DOT, 2013.

Figure 7. Proposed toll/managed lane system, San Antonio region. Source: San Antonio–Bexar County MPO, 2011. Congested Average Trip Length (CATL) In Minutes of Free Path and Tolled Path Option Under the 2035 Build and No-Build Networks Segmentation of 2035 Home-Based Work Person Trips by Potential Time Savings Number of 2035 Home- Based Work Person Trips Build Network Average Trip Length Using a Toll Path Build Network Average Trip Length Using a Free Path No-Build Network Average Trip Length Using a Toll Path (2) No-Build Network Average Trip Length Using a Free Path En vi ro nm en ta l J us ti ce Zo ne s (n = 3 64 ) Trips that can save 0+ minutes using a new toll facility (1) 55,593 28.92 31.94 n/a 36.11 Trips that cannot save 0+ minutes using a new toll facility 877,050 20.01 20.07 n/a 22.37 N on -E nv ir on m en ta l Ju sti ce Z on es (n = 2 87 ) Trips that can save 0+ minutes using a new toll facility (1) 138,330 31.91 34.89 n/a 40.87 Trips that cannot save 0+ minutes using a new toll facility 511,016 21.05 21.16 n/a 24.52 Source: San Antonio–Bexar County MPO, 2011 Notes: (1) Trips that can save time on tolled facilities are determined through a TransCad Selected Link analysis for all trips. (2) There are no toll facilities in the No-Build Network. Table 2. Environmental justice analysis with 2035 home-based work person trips, San Antonio–Bexar County region.

210 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Resources Georgia DOT. 2013. Technical Memorandum Update—Evaluation of Tolling Effects on Low Income Popu- lations: Northwest Corridor Project. Retrieved from http://www.nwcproject.com/media/pdfs/reeval/ evltllngefflip13.pdf. San Antonio-Bexar County MPO. 2011. Appendix F—San Antonio–Bexar County MPO Regional Toll Analy- sis. Retrieved from http://www.411on281.com/us281eis/assets/File/20%20-%20Appendix%20F%20-%20 SA-BC%20MPO%20Regional%20Toll%20Analysis.pdf.

211 What Is It? VOT and its measurement is a key factor and variable in trans- portation planning, project evaluation and asset assessment, and performance. The classical microeconomic theory of time allo- cation defines VOT as the opportunity cost of an additional unit of leisure (Concas and Kolpakov, 2009). Economists have used the prevailing average wage rate as a proxy for VOT, but research has found that VOT can be higher or lower than the current wage for an individual. VOT can be influenced by several factors, including trip purpose, trip timing, trip urgency, work schedule flexibility, and socio-demographic characteristics, among other factors. While surveys and empirical research have sought to account for the factors that affect VOT, they may at times be impractical or infeasible to administer. For this reason, the traditional shadow price measure has often been equated with the gross hourly wage rate plus fringe benefits. The value of travel time savings (VTTS) from travel time improvements has been based on fractions of the gross hourly wage rate for categories of travel (see Table 1). The economic theory behind the valuation of travel time, the rationale for this approach, and its limitations are described in the U.S. DOT’s revised The Value of Travel Time Savings: Departmental Guidance for Conducting Economic Evaluations (U.S. DOT, 2015). U.S. DOT’s recommended values and ranges for travel time savings do not account for other components of consumer sur- plus/willingness to pay such as travel comfort, safety, and reli- ability. These recommended values also do not differentiate by the income distribution range of travelers nor do they reflect the geographic diversity of travelers. Nonetheless, the Guidance offers analysts a benchmark against which empirical estimates of VOT, derived from the survey data on income and wages within a given region, can be comparatively assessed. For this reason, one approach taken has been to first quantify VOT and then apply median values across a range of income brackets within a corridor (Table 1 shows an example of this analysis). This refined step captures the entire population distribution by income class within a region to better reflect the variation in the value of travel time savings t o o l 9 Analyzing the Value of Time/Willingness to Pay in Environmental Justice Assessments Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Revealed and Stated Preference Surveys • Discrete Choice Modeling • Toll Optimization Modeling • Use of Value of Travel Time Savings to Inform Mitigation Policies for Low-Income Travelers Affected Populations • Low-Income • Minority Examples Featured • Florida DOT, I-95 Express Lanes • Los Angeles County Metropolitan Transpor- tation Authority, I-10 and 1-110 ExpressLanes • San Bernardino Associated Governments, I-10 and I-15 ExpressLanes

212 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox among travelers. These values can then be compared to specific tolls and toll regimes in effect or being proposed within various corridors. The value of reliability (VOR) is closely related to the concept of VTTS and the willingness to pay (WTP) for travel improvements. Researchers have found evidence of VOR above and beyond the value of travel time, indicating that risk-averse travelers would be willing to pay a monetary amount to decrease their travel time uncertainty. For example, researchers have found evidence of WTP for reliability above small revealed travel time savings (Burris et al., 2012), while a synthesis of VOT/VOR studies found that travelers value reliability as much as travel time savings (Concas and Kolpakov, 2009). Thus, even in areas with low prevailing wage rates, value pricing may be successful if the travelers highly valued the reliability of their travel time that could potentially be provided by tolled managed lanes (Concas and Kolpakov, 2009). VTTS is the main category of social benefits for transportation investments for ex ante project evaluations. Compared to accident cost savings and vehicle operational and ownership cost savings, travel time benefits are typically 60 to 80 percent of the total social benefits estimated for these projects (Concas and Kolpakov, 2009). VOT is applied in TDMs for use in predicting mode choice and lane usage in tolled managed lane projects (see tool, “Using Travel Demand Models for EJ Assessments”). VOT is a key variable in assessing congestion pricing schemes. VTTS is compared between the general purpose lanes and tolled managed lanes. The comparison of VOT to toll regimes used in dynamic pricing schemes is crucial in assessing the viability and socioeconomic impact of these projects and in conducting an EJ assessment of effects on low-income and other disadvantaged populations. Category Per Person-Hour as a Percentage of Total Earnings Recommended Hourly Earnings Rates for Determining Values of Travel Time Savings (2013 U.S. per person- hour) Recommended Hourly Values of Travel Time Savings (2013 U.S. per person-hour) Local Travel Personal Business All Purposes ** 50% (35%-60%)* 100% (80%-120%) $25.00 $24.40 $12.50 ($8.70-$15.00) $24.40 ($19.50-$29.30) Intercity Travel Personal Business All Purposes ** 70% (60%-90%) 100% (80%-120%) $25.00 $24.40 $17.50 $24.40 ($19.50-$29.30) Source: U.S. DOT, 2015 Notes: *Plausible ranges shown in parentheses. **Weighted averages, using distributions of travel by trip purpose on various modes. Distribution for local travel by surface modes: 95.4% personal, 4.6% business. Distribution for intercity travel by surface modes: 78.6% personal, 21.4% business. Table 1. Recommended values of travel time savings and plausible ranges for surface transportation (per person-hour as a percentage of total earnings and $ per person-hour).

Analyzing the Value of time/Willingness to Pay in Environmental Justice Assessments 213 Why Is It Effective in Environmental Justice Analysis? VOT and its accurate measurement can inform an assessment of EJ effects through its close consideration of the factors that influence potential differences in travel behavior. For EJ, VOT by income distribution and trip purposes are important variables to measure and estimate within a region or corridor. VOT can be used to inform the development of toll pricing and account policies that are sensitive to the behavioral differences of low-income travelers. The decision of managed lane users to value their travel time savings more than the cost of a particular toll at a particular time of day or day of the week can be assessed for its income equity effects. The decision itself exemplifies how important it is to accurately measure VOT/VTTS across a broad spectrum of income groups if the potential income equity impacts for a given system or project are to be assessed. Are some low-income travelers paying the toll because they have no other alternatives? Are the tolls at all times of the day higher than their individual values of travel time? Does the particular tolled managed lane waive tolls for those who are carpooling? Or do most low-income drivers travel alone? Does the availability of the managed lanes optimize travel flows in ways that will benefit those who elect to take the “free” general purpose lanes? These pricing and time-related questions necessitate a detailed travel time valuation and road user sur- vey analysis covering all groups of travelers using the system, or those who may potentially use it in the future, to analyze the equity dimensions. A key step in understanding the impacts on low-income users is to first evaluate the distribu- tion of general purpose and managed lane users by income class. Since VOT is based on fractions of hourly gross wage rates per each income or wage class, this step is crucial in assessing income equity. Analyzing travelers within a region or corridor according to the distribution of income is a first step in measuring potential travel impacts. Analysts can obtain relevant income data from the American Community Survey, detailed socioeconomic information estimated at the TAZ level, and telephone or online surveys of travelers drawn from license plate or other intercept methods, among other approaches. Some of this information can be challenging to obtain, but is essential for assessing the income equity impacts of congestion pricing schemes. When assessing potential EJ impacts using VOT and reliability per income category, it is impor- tant to evaluate the effects of the managed lane implementation comprehensively. The operation of the tolled managed lane can yield benefits in terms of the VTTS for all income categories for both the tolled managed lane and the general purpose lane users in some cases, and this phenom- enon should be considered to ensure a balanced appraisal of potential EJ impacts attributable to the HOT lane conversion or construction. Understanding the nature of congestion pricing impacts per income class and their VTTS incidence should inform how project sponsors design toll policies and managed lane mitiga- tion strategies to address potential travel behavior differences by income. For example, one proposed toll design scheme calls for developing pricing mechanisms that consider income equity and are deployable using current infrastructure and technology for HOT lane facilities (Paleti and Peeta, 2014). This study integrates revenue maximization and equity principles into an optimal toll determination problem. A key feature of the proposed methodology is a multi- toll pricing scheme that is consistent with current technology and infrastructure and factors income equity and level-of-service constraints. What Are Some Techniques for Implementing This Tool? Empirical methods for the valuation of travel time are based on trip characteristics and user socioeconomic data obtained from both stated preference and revealed preference surveys. Stated preference surveys ask individuals to respond to a set of hypothetical questions or scenarios. The

214 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox respondents may have no experience or may not have encountered the subject matter of the ques- tion. Stated preference surveys are particularly useful in making an estimate of travel behavior prior to implementation of a project or service and can encourage an exploration of responses to tolls that may not exist in the transportation system. Revealed preference data usually reflects actual trip behavior on a particular corridor. Stated preference asks how one would behave, while revealed preference elicits how a traveler actually behaved. Once the survey data is compiled, a statistical model is then applied or fit to this data set to obtain conclusions about VTTS and VOT for the given corridor. The model usually applied is a discrete choice or logit model that casts the binary decision of whether the trip was taken in the HOT lane (and paid a toll) or was taken in the general purpose lane (and did not pay a toll). VTTS can be defined as differences in trip times between the GP and HOT lanes (see accompanying text box, Discrete Choice Model Computation of Value of Travel Time Savings). Discrete Choice Model Computation of Value of Travel Time Savings For the I-95 Express Lanes study, two utility functions (Equation 1) were esti- mated from data collected in an online revealed preference survey. These func- tions represent the utility, or satisfaction, to the traveler from choosing to drive in the express lanes and pay the toll, or not, based on the option that would yield the highest utility. U X Xij i i j j i j (Equation 1), , ,= β + β + ε Where: Uij = utility of alternative i to traveler j i = set of alternatives available to traveler j Xi = vector of measurable attributes of the alternatives (e.g., toll, travel time savings) bi = vector of the coefficients of Xi Xj = vector of measurable attributes of the traveler (e.g., income, education) bj = vector of the coefficients of Xj ei,j = unobservable factors The value of travel time savings (VTTS) or value of time (VOT) can be derived as the marginal effect of this analysis. It is estimated (Equation 2) by comparing the disutility of higher travel time with the disutility of a toll. With a discrete choice model, VTTS can be computed as the ratio of the coefficients of travel time sav- ings (TTS) and the toll cost (toll) obtained from the discrete choice model results. (Equation 2), , , , VOT VTTS U TTS U toll i j i j i j i i j i TTSi tolli = = ∂ ∂ ∂ ∂ = β β Source: Perk et al., 2011 Example 1: I-95 Express Lanes, Florida. Researchers designed and implemented a revealed preference survey for the I-95 express corridor in Miami-Dade County (Perk et al., 2011). The online survey was distributed via email, drawing on databases maintained by South Florida

Analyzing the Value of time/Willingness to Pay in Environmental Justice Assessments 215 Commuter Services and Florida DOT District 6. With no data of individual trips along the I-95 corridor, the online survey was designed to collect information on individual trips made on the general purpose and express lanes. Survey participants were requested to provide information on two recent trips on the I-95 and respond to attitudinal and demographic questions. Respondents were asked to identify the date of their trips and estimate their travel speeds and toll fees. These responses were then compared and validated against actual data on tolls and speeds collected by Florida DOT District 6. Speed and volume information for the general purpose and express lanes were recorded, using a microwave vehicle detector system (MVDS) that deploys a set of detectors along the corridors. This information was used to validate the reported speeds. Similarly, a toll chronology, which includes a history of actual toll levels, was used to validate the reported toll information. The actual toll and speed data was used to develop the researcher’s estimate of VTTS. The researchers were able to make an estimate of VTTS based on revealed preferences and the online survey instrument (see text box, Modeling the Value of Travel Time Savings). Travel time savings were defined as the difference in travel time, in minutes, to cover the 7.3 mile segment of the I-95 Express Lanes facility versus the general purpose lanes at a given point of time. The VTTS estimates covered a range of income classes for a bracket range spanning $2.27 to $79.32 per hour and a mean of $32 per hour. Figure 1 presents a frequency distribution of VTTS, which indicates that the highest percentage of survey respondents (18.7%) have an estimated VTTS ranging from $45 to $60 per hour, based on annual household income. The shape of the VTTS distribution was found to be generally similar to the annual household income distribution of the survey respondents (not shown here) and in accordance with the general assumption that VTTS varies with the wage rate. The computed VTTS for these I-95 travelers was estimated to be about 49 percent of their estimated gross hourly wage (Perk et al., 2011). The researchers concluded that these VTTS estimates were generally consistent with recent transportation research that found VTTS in the range of 20 percent to 50 percent of the traveler’s hourly wage (and higher for revealed preference results). Modeling the Value of Travel Time Savings Toll pricing policies can be informed by accurate estimates of value of travel time savings (VTTS). For the I-95 Express Lanes project, the revealed preference survey sought information on both annual household income and annual personal income. The estimated model for annual household income performed better and was included in the final research findings. The equations for the integrated toll/ household income variable (“toll_householdinc”) and the VTTS are shown below: toll_householdinc toll annual household income 2000 (Equation 3)= VTTS annual household income 2000 60minutes hour (Equation 4)Time toll_householdinc = β β × × Where: VTTS = value of travel time savings ($/hour) bTime = coefficient of the variable time (represents travel time savings) btoll_householdinc = coefficient of the variable toll_householdinc (integration of toll and income) Source: Perk et al., 2011

216 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Example 2: Los Angeles ExpressLanes Program, I-10 and I-110, Los Angeles County. To assess the impact of the ExpressLanes program initiative on low-income travelers, the Los Angeles County Metropolitan Transportation Authority (Metro) consultant first examined the income distribution of travelers. Particular attention was given to the travel behavior of the low- income segment; estimated travel time savings for use of the managed lanes were then compared to the costs of operating a vehicle and the projected toll levels on the managed lanes to assess the attractiveness of this option for this low-income segment. Referencing the transportation research literature, the Metro consultant set the appropriate value of time at 60 percent of the wage rate for a low-income household (an estimated $35,000 annual income), or a $10.10/hour rate for operating a single-occupancy vehicle. For carpoolers the value of time was assumed to be 1.82 times that value for a total of $18.38/hour. These VOT estimates were then compared to vehicle marginal VOT as estimated by a toll opti- mization model (TOM). The TOM calculated vehicle marginal VOT for small segments of the roadway and weighted them based on vehicle miles traveled on each segment. Vehicle marginal VOT ranged from as low as $24.95 to as high as $67.78 for a single-occupancy vehicle. The analytical exercise revealed that there was no situation in which the ExpressLane would be desirable to a low- income single-occupancy vehicle at the $10.10/hour or even at an $18.38/hour VOT for carpoolers. Nonetheless, the managed lanes serve as a type of insurance policy offering mobility and access for all income groups. It has been argued that the “option value” may be relatively more important to lower-income groups in terms of access to jobs or childcare because these groups have less schedule flexibility in their travel time (Kuehn, 2009). This reliability can be critical for a traveler who must meet strict scheduling requirements, and it may not be correlated with VOT (Concas and Kolpakov, 2009). There are circumstances in which a low-income driver’s VOT might increase significantly, leading to a willingness to take the tolled ExpressLane (e.g., being on time for work, picking up a child from day care, emergency healthcare). In recognition of this toll payment choice, it was determined that it would take a “toll credit” and a much higher VOT—120 percent of the wage rate rather than 60 percent as in the base case example, perhaps suggestive of an urgent situation for the low-income driver—to find a price point in which the tolled managed lanes would be attractive. This analytical exercise, in part, informed Metro in 2.7% 6.7% 5.3% 12.0% 2.7% 16.0% 13.3% 16.0% 4.0% 18.7% 2.7% 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0% Fr eq ue nc y (% o f S am pl e) Value of Travel Time Savings ($/hour) Figure 1. VTTS distribution for survey respondents traveling on I-95, Miami–Dade.

Analyzing the Value of time/Willingness to Pay in Environmental Justice Assessments 217 developing its Low-income Assistance Plan and Transit Credit Rewards Programs as forms of mitigation. Example 3: I-10 and I-15 ExpressLanes, San Bernardino County. For the San Bernardino Associated Governments (SANBAG), an equity assessment was prepared for the I-10 and I-15 corridors in San Bernardino County (see Figure 2). The assessment concluded that the express lanes will have several benefits for low-income drivers. Notably, the travel demand modeling demonstrated that travel times in the general purpose lanes will be faster on both I-10 and I-15 if express lanes are implemented, compared with other project alternatives. Drivers using the free general purpose lanes would be better off if the express lanes were implemented than if they were not. The express lanes also would provide a new travel option for low-income drivers, which did not exist under the current road configuration. The analysis of potential toll levels indicated that there could be times when a low-income driver would also find the time savings from the express lanes attractive. What Are Its Limitations? To strengthen the analysis of VOT in EJ assessments of managed lanes projects, more research focus could be placed on assessing the actual behavior of the users of managed lanes that have been in operation for several years. More revealed preference types of studies of these projects Figure 2. Median household incomes by TAZ were mapped adjacent to the study corridor to assist in identifying income quintiles to support focused assessment of low-income populations and travelers.

218 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox with established user behavior histories (at various frequencies, seasons, and over complete macro- economic cycles) would provide a greater empirical base for drawing conclusions related to impacts on users by income class. As suggested by the I-95 Express Lanes example and some of the resources at the end of this section, some of these studies have already been undertaken. Toll facilities with operating histories can take advantage of some of the available online survey technology employed to gather traffic data and other relevant survey explanatory data char- acterizing EJ populations by region. These additional research efforts and case studies could inform the VTTS estimation procedures and establish a more robust dataset that might yield additional insights into VOT/VTTS, the assessment of impacts to low-income populations, and the potential effectiveness or need for mitigation strategies. What Types of Resources and Costs Are Required? Implementing surveys to collect data for VOT models and constructing robust VOT models themselves can be labor intensive and time-consuming. Some steps can be streamlined but may not yield the kind of detailed data that would provide the most reliable estimates of VOT/VTTS. Where possible, some data can be obtained from previous impact studies (environmental impact studies, environmental assessments) or corridor/road user surveys that have identified and assessed EJ communities that may overlap with project travel corridors. Obtaining EJ popu- lation and income characteristic information for these communities may possibly streamline the research efforts. In addition, established benchmarks for some values can be used to provide a frame of reference or ballpark estimate if resources are limited. Still, it is important to recognize that there is a growing body of data being amassed from electronic tolls and, ideally, this data should be made available to mine to improve the quality of research and understanding of the distributional impacts of toll pricing. Who Has Used It Successfully? The examples previously described and the resources section provide useful information on key definitions, analytical steps, and implications of the analysis of VOT and VTTS for assessing the travel behavior effects by income in support of a comprehensive EJ assessment. Resources Burris, M., Nelson, S., Kelly, P., Gupta, P. and Cho, Y. 2012. “Willingness to Pay for High-Occupancy Toll Lanes: Empirical Analysis from I-15 and I-394.” Transportation Research Record: Journal of the Transportation Research Board, No. 2297. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 47–55. Burris, M., and Xu, L. 2006. “Potential Single-Occupancy Vehicle Demand for High-Occupancy Vehicle Lanes.” Transportation Research Record: Journal of the Transportation Research Board, No. 1960. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 108–118. Concas, S. and Kolpakov, A. 2009. Synthesis of Research on Value of Time and Value of Reliability. Retrieved from http://www.nctr.usf.edu/wp-content/uploads/2012/05/77806.pdf. FHWA. 2008. Income-Based Equity Impacts of Congestion Pricing—A Primer. Retrieved from http://www.ops. fhwa.dot.gov/publications/fhwahop08040/fhwahop08040.pdf. FHWA. 2014. Guidebook for State, Regional, and Local Governments on Addressing Potential Equity Impacts of Road Pricing. Retrieved from http://www.ops.fhwa.dot.gov/publications/fhwahop13033/index.htm. Kuehn, D. 2009. “Environmental Justice and the Distribution of Benefits from Highway Pricing Program.” Presented at the 88th Annual Meeting of the Transportation Research Board, Washington, D.C. Metro. 2010. Metro ExpressLanes Project, Draft Final Low-Income Assessment, LA ExpressLanes Program. Retrieved from https://www.metroexpresslanes.net/en/about/low_income_draft_final_report.pdf.

Analyzing the Value of time/Willingness to Pay in Environmental Justice Assessments 219 Network Public Affairs. 2013. Equity Assessment Report for I-10 and I-15 in San Bernardino County. Retrieved from http://www.1015projects.com/files/managed/Document/119/86-406-sanbag-equity-assessment- report—final-nov-2013.pdf. Paleti, C. and Peeta, S. 2014. Design of Equitable Toll Prices for High-Occupancy-Toll Lanes. Presented at the 93rd Annual Meeting of the Transportation Research Board, Washington, D.C. Patil, S., Burris, M., Shaw, D., and Concas, S. 2011. “Variation in the Value of Travel Time Savings and Its Impact on the Benefits of Managed Lanes” in Transportation Planning and Technology. Vol. 34, pp. 547–567. Retrieved from http://agecon2.tamu.edu/people/faculty/shaw-douglass/TTP2011.pdf. Patterson, T. M., and Levinson, D. 2008. Lexus lanes or Corolla lanes? Spatial use and equity patterns on the I-394 MnPASS lanes. Unpublished manuscript, Department of Civil Engineering, University of Minnesota. Retrieved from http://nexus.umn.edu/Papers/MnPASSEquity.pdf. Perk, V. A., DeSalvo, J. S., Rodrigues, T. A., Versoza, N. M., and Bovino, S. C. 2011. Improving Value of Travel Time Savings Estimation for More Effective Transportation Project Evaluation. Retrieved from http://www. nctr.usf.edu/wp-content/uploads/2013/09/77921.pdf. Tilahun, N., and Levinson, D. M. 2006. “A Moment of Time: Reliability in Route Choice Using Stated Preference.” Presented at the 87th Annual Meeting of the Transportation Research Board, Washington D.C. Transportation Research Board. 2006. NCHRP Synthesis 364: Estimating Toll Road Demand and Revenue. Trans- portation Research Board of the National Academy of Sciences, Washington, D.C. U.S. DOT. 2015. The Value of Travel Time Savings: Departmental Guidance for Conducting Economic Evaluations–Revision 2. Retrieved from https://www.transportation.gov/sites/dot.gov/files/docs/Revised%20 Departmental%20Guidance%20on%20Valuation%20of%20Travel%20Time%20in%20Economic%20 Analysis.pdf.

220 What Is It? User cost assessments examine the economic or financial impact on households of various transportation options, including toll road pric- ing and their “burden” effects on household budgets. The approach estimates the cost of using the tolling facility for low-income and non- low-income households and compares how these costs may differ and how they may be experienced differently within and between various household income levels. Why Is It Effective in Environmental Justice Analysis? The frame of analysis considers possible regressive effects of travel alternatives (including but not limited to toll road pricing) for individual and/or groups of low-income households. Additionally, as transporta- tion costs rise, a potential risk of social exclusion exists for low-income households as a consequence of tolling. Social exclusion means that a transportation system by virtue “of its price, areas of service, or vehicle category is perceived to be biased against a social class, poor neighbor- hood or financial reach of a percentage of the population” (Madi et al., 2013). The outcome of social exclusion is that affected individuals or groups are precluded from participating fully in the economic, social, and civic activities of the society in which they live. Social exclusion warrants close attention in regional systems planning and project facil- ity environmental studies. What Are Some Techniques for Implementing This Tool? User cost analyses require an estimate of the costs experienced by a traveling household on a daily or annual basis, an estimate of the share of their income spent on these charges, and the extent to which consumption of other goods and services might be affected by the loss of this income. In some cases, this analysis will incorporate a valuation of the time impacts (i.e., time savings or increase in travel time) resulting from the road pricing option or from a resulting shift to public transportation or other longer routes to avoid pricing. Assessing the income equity of a road pricing alternative requires analysis of the economic impacts for households of varying t o o l 1 0 Assessing User Costs and Household Burden Effects Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act Tools and Techniques • User Cost Models • Travel Demand Model Affected Populations • Low-Income • Minority Examples Featured • Texas DOT, U.S. 281 Environmental Impact Statement • Kentucky Transportation Cabinet and Indiana DOT, Ohio River Bridges • Washington State Transportation Center, Puget Sound Metropolitan Region

Assessing User Costs and Household Burden Effects 221 income levels. For example, the economic and time impacts borne by low-income households would be compared with those experienced by middle- and high-income households. In addition to exploring the “vertical equity” impacts, a fair and comprehensive analysis would also consider the “horizontal equity” impacts (i.e., the differences between households at the same income level). Because people from low-income households may exhibit significantly distinct travel behaviors, it is reasonable to consider how several contextual factors may influ- ence the impacts that are borne, such as car ownership, residential or work location, the quality of available alternative transportation modal options (e.g., transit, walking, or carpooling), and work schedule flexibility. There are different ways to estimate user costs and their effects on the household budget in the context of evaluation of toll pricing. Some examples are presented below. Example 1: U.S. 281 Corridor EIS, Texas Department of Transportation. The U.S. 281 Cor- ridor EIS evaluates the impacts of improvements along an 8-mile corridor. The EIS draws on regional network level analyses prepared by The Alamo Area Metropolitan Planning Organi- zation (formerly the San Antonio–Bexar County Metropolitan Planning Organization) in its Regional Toll and Managed Lane Analysis that analyzed the effects of the existing and planned network of toll and managed lanes on EJ populations. Drawing on data from this analysis, the U.S. 281 Corridor EIS assessed the effects of the toll component on EJ populations at the project level. In so doing, it addressed a set of considerations developed jointly by the state of Texas and the FHWA. In 2009, FHWA and Texas DOT prepared its Joint Guidance for Project and Network Level Environmental Justice, Regional Network Land Use, and Air Quality Analyses for Toll Roads (April 23, 2009), which set forward appropriate considerations for system-level planning using network analysis as part of the metropolitan planning process and for project facility studies. The guidance makes clear that projects with tolling components should be evaluated to deter- mine anticipated effects on EJ populations within the region, including the impacts to travel time and/or out-of-pocket costs. An approach to illustrating the potential economic impacts from out-of-pocket costs for individuals using toll facilities is described as follows: One method to accomplish this is to multiply the anticipated toll cost to use the proposed toll facility by an estimate of 500 trips per year (i.e., 250 round trips to work per year). This can be put into context by discussing what percentage of household income this cost represents for a household at the poverty level vs. a household at the median household income. If variable toll rates are used, an analysis at the high, low, and mid-range toll rates should be provided (FHWA and Texas DOT, 2009). Following this approach, the cost analysis refers to the Alamo Regional Mobility Authority’s Amended and Restated Policies and Procedures for Toll Collection Operations on the Alamo RMA Turnpike System. The analysis examines a range of tolling price scenarios from $0.17 per mile to $0.50 per mile and, converting these costs into a yearly basis, examines the burden tolling will have on those living at the median household income and at the poverty line. Table 1 shows their resulting toll cost burden estimate as a percentage of household income and at the poverty line. The analysis confirms that low-income EJ populations are likely to spend a greater portion of their household income on tolls and, depending on price, this toll burden could reach 10 percent for a person living in poverty who needs to travel regularly by auto at the highest toll rate (i.e., $.0.50 per toll mile). Ultimately, this adverse burden effect received little weight in the finding determination related to EJ impacts described in the EIS because Texas DOT was able to demonstrate that a well- functioning alternative, a non-toll frontage road, was available for travelers along the alignment. Through its transportation modeling, the EIS examined several travel time measures, including

222 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox average time of trip, time savings per trip compared to the no-build alternative (in minutes), and total time savings (in hours) and compared trips originating from EJ zones with all other trips in the future condition. For those unwilling or unable to afford the toll road option, the “at-grade non-toll frontage road option” offered improved travel times in comparison to the no-build alternative. This option, it was argued, would provide a benefit to EJ and non-EJ communities. Example 2: Louisville–Southern Indiana Ohio River Bridges Project, Kentucky Transpor- tation Cabinet and Indiana DOT. For its Supplemental Final EIS, KYTC and Indiana DOT prepared an average user cost analysis for a proposed interstate bridge river crossing along the Ohio River. In contrast to the San Antonio region example, this user cost method focused on vehicle user costs, which include the cost of the toll but also non-toll costs such as vehicle operat- ing costs and the cost of time in traffic. Output from a “time-of-day” travel demand forecasting model developed for the project provided the basis for measuring the value of time. Table 2 reproduces a table from the Supplemental Final EIS that identifies the average cost per trip in 2030. Trips were segmented by three categories—intrastate, trips across bridges, and bridge and non-bridge—to reflect the range of trips taken regionally and to differentiate the costs of trips originating from “EJ Community Cars” (i.e., car trips that originated from an area identified as an EJ community) and “Non-EJ Community Cars” (i.e., car trips that originated from outside EJ communities). The Supplemental Final EIS found that the costs of the tolls would not be “predominantly borne by EJ populations” citing the fact that the average cost increase for all trips (both intrastate and cross-river) would be “minimal” (approximately 3%) and that the average user cost for EJ community cars would still be significantly less (almost $2.00 or 24 percent less) than the average user cost for all non-EJ community cars. However, the Supplemental Final EIS also acknowl- edged that the increase in average user costs would be appreciably greater for cross-river trav- elers originating in EJ areas than for those originating trips outside those areas (non-EJ zones): average user costs per trip for EJ community members in their cross-river travel would rise by 21 percent (rising from $6.75 to $8.16 per trip) while non-EJ community members’ cross-river travel user costs would be 11 percent (rising from $9.15 to $10.13) after toll implementation. FHWA found the disparity in the increase in average user costs to be a sufficient basis for finding in its Revised Record of Decision that EJ populations were likely to experience a Total Cost Per Mile Daily Round Trip Cost Yearly Cost 1 Percent of Median Household Income 2 Bexar County ($49,141) Comal County ($63,480) Poverty Line ($19,790) $0.17 $2.72 $680 1.4% 1.1% 3.4% $0.32 $5.12 $1,280 2.6% 2.0% 6.5% $0.50 $8.00 $2,000 4.1% 3.2% 10.1% Source: Texas DOT and Alamo Regional Mobility Authority, 2015 Notes: 1Assumes use of the tolled/managed lane along the entire U.S. 281 corridor, at a frequency of 5 round trips per week for 50 weeks of the year; 2U.S. Census Bureau, 2008-2012 American Community Survey, Median Household Income in the Past 12 months. Table 1. Estimate of toll costs—U.S. 281 corridor, final EIS example.

Assessing User Costs and Household Burden Effects 223 disproportionately high and adverse effect as a result of the toll implementation. Although not predominantly borne by the EJ population, the consequences of toll implementation would be “appreciably more severe or greater in magnitude for EJ populations” (U.S. DOT, 2012). The burden impact of tolls on low-income households and in terms of statewide per capita income was also calculated in the Supplemental Final EIS. The burden impact of a $1.00 toll each way was reported with key assumptions, noting that for persons earning at the poverty line level, the toll would account for 4 percent of their total income and about 2 percent for persons earning at the statewide per capita income. Table 3 presents this computation and also compares Average Cost Per Trip EJ Community Cars Non-EJ Community Cars External Cars Trucks Regional Average No Action Alternative Intrastate Trips (Non-Bridge) $3.48 $4.69 $10.42 $18.34 $6.05 Trips Across Bridges $6.75 $9.15 $14.87 $46.45 $16.55 Bridge & Non-Bridge $3.68 $4.89 $11.53 $22.50 $6.87 Final EIS Selected Alternative Intrastate Trips (Non-Bridge) $3.43 $4.51 $10.28 $18.12 $5.88 Trips Across Bridges $6.40 $8.36 $14.11 $44.48 $14.92 Bridge & Non-Bridge $3.65 $4.71 $11.29 $22.03 $6.67 Modified Selected Alternative With Tolls Intrastate Trips (Non-Bridge) $3.44 $4.55 $10.32 $18.17 $5.92 Trips Across Bridges $8.16 $10.13 $15.82 $49.65 $17.13 Bridge & Non-Bridge $3.79 $4.84 $11.77 $22.84 $6.90 Source: KYTC and Indiana DOT, 2015 Table 2. 2030 average cost per trip—Ohio River Bridge Final EIS example. One-Way Toll Daily Round Trip Cost Yearly Cost 1 Percent of Household Income 2 Year 2013 Poverty Line ($11,490) Indiana 2013 Per Capita Income ($24,796) Kentucky 2013 Per Capita Income ($23,668) $1.00 $2.00 $480 4.2% 1.9% 2.0% $1.50 $3.00 $720 6.3% 2.9% 3.0% $2.00 $4.00 $960 8.4% 3.9% 4.1% $5.00 $10.00 $2,400 20.9% 9.7% 10.1% $10.00 $20.00 $4,800 41.8% 19.4% 20.3% Source: KYTC and Indiana DOT, 2015 Notes: 1Assumes use of the bridge at a frequency of 5 round trips per week, 4 weeks a month for 12 months; 2U.S. Census Bureau, American Community Survey, Median Household Income in the Past 12 months, 2014. Table 3. Estimate of burden impact of low-income and per capita income under toll rate scenarios.

224 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox the sensitivity of the burden effects at various higher toll rates considered as part of the project’s traffic and revenue financial analyses. The financial burden sensitivity effects are shown here but were not presented in the EJ section of the Supplemental Final EIS. Regardless of the toll level, as expected, a low-income person who uses the bridges for a daily commute would devote a greater percentage of their annual income for tolls than a person from the non-low-income population, and these rates can rise to levels that make taking the trip unsustainable for a household. In considering the burden of this effect, it is important to consider to what extent low-income vehicle users may actually use the bridge crossing on a daily basis. In this study, based on user data obtained from a telephone survey, the Ohio River Bridge Users Study found that some 36 per- cent of low-income populations and 57 percent of minority populations were expected to cross the Ohio River by car every weekday or several times per week. Example 3: Puget Sound Metropolitan Region, Washington State. Drawing from a study for the Washington State DOT, researchers from the Washington State Transportation Center considered, on a regional basis, the potential economic cost of tolling for low-income and non- low-income households. They found that most poor households would not be substantially affected by tolling. However, individuals who do not have better alternative routes will experi- ence a significant decrease in their economic well-being. For this analysis, the authors use the Household Activity Survey (HAS) for the Puget Sound region, a regional activity-diary survey that provides information on household travel behav- ior. It contains basic demographic information and the exact locations of home and work for respondents. The researchers made estimates of journey-to-work routing, utilizing ArcMap route-finding algorithms that used geocoded data from the HAS dataset over a regional roadway network. Their efforts highlight the degree to which low-income travelers are commuters on key segments considered for toll implementation (see Table 4). Findings from this GIS method suggested that if all 12 segments were tolled, 31 percent of low-income commuters would experience an increase in out-of-pocket expenditures. Notably, more than two-thirds of low-income commuters use routes that do not involve any of the con- templated toll segments, and 22 percent use only one or two segments. The analysis finds that very few low-income commuters use the two bridges (segments 3 and 6) or segment 12, thus tolls on the two bridges would affect less than 3 percent of the region’s low-income commuters. The authors make estimates of low-income and non-low-income households’ use of poten- tially tolled segments to estimate the potential annual cost of tolls for both groups and assess whether tolls impose a disproportionate burden on low-income households (see Table 5). The researchers emphasize the importance of a geographic-specific approach when estimat- ing the distributional impacts of highway tolls. In their case example, it is clearly evident that a segment of low-income users would be significantly financially burdened in the absence of a non-toll alternative (e.g., bridge crossing). Nonetheless, researchers argued for assessing the tolling decision and its distributional effects on the broader regional basis—not just for toll users—to avoid distorting the policy questions before decision makers. It was necessary to “move beyond the conventional focus on the users of tolled facilities to analyze a more inclusive set of households” (Plotnick et al., 2011). They offered several arguments. By looking at all commuting households, or all households, rather than solely the toll facility users, researchers found that the degree of toll regressivity diminishes. The geographic distribu- tion analysis illustrates that tolls are not borne equally by all low-income households or, for that matter, by all non-low-income households. In fact, the costs differ by whether households drive,

Assessing User Costs and Household Burden Effects 225 Highway Segment Percent of Segment Users Who are Low- Income Percent of Low-Income Commuters Who Use Segment Percent of Non-Low- Income Commuters Who Use Segment 1. I-5 north from SR 520 to I-405 (serves Seattle, northern suburbs) 8.2 8.5 14.4 2. I-405 north from SR 520 to I-5 (serves Bellevue, Redmond, other eastern suburbs) 9.1 6.1 9.1 3. SR 520 bridge across Lake Washington 3.2 1.0 4.5 4. I-5 between SR 520 and I-90 (serves Seattle) 8.2 9.3 15.8 5. I-405 between SR 520 to I-90 (serves Bellevue, Redmond, other eastern suburbs) 8.0 5.0 8.7 6. I-90 bridge across Lake Washington 4.6 1.3 4.0 7. I-5 south from I-90 to I-405 (serves Seattle, southern suburbs) 7.1 5.1 10.0 8. I-405 south from I-90 to I-5 (serves eastern suburbs) 10.9 7.5 9.2 9. I-5 south from I-405 to King County line (connects Seattle and Tacoma) 12.3 5.6 6.1 10. SR 167 south of I-405 junction (connects Seattle and Tacoma) 13.3 7.9 7.8 11. SR 999 from W. Seattle bridge to tunnel (major Seattle artery) 21.9 2.1 1.1 12. I-90 east of I-405 (connects Puget Sound region to eastern WA) 3.7 1.2 4.7 All Segments 9.2 31.0 46.2 Source: Plotnick et al., 2011 Table 4. Use of potentially tolled highway segments by low-income and non-low-income commuters. Low-Income Households Non-Low-Income Households Annual Cost of Tolls (1) Percent of Income (2) Annual Cost of Tolls (1) Percent of Income (2) Full System Tolling—$1/Segment All households $235 1.50% $600 0.80% Commuting households $403 2.60% $710 0.90% Segment commuters $994 6.40% $1,334 1.70% SR 520 Bridge—$2 One-Way Toll All households $6 0.04% $6 0.05% Commuting households $10 0.06% $10 0.06% Segment commuters $31 0.20% $31 0.12% SR 520 Commuters $960 6.20% $960 1.30% Source: Plotnick et al., 2011 Notes: (1) Assumes 240 work days per year; (2) Uses incomes of $15,600 and $76,350, the respective median among low- income and non-low-income households. Table 5. Hypothetical annual toll burdens for low-income and non-low-income households.

226 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox whether they drive for work, and by the specific roads or facilities that they use. The approach points to differences within income groups as well as between groups. By framing the distributional analysis to the regional scale, rather than just to the effects on toll users, researchers argued that it accords “with standard practice in distributional studies of taxes and income support programs and would offer more insight into how highway tolls may affect equity among all residents of the region” (Plotnick et al., 2011). In this case, the research found that relying on the toll revenues for financing the bridge would be distributionally neutral on a regional basis. This would be preferable to using sales or fuel taxes for financing transpor- tation, which research suggests (e.g., Schweitzer and Taylor 2008) is more regressive. In this light, even if tolling is regressive, and even if it is burdensome to low-income households, the appropriate question is whether tolling or road pricing will burden the poor more than other ways of paying for roads. The financial burden for a small segment of low-income users may cause adverse effects but the goals of congestion reduction (and improving air quality indirectly) are sufficient, accord- ing to researchers, to warrant the imposition of tolling as the most appropriate means for cap- turing the true costs of sole occupancy driving to society. Consistent with their distributional framework, researchers turned briefly, if not satisfactorily, to whether there were alternatives for reducing congestion that would prove less burdensome for poor users of the prospective toll facilities. Researchers considered the role of public transportation in addressing conges- tion, finding it somewhat promising, but cautioned that public transportation was only partially funded through fares and had to be subsidized. The current methods of funding (retail sales, gasoline or property taxes) will also bring additional costs to poor families, including those that will not use the services. What Are Its Limitations? These examples suggest that the average low-income household, if dependent on the use of a toll facility to reach an essential destination, is likely to pay a greater percentage of their annual income on tolls than the average non-poor household. But does this disparity in user costs consti- tute disproportionately high and adverse effects for low-income and minority populations? And, if it does, are there practicable alternatives to these impacts or practicable mitigation measures? Pursuant to the FHWA Order 6640.23A, adverse effects must be predominantly borne by low- income or minority populations or the adverse effect must be appreciably more severe or greater in magnitude than the adverse effect suffered by non-low-income or non-minority populations. The examples suggest that this assessment requires weighing several contextual factors that influence the mobility and travel options for low-income households. These factors include: the location of employment and other essential destinations for accessing opportunities rela- tive to the home for the advancement of the individual or family; the feasibility of substituting modal alternatives (e.g., public transportation, carpooling) or taking non-tolled routes to reach essential destinations; the anticipated pre-and post-toll travel patterns (e.g., travel times and user costs); toll collection mechanisms (e.g., necessity of having a transponder account and replenishment mechanisms); and how the net toll revenues are likely to be recycled (e.g., public transportation, toll discount programs), among other considerations. The Texas and Ohio River Bridges examples describe analyses that draw on travel demand models in formulating user cost estimates. In both cases, the analyses are founded on identifying the zone of origin as a place that is predominantly composed of EJ populations or non-EJ popu- lations. The models, as designed, cannot distinguish whether the trip itself is, in fact, taken by a low-income person. The income equity analysis would be improved if income segmentation were preserved in the trip generation, trip distribution, and mode choice models.

Assessing User Costs and Household Burden Effects 227 The Washington State example relies on a GIS method that is straightforward for estimating the cost of using tolled facilities. However, the approach does not incorporate routines embedded in more sophisticated travel demand models, including potential route deviations to avoid tolls, mode switching carpools or public transit, and the effects of VOT considerations on travel behavior choices. Despite these limitations, the focus of the user cost and household burden analysis properly recognizes the need to assess the distributional effects of toll pricing as part of an EJ analysis when considering toll implementation or rate changes. What Types of Resources and Costs Are Required? The user cost methods described here are generally not very expensive and can be undertaken as policy research for statewide and metropolitan planning activities or for project development during the environmental review stages of transportation decision-making. While less common, toll agencies may also have reason to review the distributional effects of tolling, particularly in advance of toll changes, to support deliberations or public inquiries. The analyses can be readily informed by available travel survey, census materials such as the American Community Survey’s PUMS, or by travel demand modeling already developed by the MPO or prepared in support of the transportation-related and socioeconomic-related environmental studies during the National Environmental Policy Act stages, or for traffic and revenue studies. Who Has Used It Successfully? Various approaches for employing the user cost method to consider the burden on low- income populations have been described here with reference to three case examples. The discus- sions highlight how the techniques were employed along with some potential limitations. More information can be found by examining the references presented in the resources section. Resources FHWA and Texas DOT. 2009. “Joint Guidance for Project and Network Level Environmental Justice, Regional Network Land Use, and Air Quality Analyses for Toll Roads,” Washington, D.C. KYTC and Indiana DOT. 2015. Assessment of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations. Retrieved from http://kyinbridges.com/wp-content/ uploads/2015-04-24_EJ-Assessment-Plan.pdf. Madi, M., Wiegmann, J., Parkany, E., Swisher, M., and J. Symon. 2013. Guidebook for State, Regional, and Local Governments on Addressing Potential Equity Impacts of Road Pricing. FHWA-HOP-13-033. U.S. DOT, Washington, D.C. Plotnick, R. D., Romich, J. L., Thacker, J., and Dunbar, M. 2011. “A Geography-Specific Approach to Estimat- ing the Distributional Impact of Highway Tolls: An Application to the Puget Sound Region of Washington State.” Journal of Urban Affairs, Vol. 33, No. 3, pp. 345–366. Schweitzer, L., and Taylor, B. D. 2008. “Just Pricing: The Distributional Effects of Congestion Pricing and Sales Taxes.” Transportation, Vol. 35 No. 6, pp. 797–812. Texas DOT and Alamo Regional Mobility Authority. 2015. United States Highway (US) 281. Final Environmen- tal Impact Statement, Appendix E: Environmental Justice Project Level Toll Analysis, Technical Report. Retrieved from http://www.411on281.com/us281eis/assets/File/2015/20%20-%20Appendix%20E%20- %20Enviromental%20Justice%20Project%20Level%20Toll%20Analysis.pdf. U.S. DOT. 2012. Revised Record of Decision. Louisville-Southern Indiana Ohio River Bridges. Jefferson County, Kentucky and Clark County, Indiana. FHWA-KY-SEIS-12-01-F. Retrieved from http://updates.kyinbridges. com/wp-content/uploads/2012/06/LSIORB-Revised-ROD-w-signature-6-20-2012.pdf.

228 What Is It? The evaluation of disproportionately high and adverse effects on low-income and minority populations is the root of an EJ analysis. “Disproportionately high and adverse effects” are defined, according to FHWA Order 6640.23A, as those that are predominantly borne by a minority and/or low-income population or will be suffered by the minority and/or low-income population and is appreciably more severe or greater in magnitude than the adverse effect that will be suffered by the non-minority or non-low-income population. Although EO 12898 directs federal agencies to develop a strategy to identify and address the disproportionately high and adverse human health or environmental effects of their actions on low-income and minority populations, no guidance is provided as to the criteria for the determination of “disproportionately high and adverse” effects. The Council on Environmental Quality’s 1997 guidance on EJ is purpose- fully vague to allow agencies to integrate analyses of EJ concerns in an appropriate manner. The Texas DOT and the Washington State DOT have developed quantitative methods to evaluate disproportionately high and adverse effects. Why Is It Effective in Environmental Justice Analysis? For some projects (regardless of tolling), making the determination of whether there is a disproportionately high and adverse impact is informed by the location of the affected protected populations and the project footprint. However, for some transportation and tolling-related projects, low-income and minority populations outside of the project footprint can experience the impact from the imposition or increase in tolls. The ability to quantify impacts as they relate to tolling can help to ensure that low-income and minority populations do not suffer a dis- proportionate burden compared to the general population. In addition to quantifying potential impacts, it is important to compare the impacts on the low-income and/or minority populations with respect to the impacts on the overall population within the project area. Fair distribution of the beneficial and adverse effects of the proposed action is the desired outcome. T O O L 1 1 Evaluating Disproportionate Effects with Quantitative Methods Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act Tools and Techniques • Statistical Analyses • Comparison of Impacts and Population Percentages Affected Populations • Low-Income • Minority Examples Featured • Texas DOT, Guidebook on Environ- mental Justice Impacts of Toll Roads • Washington State DOT, Determining Project Effect on EJ Populations

Evaluating Disproportionate Effects with Quantitative Methods 229 What Are Some Techniques for Implementing This Tool? Texas DOT and Washington State DOT have developed different approaches for evaluating disproportionately high and adverse effects quantitatively. Example 1: Using Statistics to Evaluate Potential Disproportionately High and Adverse Effects, Texas DOT. The Texas DOT Guidebook for Identifying, Measuring and Mitigating Environmental Justice Impacts of Toll Roads (Prozzi et al., 2006) includes guidance on evaluat- ing dis proportionate impacts. Texas DOT addresses the issue for quantitative resource topics using statistical tests of significance to establish disproportionate impacts as follows. The fol- lowing excerpt from the guidebook begins with a statement on disproportionate impacts. Are the EJ Communities Disproportionately Impacted by the Toll Road? This is arguably the least well-defined aspect of EJ analysis. No guidance is available from Title VI or EO 12898 as to the criteria for adverse or disproportionate and limited guidance is provided by the CEQ. Step 5 of the EJ Evaluation Method (EJEM) [sic] thus determines whether the impacts imposed by a toll road on zones with medium and high concentrations of EJ populations are statistically significantly higher compared to zones with low concentrations of EJ populations . . . Step 5 of Texas DOT’s EJ evaluation method includes two elements. First, the analyst should determine if the measured impacts associated with the tolled alternative (Alternative 2 in Table 1) are statistically significantly higher than the measured impacts associated with the non-tolled alternative (Alternative 1 in Table 1) for the different concentrations of low-income and/or minority populations (i.e., low, medium, high). Second, if the impact imposed by the tolled alternative is determined to be higher and statistically significant, the analyst should determine whether the impact imposed on zones with high and medium concentrations or low-income and/or minority populations is statistically significantly higher than the impact imposed on zones with low or no concentrations of low-income and minority populations. Table 1 provides a graphical representation of the vertical and horizontal comparisons that need to be undertaken. The statistical test to determine whether there is a statistically significant difference between the impacts imposed by the tolled and non-tolled alternatives (i.e., vertical comparison) is the “paired t-test” based on paired data analysis (see text box, Analysis of Paired Data Using a One-Sample t-Test). In the example provided by Texas DOT, the data do not provide enough evidence to conclude that the difference in access to employment between the two alternatives is statistically significant. If there were a statistically significant impact shown by the paired t-test, the analyst would move to the next step to determine whether the impact imposed by the alternative is experienced Table 1. Comparisons required to determine significant impacts. Alternatives EJ Concentration Zones Low Medium High 1 (non-toll road condition) MI01 ↕ ↔ MI02 ↕ ↔ MI03 ↕ ↔2 (toll road condition) MI11 MI12 MI13 Source: Prozzi et al., 2006 Notes: ↕ = comparison between the toll and non-toll alternative; impacted EJ concentration zones given a statistically significant impact; MI = measured impact. ↔ = comparison between

230 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox differently by the different concentrations of low-income and minority populations. The statis- tical test to determine whether the impact on zones with high and medium concentrations of low-income and/or minority populations is significantly higher than on zones with no or low concentrations of low-income and/or minority populations is a “large sample test” based on differences between population proportions. The text box, Inferences Concerning a Difference between Population Proportions, provides a hypothetical example of a traffic noise impact at a toll plaza and evaluates whether or not the impact is experienced differently by different concentrations of low-income and/or minority populations. In the example provided by Texas DOT, there is not a statistically significant difference between the noise impacts experienced by the zones with high/medium concentrations and the zones with no/low concentrations of low-income and/or minority populations. Analysis of Paired Data Using a One-Sample t-Test A transportation agency is considering the conversion of a planned non-toll road into a toll road prior to opening the road to the public. To assess whether a disproportionate impact will be imposed, access to employment by EJ concentration zones has been estimated using TransCAD. The table below shows the number of employment opportunities that can be reached within 30 minutes by car in zones with high concentrations of low-income and/or minority populations given the two alternatives. Does the data suggest that the number of employment opportunities accessible within 30 minutes by car in zones with high concentrations of EJ populations is significantly less, given the toll road compared with the non-toll road, at a 0.05 significance level? Number of Jobs Accessible within 30 Minutes by Car High Concentration Zones of EJ Populations Toll Road Condition Non-Toll Road Condition Difference (D) 1 19 15 4 2 21 20 1 3 18 22 –4 4 5 8 –3 5 34 25 9 6 12 17 –5 The hypothesis of interest is H0: I2 - I1 = 0 (versus I2 - I1 < 0). At level 0.05, H0 should be rejected if t ≤ -t0.05,5 = -2.015. Since the value of the test statistic is 0.15, H0 cannot be rejected. Therefore, the data does not provide enough evidence to con- clude that access to employment in zones with high concentrations of low-income and minority populations is less given the toll road at a 0.05 significance level. Source: Prozzi et al., 2006

Evaluating Disproportionate Effects with Quantitative Methods 231 Example 2: Comparison of Adverse Effects between Low-income and/or Minority Popu- lations and Non-EJ Populations, Washington State DOT. In contrast with Texas DOT, Washington State DOT considers percentage differences for effects rather than statistical tests to determine disproportionate impacts (Washington State DOT, 2014). The overall guidance is useful and straightforward, and it provides another valid perspective on assessing and making a reasoned determination of the findings for disproportionate impact. An overview of the methodology used in determining the effect of a project on low-income and/or minority populations is provided by Washington State DOT. The method includes a four-step model based on a comparison of the demographic data and the geographic extent of identified adverse impacts: 1. Determine the percentage of minority and low-income populations in areas with anticipated adverse effects. 2. Determine the percentages of minority and low-income populations in areas without antici- pated adverse effects. 3. Compare the two percentages. If the areas where there are adverse effects have a higher per- centage of minority or low-income residences or businesses than the area without adverse effects, then the effect is disproportionate. Inferences Concerning a Difference between Population Proportions A traffic noise analysis reveals that neighborhoods located near a toll plaza are exposed to noise levels that exceed the FHWA’s noise abatement criteria (67 dbA). An analyst has identified the affected EJ concentration zones by overlaying the racial characteristics of these zones with the results from the noise analysis (see table below). Do the data suggest that the proportions of populations in zones with high/medium concentrations of EJ populations affected by excessive traffic noise are less than the proportion of populations in zones with no/low concen- trations of EJ populations at a 0.025 significance level? EJ Concentration Zones High/Medium No/Low Total Total population in the study area m = 569 n = 178 747 Population exposed to noise level > 67 dbA x = 301 y = 156 457 Sample population p1 = 0.529 p2 = 0.876 p3 = 0.612 Let p1 and p2 denote the two population proportions. The hypotheses of interest are H0: p1 - p2 = 0 versus Ha: p1 - p2 < 0. At a 0.025 significance level, H0 should be rejected if Z ≤ -Z.025 = -1.96. Since the value of the test statistic is -8.30, H0 must be rejected. The p-value is so minuscule that at any reasonable level a, H0 should be rejected. The data thus strongly suggest that zones with high/medium concen- trations of EJ populations are not disproportionately affected by traffic noise compared to zones with no/low concentrations of EJ populations. Source: Prozzi et al., 2006

232 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox 4. Compare the severity of the impact, potential mitigation, and other project benefits that may offset the impact to verify or refute the determination of a disproportionately high and adverse effect. Consider public perception of the severity of the impact in the analysis (Washington State DOT, 2014). Washington State DOT uses the term “right-size,” meaning the analysis effort should match the size and complexity of a given project. The guidance provides details for each of the four steps; however, this tool focuses on the comparison of potential impacts between populations. The details of the comparison steps in the guidance are shown in the accompanying text box, Making a Determination of Disproportionately High and Adverse: Comparing the Adverse Effects and Benefits Borne by Different Populations. Making a Determination of Disproportionately High and Adverse: Comparing the Adverse Effects and Benefits Borne by Different Populations Washington State Department of Transportation (WSDOT) offers guidance in making a determination of projects effects on EJ populations. Below is a para- phrased excerpt of that guidance. • Collect resource discipline effects analysis (e.g., noise, air toxins, Section 4(f), relocations, public services and utilities, land use, transportation, visual, hazardous materials, and cultural resources). – Only disciplines that have adverse effects will need to be considered in the analysis. If none of the disciplines have adverse effects, the WSDOT Guidance recommends skipping to the final step of using best professional judgment. • Review public comments for a given project and summarize the issues of con- cern for minority and/or low-income populations by neighborhood or location. – If the minority and/or low-income population does not feel that the project effects are adverse, then there is no controversy. • Overlay the adverse effects for each resource discipline on a map showing the demographic data previously collected. – Show demographic data for each census block as a percentage to identify population percentages (i.e., for each minority category, for low-income, and for the population that is neither minority nor low-income). – Show the adverse environmental effects for each resource in a way that allows visualization of the type and severity of the effect. The complexity of the graphics will vary depending on the project. A series of maps or a series of overlays may be used if the data are complex. Contact the WSDOT Graphics Office for assistance. – Confirm that the geographic data, discipline data, severity ratings, and demographic data are correct. Include the data sources on the map and note if preliminary findings for any of the disciplines are used. • Determine the minority and low-income population percentages in areas with and without anticipated adverse effects. – Determine the minority and low-income population percentages for the resource analysis areas for each discipline. – Determine the location and severity of the anticipated adverse effects.

Evaluating Disproportionate Effects with Quantitative Methods 233 What Are Its Limitations? Texas DOT’s example using the paired t-test illustrates how to determine whether or not there is a statistically significant access burden compared to a non-toll road. This analytical approach may be particularly valid for system or metropolitan-level planning analyses. In the example, the data do not provide enough evidence to conclude that the toll road delivers less access to employment in zones with high concentrations of low-income and minority population given the 0.05 significance level. This does not necessarily mean that there is no impact related to access to employment for low-income and/or minority populations in the study area. In addition, regardless of system-wide or project-specific analyses, evaluating differences between opportu- nities for or access to employment may not provide information about whether the people living in the zones will be affected by access (e.g., travel time on tolled versus non-tolled facilities) to the jobs they currently have. In evaluating the potential effects of tolling facilities, the analyst is likely to look at use of the toll road versus use of a non-tolled route. The comparison of percentages approach outlined by Washington State DOT may be more suited to project footprint impacts than travel behavior impacts from tolling, unless the tool is used in concert with an approach that compares travel time between tolled and non-tolled alternatives. Washington State DOT’s comparison of percentages may also appear limiting in that there is not a prescribed threshold or bright line statistical test for how much greater the burden must be to be considered disproportionately high and adverse. For example, with respect to tolling, it • Compare the percentages and make a preliminary determination about the anticipated impacts on minority and/or low-income populations. – If the areas where the anticipated adverse effects have a higher percentage of minority and low-income populations, there may be a “disproportion- ately high and adverse” effect. – A “disproportionately high and adverse” determination may be made if: � The severity of the adverse impact is appreciably greater for minority and low-income populations compared to areas without minority and low- income populations. � More adverse environmental impacts (multiple resources or issues of concern) occur in areas with minority and low-income populations (regardless of severity) than in areas without minority and low-income populations. � Proposed mitigation is not sufficient to reduce either the level of severity or number of adverse effects for minority and low-income populations to be commensurate to that experienced by areas without minority and low-income populations. � The project benefits do not affect minority or low-income populations to the same degree as other populations. � The project is controversial and public comment shows that minority and/or low-income populations do not perceive that the project benefits them or that the proposed mitigation is adequate. • Use your best judgment and professional expertise to make a final determination. Source: WSDOT, 2014

234 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox can be challenging to determine what percentage of a traveler’s budget would be used for tolls without creating an adverse effect that is appreciably more severe or greater in magnitude for a low-income household. The Washington State DOT approach, however, acknowledges that the EJ impact determination is more than a technical or quantitative exercise; public involvement and the level of controversy, severity of the impacts, mitigation strategies and the potential off-setting benefits will inevitably play a major role in shaping any final determination. The evaluation pro- cess and the rationale for the impact determination will be documented to justify findings. What Types of Resources and Costs Are Required? Both methods rely on similar resources with regard to identifying potential low-income and/or minority populations (see tool, “Developing a Socioeconomic Profile and Community Characteristics Inventory for EJ Assessments”) and both methods would draw on readily avail- able tools (e.g., GIS, computer-aided design, travel demand models, noise and air quality models) and the expertise of a multi-disciplinary team of subject matter experts in key disciplines to iden- tify potential adverse and beneficial impacts. GIS software would be used to overlay the impacts and demographic information. The Texas DOT methods require the analyst to be well-versed in statistics; the Washington State DOT method does not. Who Has Used It Successfully? The examples briefly described the analytical steps and criteria used to make a determination of disproportionately high and adverse effects on low-income and minority populations. The methods and potential limitations of using solely quantitative methods to make a determination are also discussed. Resources Council on Environmental Quality. 1997. Environmental Justice: Guidance under the National Environ- mental Policy Act. CEQ, Executive Office of the President. Issued December 10, 1997. Retrieved from http://www.whitehouse.gov/CEQ/. Prozzi, J., Victoria, I., Torres, G., Walton, C. M., and Prozzi, J. 2006. Guidebook for Identifying, Measuring and Mitigating Environmental Justice Impacts of Toll Roads. Center for Transportation Research, University of Texas at Austin. Retrieved from http://www.utexas.edu/research/ctr/pdf_reports/0_5208_P2.pdf. Washington State DOT. 2014. WSDOT, ESO, NEPS/SEPA Compliance. TSK 458–d: Determining Project Effect on EJ Populations. Retrieved from http://www.wsdot.wa.gov/NR/rdonlyres/931F369F-F3C6-421D-98DC- 253D8E7518E0/0/EJ_DetermineEffect.pdf.

235 What Is It? Cash replenishment options address a possible barrier to the use of electronic tolling facilities caused by an absence of a credit card or a debit card to purchase or make a needed deposit to acquire a transponder or replenish an account. Why Is It Effective in Environmental Justice Analysis? One of the key considerations for addressing EJ considerations with toll implementation is equitable access to toll facilities. As agencies implement electronic tolling systems, individuals without a credit card or bank account face the prospect of being excluded or financially penal- ized for their use through the payment of a higher toll for using the cash lane or for a video transaction in the absence of a transponder. Popula- tions in this circumstance may experience a form of social exclusion (i.e., compromised access to jobs and other amenities critical to sustain- ing a household’s quality of life). While there are many cited benefits to electronic tolling (e.g., traffic safety, operating efficiencies, financial control of operations, less congestion and delays, improved air quality, and smaller land consumption at tolling facilities), addressing the needs of population groups without access to credit cards or bank accounts is paramount to mitigating undue burdens associated with purchasing transponders and replenishing user accounts. The magnitude of this issue can be better understood after reviewing some key definitions and statistics about unbanked and underbanked households. An unbanked person is someone with no bank account of any kind—no checking, savings, or credit card. An underbanked person may have a bank account but uses services like check cashing, money orders, and payday loans. Recent studies indicate the following patterns and trends: • 72 percent of U.S. consumers had at least one credit card in 2012 according to researchers at the Federal Reserve Board in Boston (Schuh and Stavins, 2014). This translates into about 65 million adult consumers without a credit card. • In the aftermath of the 2008–2009 Great Recession, more young people may be opting out of using credit cards. One Gallup Poll estimate found that the percentage of the U.S. population t o o l 1 2 Instituting Cash Replenishment Options for Unbanked and Underbanked Populations Framework Step • Identify and Assess Mitigation Strategies Stages in Decision-Making • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Financial Operations Planning • Market Research • Procurement Affected Populations • Low-Income • Unbanked and Underbanked Examples Featured • Puerto Rico Highway and Trans- portation Authority, Puerto Rico AutoExpreso • Metropolitan Transportation Authority, New York City, E-ZPass Cash Reload Card • Florida Turnpike Enterprise, SunPass Mini • Washington State DOT, Electronic Benefit Card

236 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox without a credit card had increased by 7 percentage points (22% to 29%) between 2008 and 2014 (Swift, 2014). While some individuals may choose not to have a credit card, it is likely that the majority of those without a credit card cannot acquire one due to financial constraints. • 7.7 percent of all U.S. households were estimated to be unbanked in 2013, totaling some 9.6 million households (FDIC, 2014). • 20 percent (24.8 million) of U.S. households were underbanked in 2013, meaning that they had a bank account but also used alternative financial services outside of the banking system (FDIC, 2014). • Among the most likely to be unbanked are non-Asian minority populations, lower-income households, younger households, and the unemployed. Working-age disabled households also exhibit higher patterns of being unbanked and underbanked (FDIC, 2014). Juxtaposing these statistical patterns with car ownership rates, it is reasonable to conclude that many households with cars may be without access to a credit card or do not rely upon traditional financial institutions. According to the 2014 U.S. Public Use Microdata Sample (PUMS), only 9.2 percent of U.S. households were car-free, which is a percentage well below that without credit cards or bank accounts. In the absence of cash options for the purchase of a transponder or replenishment of an account, a typically low-income segment of the U.S. population must absorb a higher toll cost for a video transaction, or forego the use of an all-electronic tolling facility, when they require or need to use a toll road. What Are Some Techniques for Implementing This Tool? Transportation agencies have gradually begun to develop policies and services that respond to the potential difficulties in securing transponder access and replenishing accounts. The follow- ing information showcases how four different agencies are implementing options for transpon- der access and cash replenishment. Example 1: Puerto Rico AutoExpreso. In light of the fact that some 40 percent of Puerto Ricans were estimated to be without a bank account, the Puerto Rico Highway and Transporta- tion Authority implemented a system that allows its users to swipe their AutoExpreso card at gas stations and convenience stores and proceed to pay inside the store (Parkany, 2005). Users can purchase transponders at these retail outlets and replenish their accounts. The program has been designed to preserve the anonymity of users, which directly addresses the privacy concerns of some drivers reluctant to use transponders. A light on the toll tag also signals drivers that their account balance is running low and that it is time to replenish it. Example 2: E-ZPass Cash Reload Card—Market Research. To explore ways to encourage its cash users to convert to its E-ZPass electronic toll collection system, the New York City Metropolitan Transportation Authority’s (MTA) Bridge and Tunnels, legally known as the Triborough Bridge and Tunnel Authority, conducted a stated preference market survey. Market research led the MTA to conclude that installation of ATM-like kiosks and in-store refill methods could significantly increase the willingness of cash customers to make the conversion. Subsequently, MTA introduced its Cash Reload Card for customers who want the ease of use and savings of E-ZPass but greater control of their cash. The Cash Reload Card is supported by a credit card network, Visa ReadyLink, which offers cash replenishment options through reloadable prepaid credit cards at several hundred locations in the New York Metropolitan region and at more than 2,500 stores nationally (e.g., pharmacies, convenience stores, etc.). The Cash Reload Card can be acquired at the E-ZPass cash lane, by mail, at walk-in customer service centers, or by calling a toll-free customer service number. Mobile phone users that accept text messages can opt-in for mobile alerts, notifying them of key account activity and triggers, including low account balances. The minimum reload amount is either $20, or an amount sufficient to bring an account balance to a minimum of $10, whichever is greater. There is a fee of $2 or less to reload the card through an MTA preferred retailer or kiosk locations.

Instituting Cash Replenishment options for Unbanked and Underbanked Populations 237 Example 3: Florida SunPass’s Criteria for Selection of Retail and Money Service Vendors. Florida Turnpike Enterprise offers an affordable transponder, the SunPass Mini, at $4.99. When the SunPass Mini was introduced, it was essentially free because the $4.99 was added back to the customer’s SunPass account once activated. The SunPass Mini transponder is available at nearly 2,000 locations throughout the state, including pharmacies, grocery stores, and financial services as well as the American Automobile Association (AAA). Transponders are also available at all Turnpike service plazas, Turnpike gas stations, and SunPass Service Centers. For customers who do not have a credit card or a debit card, transponders can be replenished with cash at nearly 5,000 retail locations throughout the state of Florida. Marketing materials, including the website for SunPass, are provided in English and Spanish. The Florida Turnpike utilized a request for proposal process to select retail and money ser- vices vendors that contained several operating requirements (see text box, Example of Florida Turnpike Enterprise SunPass Vendor Requirements). The evaluation criteria applied to review money services proposals, included: • Quality and completeness of proposed solutions that meet the requirements and objectives identified in the scope of services; • Perceived customer friendliness and convenience of services proposed; • Customer convenience of retail locations, including number of locations and statewide coverage; • Schedule for development and implementation in light of the agency’s plans for operation; • Provides best geographic fit with the Florida DOT’s all-electronic tolling facilities; • Deployment as described in the scope of services; and • Innovation of additional services provided in addition to those identified in the scope of services. Example of Florida Turnpike Enterprize SunPass Vendor Requirements The request for proposal criteria included the following requirements for retailers to serve as SunPass Vendors: 1. Retailer must be in the retail business engaged in the sale of products or prod- ucts services at a retail price to the consumer for at least five (5) years. 2. Retailer must be engaged in the sale of a variety of retail products and services. 3. Retailer must be a statewide Florida business with at least 75 statewide retail outlets with a broad distribution network throughout the State of Florida. The distribution network of retail outlets must at a minimum include locations in the Jacksonville, Orlando, Tampa/St. Petersburg, the Florida Panhandle, and South Florida markets. 4. Retailer must have walk-in store locations operating under the same trade name and same ownership. Retailer must display and sell SunPass Portable and Mini Transponders at all Florida store locations. 5. Retailer shall have no more than two Florida warehouse distribution centers where all shipments of SunPass Transponders by the Department will be received by the retailer. 6. Retailer must agree to order and receive pallet quantities as follows: 2,304 SunPass Mini Transponders (one pallet); 576 Portable Transponders (one pallet). 7. Retailer cannot be under the jurisdiction of the bankruptcy courts of the United States. 8. Retailer shall be responsible for any and all costs associated with the sales of transponders at its stores.

238 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Example 4: Washington State DOT, Electronic Benefit Card. Washington State DOT estab- lished its “Good To Go” card to handle the processing of electronic toll transactions on all its toll facilities. Drivers may set up a prepaid account by purchasing and activating a pass that mounts on the vehicle or by registering their license plate for photo identification (i.e., a video transac- tion). For those without credit cards or bank accounts, one payment option is to visit a customer service center located in Seattle, Bellevue, or Gig Harbor to activate an account using EBT or cash. An EBT is an electronic system that enables state welfare departments to issue food and cash benefits via a magnetically encoded payment card. Food benefits are federally authorized benefits distributed through the Supplemental Nutrition Assistance Program, formerly the Food Stamp Program, for the purchase of food and non-alcoholic beverages. Cash benefits include state general assistance, Temporary Assistance for Needy Families benefits, and refugee benefits. Washington State permits the use of its EBT card to purchase a transponder and replenish an accompanying account. What Are Its Limitations? Instituting cash replenishment options offers a good-faith strategy for addressing transponder usage barriers for the underbanked and unbanked; however, other constraints to their usage should be recognized. Issuers of prepaid reloadable cards, such as the MTA Cash Reload Card, may not require good credit or a positive bank history to issue the card, but some identity verifica- tion information is likely to be required for setting up an account such as name, address, phone number, date of birth, and Social Security number. Supplying this identity-related information may inhibit usage for individuals with privacy concerns and those with low literacy, language barriers, or disabilities, who may be unable to complete the application-related activities. Transportation agencies should also appreciate that fees, disclosure, and customer service vary among prepaid card issuers. Some card issuers apply fees for activities such as card activa- tion, balance account information, and card maintenance. Fee disclosure policies also can vary with less credit protection and transparency than typically required for credit cards. Customer service levels may differ among issuers, which could affect a customer’s satisfaction and commit- ment to the option. Some issuers offer free customer service and opportunities to speak directly with a customer service representative but others limit interactions to automated voicemail or email contacts for consumers with questions or problems. As important as cash replenishment options may prove to be for addressing a barrier to use for a segment of unbanked and underbanked populations—often low-income households— the option itself does not fully address other recognized barriers to transponder usage for low- income persons that are embedded in toll policies and the attributes of toll accounts. These toll policies and account attributes (e.g., the size of transponder deposits, initial prepayment amounts, replenishment amount policies and automatic deductions, maintenance fees, and close-out policies) also affect an individual’s willingness to consider the option (Parkany, 2005). These initial out-of-pocket costs can make transponder usage out-of-reach financially for low- income households despite the fact that tolled facilities that institute electronic toll collection tend to incentivize transponder usage by offering discounted tolls for their use and higher prices for those who rely on video tolling or booth tolling. Because these toll policies and attributes affect travel patterns and the assessment of impacts, agencies and practitioners need to examine them as early as possible in the planning phase, as part of the project action or alternatives under consideration during the environmental review process and before the issuance of a record of decision, and for program evaluation during the operational phase. FHWA and Texas DOT’s Joint Guidance for Environmental Justice Analy- ses of Toll Roads in Texas acknowledges these considerations and seeks greater disclosure by

Instituting Cash Replenishment options for Unbanked and Underbanked Populations 239 specifying the need for agencies and practitioners to identify and assess these considerations in planning and environmental review stages (see text box, Assessing Toll Collection Methods and Cost Differences). Assessing Toll Collection Methods and Cost Differences FHWA and Texas DOT’s Joint Guidance for EJ Analyses of Toll Roads in Texas lists considerations for discussion for all projects that include toll features, including toll collection methods, cost differences, and possible barriers. • Toll collection methods (e.g., electronic toll collection, toll booths) and how it affects access and costs. • Cost differences between toll tags purchased with credit cards or debit cards versus cash, including how and where individuals may set up and maintain toll accounts. • Accommodation for limited English proficiency and persons with disabilities to allow these populations to access the toll facilities. For example, the TxTag website is available in Spanish and provides a customer service contract number for the deaf and hard of hearing. What Types of Resources and Costs Are Required? Facilitating cash replenishment options requires considerable operational pre-planning and market research prior to implementation to build awareness of the option among vendors and the public. Marketing and awareness campaigns are essential to educating the public about the benefits of the program for managing personal cash flow and the discount or cost advantage of the cash replenishment options vis-à-vis booth and video transactions. The cost for implementation of infrastructure, technology, and building relationships with vendors may differ for regions where tolling agencies have already built their infrastructure and fostered relationships with vendors. The Florida Turnpike Enterprise, for example, required significant new infrastructure, creating new kiosks and mobilizing private kiosk vendors. The Puerto Rico AutoExpreso was also highly innovative in its approach. The MTA Bridges and Tunnels initiative, by comparison, was built on its past history of operations with retailers and gas stations due to the E-ZPass On-The-Go program, which allowed customers to purchase unregistered tags from select retail stores. Who Has Used It Successfully? Various approaches for planning and implementing cash replenishment options are described in the four examples presented in this section. The discussions highlight the techniques that were employed along with some potential limitations. More information can be found by examining the references presented in the resources section. Resources Campbell, M., Spitz, G, Carpenter, C., Ramchal, K., Jacobs, D. 2011. “The Effect of Improved Replenishment Options to Convert Cash Users to Electronic Toll Collection.” Presented at the 90th Annual Meeting of the Transportation Research Board, Washington, D.C.

240 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Doering, R. W., Ruiz, L. A. S. 2004. Enhancing the Value in AVI Low Cost eGo™ Sticker Tags Operational in Puerto Rico’s AutoExpreso™ System. Presented at the International Bridge, Tunnel and Turnpike Association Technology Workshop. Florida DOT. 2010. Request for Proposals, Money Services for SunPass Transponder, RFIP-DOT-09/10-8001-SM. Federal Deposit Insurance Corporation. 2014. 2013 FDIC National Survey of Unbanked and Underbanked Households. Florida DOT. 2014. Request for Proposals, Retail Sales Opportunity for SunPass Transponder, RFIP-DOT-13/ 14-8001-SM. MTA. 2012. How to Use the MTA Cash Reload Card (English). Retrieved from https://www.youtube.com/ watch?v=v0XF9FCzqVg. MTA. No date. E-ZPass Cash Reload Card Website: Retrieved from http://web.mta.info/bandt/ezpass/reload.html. Parkany, E. 2005. “Environmental Justice Issues Related to Transponder Ownership and Road Pricing.” Trans- portation Research Record: Journal of the Transportation Research Board, No. 1932. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 97–108. Schuh, S. and Stavins, J. 2014. The 2011 and 2012 Surveys of Consumer Payment Choice. Federal Reserve Bank of Boston, No. 14-1. Swift, A. 2014. “Americans Rely Less on Credit Cards than in Previous Years”. Gallup Organization. Retrieved from http://www.gallup.com/poll/168668/americans-rely-less-credit-cards-previous-years.aspx.

241 What Is It? Federal law requires toll revenues to be used first to cover the costs of developing, operating, and maintaining the toll facility and providing a return on investment to any pri- vate investment partner. Excess revenues may be used for any transportation purpose within the subject corridor for which federal funds would normally be used. Toll payment assistance or infrastructure and transit advancements that benefit EJ communities, a technique known as revenue recy- cling, are ways to mitigate adverse financial and physical effects of the tolling projects. Why Is It Effective in Environmental Justice Analysis? Recycling toll revenues into transportation infrastructure and services can increase passenger throughput in the cor- ridor and potentially benefit all users, not just those who can afford to pay the fee, particularly if there is a non-tolled gen- eral purpose lane. Subsidizing public transportation services or its needed infrastructure and equipment can offer an option for low-income populations seeking mobility and access along tolled corridors. Dedicating a portion of the toll revenue to finance transportation improvements, such as more frequent transit services or vanpools, may offer ben- efits that offset the adverse effects of tolling and toll facilities for low-income travelers. Increasingly, transit vehicles can use managed lanes free of charge, but there must be sufficient transit service for the corridor to function as a feasible alternative to driving. Ded- icating a portion of the excess toll revenue to fund transit service in the corridor is a means for ensuring that a toll-free option exists. For toll projects, mitigation can vary depend- ing on the type of impacts borne by EJ populations, but can be contentious depending on the mode or route funded. t o o l 1 3 Recycling Tolling Revenue through Transit Investment and Low-Income Assistance as Forms of Mitigation Framework Step • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Project Design and the National Environmental Policy Act • Implementation Tools and Techniques • Revenue Recycling • Transit Services • Transit Rewards Programs • Toll Account and Transponder Policies • Toll Relief Affected Populations • Low-Income • Minority Examples Featured • State of Minnesota, I-394 MnPASS Express Lanes • Los Angeles County Metropolitan Transportation Authority, I-110 and I-10 ExpressLanes • San Diego Association of Governments, I-15 FasTrak HOT Lanes • Indiana DOT and Kentucky Transportation Cabinet, Ohio River Bridges • Elizabeth River Crossings and Virginia DOT, Elizabeth River Tunnels • Florida DOT, I-95 Express Toll Lanes • Metropolitan Transportation Authority; Port Authority of New York–New Jersey, Legacy Tolling Systems

242 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Setting aside revenues to fund discount and credit programs offers another alternative for ensuring mobility for low-income travelers. Toll relief for low-income populations, for example, can reduce the financial burden but still allow low-income commuters to enjoy the potential time-savings benefits of the priced lanes when most needed. However, the implementation of these policies requires close attention to eligibility and administrative requirements. What Are Some Techniques for Implementing This Tool? Improving public transit access and its functionality along tolled corridors is an important strat- egy to address potential adverse effects borne by low-income commuters who are in need of more affordable transportation options. Project sponsors have also opted to apply excess toll revenues toward toll discounts for low-income commuters, transit rewards programs, active transportation projects, or additional maintenance and operation expenses in adjacent lanes. States must adopt toll revenue recycling policies—and in some cases, legislation to grant authority—that define the specific uses for excess toll revenues as part of their operations and business plans. The following examples demonstrate how various agencies have implemented a variety of mitigation strategies. Example 1: I-394 MnPASS Express Lanes, Minneapolis, Minnesota. The HOT lanes on I-394, in the 10-mile corridor running west from Minneapolis, were the first of their kind in Minnesota. When working to convert the I-394 HOV lane into a HOT lane, the two strongest public criticisms were that value pricing would decrease HOV use and concerns of overall eco- nomic fairness. HOT lanes were perceived as primarily serving wealthier travelers—those who can afford to pay the toll will travel more efficiently in the toll lanes. In 1997, the state withdrew the HOT lanes proposal after public criticism, but eventually received legislative endorsement for the HOT lanes in 2003 (Minnesota DOT, 2013). Minnesota pledged to dedicate 50 percent of the excess project revenues to highway improvements and 50 percent to transit improvements within the corridor, which helped to solidify public support (see text box, Excerpt from Minnesota Legislative Statute Describing Special Revenue Fund and Purposes). Excerpt from Minnesota Legislative Statute Describing Special Revenue Fund and Purposes From Section 160.93 User Fees; High Occupancy Vehicle and Dynamic Shoulder Lanes, Subdivision 2 includes the following section on Deposit of revenues; appropriation: a) . . . money collected from fees authorized under Subdivision 1 must be deposited in a high-occupancy vehi- cle lane user fee account in the special revenue fund. A separate account must be established for each trunk highway corridor. Money in the account is appropriated to the commissioner. b) From this appropriation the commissioner shall first repay the trunk highway fund and any other fund source for money spent to install, equip, or modify the corridor for the purposes of Subdivision 1, and then shall pay all the costs of implementing and administering the fee collection system for that corridor. c) The commissioner shall spend remaining money in the account as follows: (1) One-half must be spent for transportation capital improvements within the corridor; and (2) One-half must be transferred to the Metropolitan Council for expansion and improvement of bus transit services within the corridor beyond the level of service provided on the date of implementation of subdivision Minnesota Legislative Statute 160.93 2014

Recycling tolling Revenue through transit Investment and low-Income Assistance as Forms of Mitigation 243 Minnesota even updated the goals of the I-394 MnPASS Express Lanes to include this change (Cambridge Systematics, 2006). The I-394 MnPASS goals are to: • Improve I-394 efficiency; • Maintain free flow speeds in MnPASS lane; • Collect tolls electronically; • Use excess revenues, if available, to improve highway and transit in the corridor; and • Employ new technologies for pricing and enforcement. With the pledge to recycle revenues for highway and transit improvements, Minnesota built the public support needed to implement the pricing and operational changes intended to maxi- mize the corridor’s capacity. Example 2: I-110 and I-10 ExpressLanes, Los Angeles, California. As a participant in the U.S. DOT’s Congestion Reduction Demonstration Program, Caltrans and the Los Angeles County Metropolitan Transportation Authority (Metro) were awarded $210.6 million in 2008 to convert the existing HOV lanes to HOT lanes on the I-110 and I-10 freeways (Metro, 2014a). In advance of the HOV to HOT conversion, Metro used federal funds to make substantial investments in transit station expansion, local and bus rapid transit (BRT) service, and vanpool services. Enhancements included headway reductions by as much as 83 percent, the purchase of 59 new buses at about $692,000 each, transit signal priority treatments, vanpool funding, and facility upgrades (Network Public Affairs, 2010; Metro, 2009). In 2015, the Metro Board also recommended allocating up to $6.25 million per year as an operating subsidy for the additional bus service that was implemented (Metro, 2015a). These changes were made with the belief that such improvements would help generate community buy-in of the project and prevent negative ramifications for low-income residents. Since the ExpressLanes have been in operation, toll revenue has been used for maintenance, administration, operations, toll collection, and enforcement, as required by state law. Limited revenue can be set aside as reserve funds, but any remaining revenue is to be used in the corridor where the revenue was generated, and spending must follow an approved expenditure plan of the Metro Board (California Legislative Statute & Hwy Code 188.4 2014). The board, therefore, has the authority to implement the most needed changes in the corridor as resources permit. Funds have been granted on a competitive basis to projects in three broad categories: transit systems improvements (40%), system connectivity/active transportation (40%), and highway system improvements (20%). These projects must be within 3 miles of the I-110 or I-10 Cor- ridor and public agencies that provide transportation facilities or services within Los Angeles County are eligible (Wong, 2015). The competitive grant program prioritizes innovative and sustainable transportation technologies and projects supporting local sustainability policies, and has resulted in millions of dollars being recycled into these projects. For Round 1 of the ExpressLanes program, Metro reinvested approximately $26.7 million collected as toll revenue into mobility projects in the corridor. For Round 2, occurring in fiscal years 2016 and 2017, Metro expects to allocate between $42.5 and $53.3 million (Metro, 2015a). Moving forward, Metro may change its funding priorities. In 2015, the Metro Board recom- mended setting aside an equivalent of 20 percent of the competitive grant funding for allocation to Caltrans since that agency alone is able to implement highway improvements. The Board, however, proposed that 50 percent of Caltrans’ allotment be allocated at the beginning of a project and 50 percent upon on-time completion to promote accountability and efficiency in the absence of competition (Metro, 2015a). Surplus toll revenue also funds Metro’s innovative program for low-income commuters along the ExpressLanes Corridor. The Low-Income Assistance Plan provides qualifying applicants, earning less than twice the federal poverty level, with a one-time $25 credit after setting up their

244 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox prepaid tolling account (see Figure 1). This balance can be applied toward the transponder fee, if applicable, or toward initial toll credit. The plan also waives members’ $1 monthly account maintenance fees (Metro, 2014b). Members qualify based on participation in other government benefits programs. The Transit Rewards Program rewards frequent transit users with toll credits for use in private vehicles. For every 32 one-way trips taken on a transit line in the tolling corridor, the traveler earns $5 in toll credits to be used within 90 days. The system therefore makes it possible for fre- quent transit commuters to enjoy occasional trips in the express lanes at no additional cost when the need or desire arises (Metro, 2014b). As of September 2014, the program was growing steadily and on average enrolling 220 new accounts per month (Metro, 2015b). The Transit Rewards Pro- gram is an innovative program in the transit and toll industry and has been inspired by several research activities (see text box, Research on Model Programs for Toll Credit and Rewards). The programs’ enrollment and monetary values are summarized in Table 1. Metro does not track other costs associated with these programs, such as administration or lost toll revenue. Example 3: I-15, FasTrak HOT Lanes, San Diego, California. The FasTrak HOT lane facility in Greater San Diego was the first dynamically priced HOT lane facility in the world and pro- vided an early model for building public acceptance for HOT lanes (Samuel, 1998). HOT lanes were first administered on parts of I-15 and the South Bay Expressway, opening with electronic tolling in 1998 at a cost of $1.3 billion. The I-15 Express Lanes were significantly expanded between 2008 and 2014, resulting in a four- lane, 20-mile express lane facility in the median of I-15 stretching from State Route 163 north to State Route 78 (SANDAG, 2014). In 2016, revenues from FasTrak tolls and violations were expected to reach $6 million. After subtracting expenses, reserve deposits that could be spent in the corridor were expected to be $907,000. These reserve deposits have helped to fund some of San Diego Association of Gov- ernments’ (SANDAG’s) many expenditures in the I-15 corridor, including $61 million for the Mid-City BRT stations and $70 million for a Direct Access Ramp at the Mira Mesa BRT station (SANDAG, 2015). Despite low voter support, SANDAG has used toll revenues to partially fund improved express bus service in the I-15 HOT lanes since their implementation (Samuel, 1998). The service, the Inland Breeze, offers commuter express service between the Rancho Bernardo area and down- town San Diego and has received about $500,000 per year from SANDAG (San Diego Metropoli- tan Transit System, 2014). In June 2014, the San Diego Metropolitan Transit System launched Figure 1. Low-income assistance plan for eligible Los Angeles County residents.

Recycling tolling Revenue through transit Investment and low-Income Assistance as Forms of Mitigation 245 Research on Model Programs for Toll Credit and Rewards Metro’s Transit Rewards Program rewards transit users for avoiding single-occupancy travel, helping to miti- gate potential equity and EJ issues around congestion pricing. It complements research on toll credit and incentive models for maximizing passenger throughput and operating a roadway system that is accessible to all users. FAIR Lanes: Fair and Intertwined Regular (FAIR) lanes, a concept outlined by DeCorla-Souza in 2000, was a pioneering idea in managing road congestion. In this model, during peak periods two lanes are designated as priced lanes, one or both of which were previously unpriced. To compensate for the perceived loss of a free general purpose lane, drivers who use the general purpose lanes during peak hours with a transpon- der are granted credits. These credits can then be applied towards tolled travel or transit fares. The concept also evolved toward the provision of toll and transit credits for low-income commuters to reduce disparate impacts of such measures (Parsons Brinckerhoff, 2005). High-Occupancy Toll/Credit: The FAIR lanes concept was studied in Santa Clara and Alameda counties in Cali- fornia in 2005. Because political concerns and lack of physical infrastructure made it infeasible to pilot the true FAIR concept, a modified model was applied using just one high-occupancy toll (HOT) lane, which was termed the High-Occupancy Toll/Credit (HOT/C) lanes model. The study compared 14 different scenarios employing HOT/C lanes with varying combinations of tolling and rewards policies. Results suggested that the public was generally favorable for plans benefiting the community as a whole, but there were substantial lost revenues from HOT lanes due to drivers paying with credits instead (Parsons Brinckerhoff, 2005). FAST Lanes: The FAST lanes concept expands on the FAIR lanes concept. Under this model, each driver is granted a limited number of dollar credits each month. Drivers can use these credits to drive in express lanes, and carpoolers can pool their credits as well. Once credits are exhausted, drivers are charged full price for the express lanes, which are dynamically priced based on congestion levels. Residents who choose not to use priced lanes can receive rebates for the unused toll credits in the form of property tax or vehicle registra- tion fee abatements. By compensating drivers for the perceived loss of general purpose lanes, this model has been shown to gain more political support than the other rewards programs (DeCorla-Souza, 2003; Lari and Aultman, 2010). I-30 Express Lanes Survey: With a new priced express lanes system under construction in Dallas–Fort Worth in 2015, the Texas A&M Transportation Institute implemented a pilot program to gauge drivers’ reactions to proposed incentives programs. The incentive structure differed from other programs because it rewarded not only using alternative transportation and carpooling, but also worked as a loyalty program that incentivized driving alone during off-peak hours. The incentives for off-peak driving were equal to those for carpooling: pay for 10 trips, get one free; discounts to local businesses; and various gift cards (Burris et al., 2015). Name of Program Number Households Enrolled Value of Credits Earned Reporting Date Low-Income Assistance Plan 8,882 $220,050 December 2015 Transit Rewards Program 7,000 $26,000 September 2014 Table 1. Program enrollment patterns and estimated value of credits earned for Metro assistance programs.

246 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox a new BRT service—Rapid—to provide higher frequency transit service in those lanes. Physical improvements such as direct access ramps to transit stations allowed buses and HOVs to directly access the express lanes without yielding to traffic in the general purpose lanes and offering greater reliability. The new services run to several destinations via I-15, permitting limited-stop services at greater frequency and extending peak-hour services. With transit service as the focus of its mitigation efforts, SANDAG does not offer discounts for low-income commuters to use the HOT lanes. It does offer periodic promotions to toll road users, such as a refer-a-friend program and occasional prize drawings for toll road users (SANDAG, 2016). Example 4: Ohio River Bridges Project, Louisville, Kentucky, and Southern Indiana. Two Ohio River Bridges—the East End Crossing and the Downtown Crossing—provide a crucial transportation link between Southern Indiana and Louisville, Kentucky. KYTC and the Indiana DOT conducted extensive research to determine the effects of tolling and the proposed electronic tolling system, RiverLink, on both bridges on the population of the Ohio River region, including low-income and other disadvantaged communities. In fulfillment of FHWA’s Revised Record of Decision, Indiana DOT and KYTC prepared an Assessment of the Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations (2015). That study concluded that all groups would benefit from the improved traffic flow and the enhanced transit service across the river that the tolling would provide, and that no group would be dis- proportionately affected by the toll costs. FHWA concurred with this assertion, provided the agencies implemented their proposed tolling mitigation strategies (Collins and Marquis, 2015). As mitigation strategies, tolling policies were developed to address project impacts on affected low-income and minority communities nearby. The agencies gathered information and evalu- ated the various strategies regarding EJ with the tolling policy and published a draft assessment of the economic impacts of tolls and various mitigation measures in 2013 (Indiana DOT and KYTC, 2013). After an extensive public outreach process, several mitigation measures were pro- posed for implementation (Indiana DOT and KYTC, 2015). The tolling mitigation plan makes free transponders available to users who open a RiverLink account. Drivers can obtain transponders from various locations in the community (Figure 2). Unlike purchased transponders, the free transponders are incompatible with E-ZPass tolling sys- tems in other regions of the country. The mitigation plan also concluded that a $20 minimum account deposit requirement reasonably balanced the objective of making the amount acces- sible to most drivers while minimizing the number of drivers who would overdraw their prepaid accounts (Indiana DOT and KYTC, 2015). Toll rates are fixed but vary by driver’s travel patterns Figure 2. RiverLink local and E-ZPass transponders.

Recycling tolling Revenue through transit Investment and low-Income Assistance as Forms of Mitigation 247 and transponder ownership. Thus, a standard toll for cars with transponders is $2, while com- muters exceeding 40 crossings in one month receive a $1 discount per trip, retroactively to include all trips made that month. A premium toll of $3 or $4, respectively, is charged to drivers with- out transponders who do and do not pre-register their license plate numbers with the state (RiverLink, 2015). Agencies estimate that by receiving this discount, commuters crossing the bridges twice a day, Monday through Friday, would save $960 per year, which is 8 percent of the income of a person living at the 2013 federal poverty level (Indiana DOT and KYTC, 2015). To address barriers to transponder usage that may disproportionately affect EJ communities, the mitigation plan allows for cash replenishment options at select convenient locations and mobile sites and online account maintenance. The agencies also planned to allow drivers to man- age multiple transponders under just one account to ease fees and time associated with account maintenance (Indiana DOT and KYTC, 2015). To promote transit as a form of mitigation, capital improvements were also authorized for the Transit Authority of River City along with some complementary policy initiatives, including: • Expansion of park-and-ride facilities; • Introduction of new buses for shuttle services during construction; • Promotion of vanpools; • Rehabilitation and acquisition of additional facilities to accommodate existing buses; • Construction of new bus stops and consolidation of others; • Introduction of free downtown circulator buses, which will connect to some EJ communities; • Introduction of a new express bus service between Louisville, Kentucky, and Sellersburg, Indiana; • Introduction of a public awareness campaign of transit service; • Development of an EJ community outreach plan; and • Establishment tolling exemptions for transit vehicles Additionally, the Indiana DOT and KYTC agreed to monitor conditions in EJ communi- ties after construction and “incorporate practicable measures for minimizing impacts of tolling on low-income and minority communities” (FHWA, 2012). Given that 55 percent of Transit Authority of River City riders report income under $25,000, KYTC and Indiana DOT asserted that capital improvements would lead to enhanced bus services that would, if anything, dis- proportionately benefit EJ communities (Indiana DOT and KYTC, 2015). During the course of evaluation of the mitigation plan, Indiana DOT and KYTC considered other mitigation measures that they eventually deemed not practicable due to administrative costs, lost revenue, and difficulty of enforcement. As shown in Table 2, the costs of two such proposals—a one-time credit and a toll rate discount—were found to vary greatly depending on the eligibility standards (i.e., whether individuals were eligible based on income alone or where they live and the level of the discount) (Indiana DOT and KYTC, 2015). Name of Program Capital Cost (Total) Operations Cost (Yearly) Impact on Toll Revenue Estimated Total Cost (First Year) One-time credit $1.7 million $487,000 $22,000– $206,000 $2.2 million– $2.4 million Ongoing percentage toll rate discount $7.5 million $737,000 $6.5 million–$110 million $39.5 million– $143 million Source: Indiana DOT and KYTC, 2015 Table 2. Concept estimate of potential mitigation program costs.

248 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Example 5: Elizabeth River Tunnels, Hampton Roads Area, Virginia. The Elizabeth River Tunnels project consisted of constructing a new tunnel alongside the Midtown Tunnel between the cities of Norfolk and Portsmouth, improving the existing Midtown Tunnel and Downtown Tunnel, and extending the Martin Luther King Freeway in Portsmouth (Figure 3). The new and existing tunnels were tolled in order to fund the new construction. The project was made possible through a public-private partnership between Virginia DOT and the contractor, Elizabeth River Crossings (ERC), currently operating the tolled facilities. Under the Comprehensive Agreement signed by both parties in 2011, Virginia DOT will retain ownership of the facilities while ERC agrees to manage and maintain them for a concession term on 58 years. Virginia’s governor announced an agreement in July 2015 to ensure that no tolls would be collected on the Martin Luther King extension as a means to reduce the financial burden borne by low-income residents in Norfolk and Portsmouth. To avoid charging tolls on the Martin Luther King extension, the state agreed to transfer $78 million, previously set aside for a different project in southeast Virginia. With the agreement, motorists could enter or exit the Martin Luther King extension in a variety of ways and not get charged a toll (through VA 164, I-264, US 58, and the High Street) for making trips to and from Portsmouth. Trips continuing onward through the extension on US 58, however, are tolled. In order to discourage drivers from circumventing the tolling system by exiting at the High Street ramps and then re-entering, the system assesses a toll to anyone who re-enters the freeway within 10 minutes of exiting at the High Street ramps (Virginia DOT and ERC, 2011). ERC also agreed in the Comprehensive Agreement to make payments of $500,000 a year for 10 years beginning in 2017 to help offset the cost of tolls in the Midtown and Downtown tunnels for those toll users who are the most financially stressed. Virginia DOT assembled a Toll Relief Steering Committee, a voluntary advisory group, to provide guidance in the development of the Toll Relief Program and its roll-out to the public. The committee included representatives from state and local agencies and non-profit organizations. Eligible applicants for toll relief must be residents from Norfolk or Portsmouth, earn $30,000 or less, and have a Virginia E-ZPass account. An in-person visit to one of two E-ZPass customer Figure 3. Hampton Roads improvements.

Recycling tolling Revenue through transit Investment and low-Income Assistance as Forms of Mitigation 249 service centers in Norfolk or Portsmouth is required to provide proof of residency and income eligibility. According to the agreement, the relief funds for the program are to be renewed annually and qualified participants will be expected to re-verify their income and residence each year. Under the initial terms of the Toll Relief Program launched in December 2016, enrolled partici- pants were expected to receive toll discounts after taking eight one-way trips monthly through the Elizabeth River Tunnels. Enrolled participants receive a rebate on the previous eight trips, and a 75-cent discount per trip, regardless of the vehicle class, with no limit on the number of discount trips in a month. The discounted toll price would offer a savings off the passenger vehicle toll of $1.95 during the peak period and $1.65 during the off-peak for E-ZPass customers. Moreover, passenger vehicles without E-ZPass would face considerably higher tolls for plate transactions depending on whether the plate is pre-registered ($3.60/peak–$3.30/off-peak) or not registered ($5.25/peak–$4.95/off-peak). Under the terms of the agreement, ERC agreed to set aside $500,000 annually for toll relief with any unused funds to roll over to the next calendar year. If funds are depleted before the end of the calendar year, then toll relief payments are to be suspended until the next calendar year. In April 2017, the toll benefit increased from 75 cents to $1 because of a one-time additional contribution from the operators of the Elizabeth River Tunnels. To further offset toll costs, Virginia DOT has paid ERC a lump sum of $82.5 million, using primarily FHWA grants, to reduce toll rates from February 2014 to December 2016 as the tolling project is in its initial phases (Virginia DOT and ERC, 2011). Virginia DOT also seeks to address the extraordinary and potentially punishing costs of unpaid toll violations, which may be disproportionately borne by low-income households, by placing a cap on the total potential cost of a toll violation. Thus, the amount that ERC can charge an indi- vidual for a proceeding in court for a toll violation, plus fees and civil penalties, cannot exceed $2,200 (Virginia DOT and ERC, 2011). This is noteworthy, in part, because it is significantly lower than the general cap of $21,300 placed by the state for all projects (Forster, 2015). In this way, it could be argued that the agency is addressing equity issues from tolling burdens that would otherwise occur without the cap. Virginia DOT will also be working in collaboration with ERC and the community to better inform and educate the region about the E-ZPass program, making tolling as convenient as possible for motorists (Virginia DOT, 2015). Additionally, various transit improvements will provide alternatives for crossing the bridges and maximizing capacity. These improvements include: • ERC will provide Hampton Roads Transit with an annual $2.1 million subsidy for the life of the 58-year concession period in which ERC is operating toll facilities. • Hampton Roads Transit will increase bus frequency between Norfolk and Portsmouth, purchasing seven new buses for use on three routes. • The Elizabeth River Ferry will begin service 90 minutes earlier on weekdays to accommodate more commuters (Elizabeth River Tunnels, 2012). • Virginia’s Department of Rail and Public Transportation will work with Hampton Roads Transit to make certain the state’s subsidy will be used specifically to move more people with fewer cars through the tunnels (Virginia DOT, 2015). Example 6: I-95 Express Toll Lanes, Miami, Florida. The Florida DOT opened the I-95 toll- ing project in 2008 at a cost of $122 million as part of the Miami Urban Partnership Agreement. Phase 1 starts north of Miami around I-195 and continues farther north for 10 miles. Phase 2, currently under construction, will extend north again to the Fort Lauderdale area (Florida DOT,

250 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox 2014). Registered vanpools, carpools, registered hybrid cars, and all transit vehicles have always been exempt from tolling in the express lanes and the mitigation plan called for a multitude of transit improvements and concessions (Reich and Davis, 2010). In 2010, three new BRT routes were added on I-95, an increase from the one express route that had previously been operating as part of Phase 1. The routes, collectively branded as I-95 Express, were introduced to improve passenger throughput in the corridor and improve the experi- ence for transit-reliant residents (Florida DOT, 2013). Improvements to the existing service included expanded park-and-ride capacity and facility improvements; upgrades to stationary transit infrastructure, such as bus shelters, accessibility features, and amenities; and Intelligent Transportation Systems features such as real-time arrival information and transit signal priority (National Bus Rapid Transit Institute et al., 2011). Example 7: Legacy Tolling Systems. Revenue recycling has been in practice for decades, in the form of either transit or highway investments. Most of the major turnpike projects implemented in the 1950s in the United States paid off their original debt by the mid-1980s, but none opted to end tolling. They recognized their systems’ potential to help finance new projects or subsidize exist- ing modes. For example, New York’s MTA uses surplus revenue from its tunnel and bridge tolls to subsidize its subway system (HNTB, 2015). Similarly, the Port Authority of New York–New Jersey’s Tunnels, Bridges, and Terminals operations produce a surplus that is used, in part, to subsidize the Port Authority Trans-Hudson (PATH) rail service and its capital plan. What Are Its Limitations? Mitigation plans involving revenue recycling face political, legal, and practical limitations and challenges. Agencies may experience resistance to the idea of using highway tolls for multi- modal transportation solutions, especially in less transit-rich areas. Depending on the tolling program, they may not have the legal authority to allocate revenues for multimodal transporta- tion, including transit, pedestrian, and bicycling projects. In some states, funds acquired from toll collection may only be used on the designated highway or bridge facility, or for specific authorized purposes. For example, Washington State restricts the use of toll revenue to cover operating costs, to meet obligations for debt repayment and other funding obligations for proj- ects or operations on the designated toll facility, to provide for the operation of conveyances of people or goods, and to fund improvements to the toll facility. The State has additional restric- tions specific to each of its authorized toll facilities (Washington State, 2008). In some cases, private tollway investors include non-compete clauses in contracts that pro- hibit agencies from offering transit service that would compete with the tollway and threaten profits. The profit objective of private investors may conflict with the agency’s goals of providing efficient and equitable transportation in the corridor. States can manage these conflicting priori- ties by ensuring that transit is not included in non-compete clauses (Ankner, 2008). Toll discount, exemption, and rewards programs also have certain limitations. Building com- munity awareness of features available with a mitigation plan may require extensive outreach to ensure that the program is reaching those who need it the most. Agencies must also consider how programs are designed in order to avoid unintended consequences such as encouraging drivers to take more trips than usual. For example, once a commuter on the Ohio River Bridges exceeds 40 crossings in a calendar month using their transponder, their account is retroactively credited with a 50 percent discount on the first 40 crossings, or $40, effectively reducing their per-crossing rate from $2 to $1. Therefore, if a toll user has made more than 20 but fewer than 40 trips in a calendar month, they have been charged at least $40 in tolls, as each crossing costs $2. The com- muter could actually save money by making extra trips to reach the 40-trip threshold.

Recycling tolling Revenue through transit Investment and low-Income Assistance as Forms of Mitigation 251 Physical design of infrastructure also shapes and limits forms of mitigation for toll revenue recycling. For example, if BRT services are planned for HOT lanes, they may require physical roadway improvements such as direct access ramps to improve safe and efficient access to park- and-ride lots, transit stations, and stops. What Type of Resources and Costs Are Required? Outreach efforts to EJ communities require staff time as well as resources for advertising and on-the-ground public engagement. These costs can be offset by partnering with organizations that have already gained community trust and have knowledge of local populations. Agencies need to secure participation from retail stores or community locations to sell or distribute low- cost transponders in neighborhood locales. As highlighted earlier, any tolling plan that offers discounted rates to certain drivers will also incur additional administrative costs and may result in reduced tolling revenue. The costs of capital for infrastructure and transit operations improvements, such as additional bus facilities or increased service, will vary by the scale of the project but will be significant. Who Has Used It Successfully? The examples provided earlier demonstrate that states, cities, and MPOs from across the United States have found successful ways to coordinate with transit service providers to mitigate the EJ impacts of tolling, such as along HOT/managed lanes, while seeking to maximize passen- ger throughput in priced corridors. While each region will find unique challenges depending on its budget, political climate, and infrastructure, these examples show that options are available for recycling toll revenues to upgrade infrastructure, improve transit service, and offer forms of economic assistance to low-income commuters in regions where tolling is implemented. For more information, see the references presented in the resources section. Resources Ankner, W. 2008. Road Pricing, Tolling, Mega-Projects, and other Multi-Modal Opportunities: Strategically Posi- tioning Public Transportation. Retrieved from http://www.apta.com/resources/reportsandpublications/ Documents/WDA_APTA_Final%20Report_020708.pdf. Burris, M., Han, N., Geiselbrecht, T., Wood, N. 2015. I-30 Express Lane Survey Report. Prepared for NCTCOG. Cambridge Systematics. 2006. I-394 MnPASS Technical Evaluation. Prepared for the Minnesota Department of Transportation. Collins, B. and R. Marquis. 2015. Letter to Secretary Michael Hancock, Kentucky Transportation Cabinet and Commissioner Brandye Hendrickson, Indiana DOT. DeCorla-Souza, P. 2003. Clearing Existing Freeway Bottlenecks with Fast and Intertwined Regular Networks: Costs, Benefits, and Revenues. FHWA, Washington, D.C. Elizabeth River Tunnels. 2012. “Hampton Roads Transit Partnership.” Elizabeth River Tunnels. Retrieved from https://www.driveert.com/wp-content/uploads/2012/12/ERT_factsheets_HRT-Partnership_WEB.pdf. Elizabeth River Tunnels. 2016. “Cost and Funding.” Elizabeth River Tunnels. Retrieved from https://www. driveert.com/wp-content/uploads/2012/11/ERT_factsheets_CostandFunding_Current.pdf. FHWA. 2012. Louisville-Southern Indiana Ohio River Bridges Revised Record of Decision, FHWA-KY-SEIS-12- 01-F. Retrieved from http://updates.kyinbridges.com/wp-content/uploads/2012/06/LSIORB-Revised- ROD-w-signature-6-20-2012.pdf. Florida DOT. 2013. 95 Express Annual Report, Miami. Florida DOT. 2014. “Phase 1 & 2 Entry and Exits Diagram.” 95 Express. Retrieved from http://www.95express. com/pages/project-overview/express-lanes-entry-exit-points. Forster, D. 2015. “Source: State to Eliminate Toll on MLK Freeway Extension, at a Cost to Taxpayers of $78 Mil- lion.” Virginian-Pilot. Retrieved from http://pilotonline.com/news/local/transportation/source-state-to- eliminate-toll-on-mlk-freeway-extension-at/article_e2121f23-5537-59ba-834b-5e3781c0d81c.html. HNTB. 2015. Leveraging Tolls in the 21st Century: White Paper.

252 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Indiana DOT and KYTC. 2013. “States Seek Input On Ohio River Bridges Toll Plan.” Inside Indiana Business. Retrieved from http://www.insideindianabusiness.com/story/29814455/states-seek-input-on-ohio-river- bridges-toll-plan. Indiana DOT and KYTC. 2015. Assessment of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations. Prepared for U.S. DOT. Retrieved from http://kyinbridges.com/ wp-content/uploads/2015-04-24_EJ-Assessment-Plan.pdf. Lari, A. and Aultman, S. 2010. Study of Public Acceptance of Tolling with New Capacity and Credits: Concept of FAST Miles and FEE Lanes. Prepared for Minnesota DOT. Metro. 2009. Adopted Budget FY2010. Retrieved from https://d1akjheu06qp1r.cloudfront.net/about_us/finance/ images/budget_adopted_fy10.pdf. Metro. 2014a. “Congestion Reduction Demonstration Program [CRDP].” Metro ExpressLanes. Retrieved from https://www.metroexpresslanes.net/en/about/congestion.shtml. Metro. 2014b. “Transit Rewards Program [TRP].” Metro ExpressLanes. Retrieved from https://www.metroexpress lanes.net/en/about/transit.shtml. Metro. 2015a. Board Report: Metro ExpressLanes Round 2 Net Toll Revenue Allocation Guidelines. Retrieved from http://boardsecretary.metro.net/2015/10_October/RBM%20Item%2032.pdf. Metro. 2015b. LA County Congestion Reduction Program. Memorandum of the Metro Board, October 14. Retrieved from http://media.metro.net/board/Items/2015/02_february/20150218ahcritem18.pdf. Minnesota DOT. 2013. “MnPASS Express Lanes.” MnPASS. Retrieved from http://www.mnpass.org/394/. National Bus Rapid Transit Institute, Center for Urban Transportation Research, and the University of South Florida. 2011. Miami Urban Partnership Agreement Project Phase 1 Transit Evaluation Report. Prepared for U.S. DOT. Retrieved from https://transit.dot.gov/about/miami-upa-transit-evaluation-report-phase-1a. Network Public Affairs. 2010. Metro ExpressLanes Project: Draft Final Low-Income Assessment. Prepared for Los Angeles County Metropolitan Transit Authority. Parsons Brinckerhoff et al. 2005. HOT Credit Lanes Feasibility Study. Prepared for the Alameda County Conges- tion Management Agency. Retrieved from http://ibtta.org/sites/default/files/HOT%20Credit%20Lanes%20 Feasability%20Study.pdf. Reich, S. L. and Davis, J. L. 2010. Investigation of the Feasibility of Toll and Transit Agency Equity Sharing: White Paper. Prepared for Florida DOT. Retrieved from http://ibtta.org/sites/default/files/Bus-Toll%20lane%20 proposal%20FDOT%20THEA.pdf. RiverLink. 2015. “Tolling Rates.” RiverLink. Retrieved from http://riverlink.com/tolling-info/tolling-rates. Samuel, P. 1998. “Dynamic Pricing: San Diego.” Toll Roads News, 18 July. SANDAG. 2014. “Fact Sheet: About the Interstate 15 Express Lanes.” Keep San Diego Moving. Retrieved from keepsandiegomoving.com/Libraries/Lossan-doc/CW_T_I15_1A_factsheet_update_mgu_060614.sflb.ashx. SANDAG. 2015. Final Program Budget: FY 2016. Retrieved from http://www.sandag.org/uploads/publicationid/ publicationid_1957_19285.pdf. SANDAG. 2016. “South Bay Expressway: Promotions.” South Bay Expressway. Retrieved from http://sbxthe125. com/whats-happening/promotions. San Diego Metropolitan Transit System. 2014. Adopted Fiscal Year 2015 Budget. Virginia DOT. 2016. Virginia Toll Relief Program. Retrieved from http://www.virginiadot.org/tollrelief/. Virginia DOT. 2015. “Gov. McAuliffe Announces Deal Ensuring No Tolls on the MLK Freeway Extension in Portsmouth.” Retrieved from http://www.virginiadot.org/newsroom/statewide/2015/gov._mcauliffe_ announces_deal84307.asp. Virginia DOT and ERC. 2011. Comprehensive Agreement: Relating to the Downtown Tunnel/Midtown Tunnel/ Martin Luther King Freeway Extension Project. Retrieved from https://www.driveert.com/wp-content/ uploads/2013/08/Comprehensive-Agreement.pdf. Wong, P. 2015. Metro ExpressLanes: Net Toll Revenue Reinvestment. Presentation at 2015 ITS California Annual Meeting. Retrieved from http://www.itscalifornia.org/Content/AnnualMeetings/2015/Presentations/ TS11-4-MTA-NetTollRevenueReinvestment.pdf.

253 What Is It? Public records have been used, where available, to examine the usage patterns and income distribution of users of existing tolled facilities. These studies have also been performed to analyze the distribution of tolling burdens and benefits by income level and geographic area. Why Is It Effective in Environmental Justice Analysis? Several tools have been used to explore equity and EJ issues in tolling, but these tools often rely on opinions solicited in surveys or drawn from transportation modeling output. The use of actual data from tolling facilities to assess the relationship between income and the use of tolling facilities is less common. What Are Some Techniques for Implementing This Tool? Several academic, applied research, and advocacy organizations have tried to use available agency records to analyze the equity and EJ effects of various tolling systems. The case examples presented show how avail- able records data can be used to analyze usage patterns, the demograph- ics of users, and the methods and measures used to assess disparities in usage. The research itself may be conducted in cooperation with the tolling authority or governing agency or as an independent evaluation. The research may be used to monitor the equity dimension of existing operations or advocate for mitigation solutions to ensure mobility and access for low-income or other disadvantaged populations. Example 1: New Jersey Turnpike and Garden State Parkway, New Jersey. Peters and Gordon (2008) examined New Jersey’s exist- ing highway tolling systems to present a case study of methods for mea- suring equity in public services. Working with a news organization that had filed a Freedom of Information Law request, the researchers were provided with E-ZPass toll collection data for users of the New Jersey Turnpike and the Garden State Parkway for a 1-year period (see text box, Preparing the New Jersey Turnpike Dataset for Analysis). t o o l 1 4 Examining Spatial Patterns and Distribution of Users on Existing Tolling Facilities Framework Step • Conduct Post-Implementation Monitoring Stages in Decision-Making • Implementation Tools and Techniques • Electronic Toll Account Data • GIS Mapping • Toll Burden Metrics • Performance Measures • Gini Coefficient • Lorenz Curves • Correlation Coefficients • Before-After Evaluations Affected Populations • Low-Income • Minority Examples Featured • New Jersey Turnpike & Garden State Parkway, Tolling Systems • Metropolitan Transportation Authority, New York City, Bridges and Tunnels • Southern Environmental Law Center, Metro Atlanta, I-85 High-Occupancy Toll Lanes • U.S. DOT, Urban Partnership Agreement and Congestion Reduction Demonstration Sites

254 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox The toll charges from the electronic toll transactions dataset were summarized by zip code and analyzed at various geographic scales. Two measures of equity burden were reported: (1) per capita and (2) share of income spent on toll. As shown in Figure 1, the general relationship between income and per capita toll burden varies, although the researchers found a significant number of places with low incomes and high toll burdens (Peters and Gordon, 2008). Results showed that when measuring per capita toll burden, the highest income zip codes (i.e., top 34 zip codes with greater than $50,000 per capita income) exhibited higher toll burdens, with an average burden of $23.10 per capita compared to $17.60 per capita in the lowest income zip codes (i.e., lowest 21 zip codes with less than $15,000 per capita income). The researchers observed that the toll burden for the lower-income per capita zip codes might be underreported due to barriers to using E-ZPass accounts (e.g., absence of credit cards or debit cards, less stable mailing addresses). The selection of equity measures frames the analysis and potentially can shift the point of emphasis, adding new points of consideration and complexity. As shown in Figure 2, once the toll burden was measured as a percentage of individual income spent on tolls, rather than the Preparing the New Jersey Turnpike Dataset for Analysis The researchers obtained electronic transaction data for the New Jersey Turnpike and the Garden State Parkway, covering a 1-year period ending in June 2007, with information for more than 262 million transactions totaling $200 million in tolls. The data were provided for 704 zip codes across the state with an E-ZPass account. The data were converted into Census zip codes (ZCTA) to integrate Census socio- economic data; approximately 1.5 percent of the transactions data could not be geographically assigned using this method. Recognizing that commercial activities on the roads would distort per capita data for locations with high concentrations of regional commercial activity, the researchers omitted data from select municipalities, including Trenton, the state capital; Teterboro, the site of a small airport for private aircraft with few residents; and Hainesport. These cities’ per capita toll burdens were outliers, suggesting high levels of commercial traffic that should not be considered when calculating toll burdens for individual residents. Source: Peters and Gordon, 2008 Figure 1. Relationship between toll burden and per capita income.

Examining Spatial Patterns and Distribution of Users on Existing tolling Facilities 255 absolute amount spent per capita, the higher-income zip codes and towns trended toward lower toll burdens than the lower-income areas as shown in the Figure 2 (Peters and Gordon, 2008). Although not shown here, locational equity of toll burden was also mapped at the zip code level. The researchers found that the zip codes with limited access to alternative highway services tended to experience greater toll burdens. Additionally, several of the higher-income areas of north-central New Jersey showed relatively lower levels of toll burden than the lower-income areas that were farther from the core employment nodes. When data were aggregated on a county level, rather than on a zip code level, higher toll burdens were not correlated with higher average incomes. Rather, they were more closely associated with the lack of access to alternative modes of travel, including commuter rail and untolled highways. Those counties located closest to tolled facilities and without free or less expensive alternative routes paid more per capita in tolls because drivers had few other choices (Peters and Gordon, 2008). In their conclusions, the researchers acknowledged that the measures used were “fairly crude” and the geographic areas were “relatively broad,” but they concluded that the gross burden of toll roads in New Jersey appeared to fall disproportionately on lower-income communities when measured using ZIP code data. Similar patterns were observed using data at the county level but the patterns were not found to be as consistently regressive. The presence of mass transit systems may contribute to lowering the burden of tolls in some high-income areas and in some low- income areas. The researchers also noted that the approach did not incorporate transportation network or travel modeling data; future work might benefit from further attention to explanatory factors such as modal choice and the relative proximity of alternative non-tolled roads to reach market area destinations. Example 2: Metropolitan Transportation Authority Bridges and Tunnels, New York City. Gordon and Peters (2011) analyzed survey data from an origin-destination survey to assess the distribution of income of toll facility users of eight priced bridges and two priced tunnels in and between the five boroughs of New York City (see text box, Using the Triborough Bridge and Tunnel Authority (TBTA) Origin-Destination Survey Data). The researchers applied tools for equity measurement—Lorenz curves and Gini coefficients—as a means for assessing the equity of the toll burden at the priced facilities (see Figure 3 and text box, Metrics to Measure Distributional Equity). Using these equity measures, the income distribution among toll facility users was assessed and compared to the general populations of the “market areas” for each facility. Market areas were mapped and defined as regions of increasing distances from the tolled facilities (i.e., within 5, 10, and 15 miles). The differences between the toll facility users and the general population with respect to the “equality line” embedded in the Lorenz curve were then examined. Figure 2. Relationship between percentage of income spent on tolls and per capita income.

256 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox The Queens–Midtown Tunnel was presented as a representative example to illustrate how the analytical approach and equity measure can be used to compare the income distribution of toll users with the general population in the market area, using the surrounding affected community or other background population as a benchmark (see Table 1). In this example, the Queens– Midtown Tunnel users had significantly higher incomes and a more unequal distribution of incomes (Gini coefficient of 0.78) than the general population for each of the defined market areas (Gini coefficients between 0.45 and 0.47) and the benchmarked regions of New York State or the United States. An illustration of how the Lorenz curves for Queens–Midtown Tunnel users significantly differed from the general populations is shown in Figures 4 and 5. Using the Triborough Bridge and Tunnel Authority (TBTA) Origin-Destination Survey Data The researchers, through a Freedom of Information Request, obtained an origin- destination survey conducted by the TBTA (also known as MTA Bridges in Tunnel) in October 2004. The survey dataset contained more than 60,000 observations of passenger car usage on nine TBTA facilities in New York City. The survey contained user characteristics, including residence, income class, and trip purpose, which were used to conduct an equity analysis of toll burden among users. “Market areas” were defined as diameters of increasing distance (i.e., 5-miles, 10-miles and 15-miles) around each TBTA facility. The user’s home zip code was matched to the market areas for each facility along with the user’s income class. The income class of users was then sorted for each market area. U.S. Census data were compiled for each market area to compare a social profile (income, race, household size) of the market area with the toll facility users and to support the measurement of differences in distribution. Source: Gordon and Peters, 2011 Figure 3. The Lorenz Curve provides an illustration that expresses inequality.

Examining Spatial Patterns and Distribution of Users on Existing tolling Facilities 257 Metrics to Measure Distributional Equity Peters and Gordon used two metrics of equality of distribution that complement each other—the Lorenz curve and the Gini coefficient. The Lorenz Curve. The Lorenz curve is a graph that measures the proportion of a population against the proportion of some total benefit or cost that that proportion of the population bears, such as income or toll burden (Gordon and Peters, 2011). In a perfectly equal situation, every member of a given population would earn the same income or pay the same amount in tolls, so each individual’s proportion of the total benefit of the cost would be equal. In such an example, 10 percent of the population would bear 10 percent of the cost, 50 percent of the population would bear 50 percent of the cost, and so on. This situation, represented by the equation y = x and expressed as a straight line called the equality line, is used for comparison’s sake with the Lorenz curve. In an unequal situation, there would be a gap between the Lorenz curve and the equality line. The Lorenz curve expresses inequality but is limited because it does not address the variables that contribute to it. Additionally, comparing two different unequal distributions to each other, such as toll burden and income, may leave room for errors in interpretation (Gordon and Peters, 2011). The Lorenz curve is most commonly used to chart distributions of income or wealth, but is increasingly being applied to other resources or costs. Gini Coefficient. The Gini coefficient, also called the Gini ratio or Gini index, expresses the Lorenz curve as a number between 0 and 1, with 0 representing complete equality and 1 representing complete inequality. The coefficient is calculated by measuring the area between the equality line and the actual Lorenz curve on a graph, and then dividing it by the total area underneath the y = x line (Gordon and Peters, 2011). The Gini coefficient is most commonly used to describe income inequality in a country or city, but can also be used to measure equality in the proportion of total toll burdens borne by all toll users. Source: Gordon and Peters, 2011 Table 1. Impact of distance from Queens–Midtown Tunnel on Market Area, New York State, and U.S. Demographics. 5-miles 10-miles 15-miles QMT User NYS HH U.S. HH Households 1,155,213 2,739,404 3,900,778 N/A 7,056,860 105,480,101 Mean HH Income $68,003 $55,908 $57,051 $117,591 $72,583 $67,626 Median HH Income $42,500 $37,500 $37,500 $112,500 $67,857 $66,111 Gini Coefficient 0.4731 0.4656 0.4564 0.7769 0.4985 0.4689 Note: HH is a household; NYS is New York State; QMT is Queens–Midtown Tunnel. Source: Gordon and Peters, 2011

258 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Market Power and Toll Burden. The researchers also used the equity measures to explore the relative “market power” of MTA tolled facilities to impose a higher toll burden on its users. According to the researchers’ hypothesis, toll roads operating in a context of few alternative routes or alternative modes nearby would be expected to have more ability to charge higher tolls and have users bear those burdens. They would also have larger market areas. To explore the potential effect of market power on social equity, the MTA facilities were first characterized by their relative level of competitiveness. The researchers determined whether each tolled facility enjoyed a “monopoly” or “low-competitive position” (i.e., it was the only viable option for most travelers in the area to cross the river) or was “competitive” (i.e., there were other feasible and less expensive or free travel options nearby). Reported price elasticities for each facility as well as the location and quality of each of the facilities were used to support the characterization. By this definition, the Verrazano-Narrows Bridge crossing between Staten Island and Brooklyn was deemed to be a monopoly facility and, to explore the market power concept, the demograph- ics of its users and its surrounding market area were compared with the more competitive Queens–Midtown Tunnel facility (see Figure 5). The demographics of the Verrazano-Narrows Bridge facility users and its surrounding area were found to be less different than a similar comparison of facility users and surrounding area on the Queens–Midtown Tunnel, although Figure 4. The Lorenz Curve for QMT users is compared with various market areas, or distances. Figure 5. The differences in the distribution of VNB and QMT facility users’ income and their market areas are compared.

Examining Spatial Patterns and Distribution of Users on Existing Tolling Facilities 259 the Verrazano-Narrows Bridge bridge user demographics were still more affluent than the sur- rounding communities. However, Verrazano-Narrows Bridge users had fewer travel choices than Queens–Midtown Tunnel users. While income inequality among Verrazano-Narrows Bridge users (Gini coefficient of 0.6046) was lower than Queens–Midtown Tunnel users (Gini coefficient of 0.7769), Verrazano-Narrows Bridges users had lower average incomes and had to bear more of the costs to use the facility. Using the equity measures, the researchers concluded that income is a stronger determinant of whether or not someone will use the tolled facilities in a competitive environment than a non- competitive one, because low-income users in competitive environments have options other than tolled roads (Gordon and Peters, 2011). However, when there are few alternatives, the proximity of the users to the toll collection point is a more important determinant of toll burden. In their concluding remarks, the researchers suggested that the equity metrics could inform pricing policies for the metropolitan network of facility crossings. Greater sensitivity to “vertical equity” could be instituted by matching segments of the facility crossings network to the rela- tive income of the average users of the facility. Thus, toll rates could be set to ensure that high- median-income facilities with transit options are set higher than low-median-income facilities with few transit alternatives (Gordon and Peters, 2011). Example 3: I-85 High-Occupancy Toll Lanes, Metro Atlanta. In October 2011, the I-85 HOT lane project in metro Atlanta converted an existing carpool lane to a managed lane, becoming the first operating managed lane in a region planning a network of such lanes. With concerns that little research had been done to examine the role of income in driver choice to use the priced lanes and its equity impacts, the Southern Environmental Law Center (SELC) published a report, A Highway for All? Economic Use Patterns for Atlanta’s HOT Lanes (2013) to assess if the facilities were being used as “Lexus Lanes,” as some critics alleged. At the time of its preparation, pre-and post-implementation user surveys were planned for the corridor (see case example, “Conducting Pre- and Post-Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I-85 Corridor”), as part of FHWA’s performance assessment of its funded CRD projects, but the equity analysis results from the survey had not been released. The research team obtained an electronic transactions dataset for the I-85 HOT lanes col- lected by the Georgia State Road and Tollway Authority (see text box, Preparing the Georgia State Road and Tollway Authority Electronic Transactions Dataset for Analysis). The SELC analyzed income and per capita use of HOT lanes for the zip codes in Metro Atlanta. The research team calculated a Pearson Correlation Coefficient (R) to analyze the strength of the linear relationship between the two variables. The analysis found a statistically significant positive (R = 0.44) cor- relation between the median income for a zip code and per capita HOT lane use among the zip code’s residents. The moderately strong correlation suggested, according to the SELC researchers, that there are other variables affecting HOT lane usage, most of which the SELC specification and analysis did not capture. The researchers also noted that these findings were consistent with other HOT lane studies such as the I-394 MnPass Express Lane project and the State Route 91 Value- Priced Express Lane project (SELC 2013). Role of Distance in HOT Lane Use. Researchers noted that prior HOT lane studies had found trip distance to influence the rate of HOT lane use and trip price (Patterson and Levinson, 2008). The SELC researchers may have expected that longer trips would correlate with higher incomes of users, but the regional demographic patterns of housing affordability and job loca- tion shape the operating context of the corridor and influence usage patterns. Therefore, SELC researchers mapped the average trip distance, median income, and the average trip cost of the I-85 corridor (see text box, Measuring the Strength of Relationship between Median Income and Per Capita Usage and Figure 6). The analysis and mapping showed how this prevailing context

260 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Preparing the Georgia State Road and Tollway Authority Electronic Transactions Dataset for Analysis The Southern Environmental Law Center (SELC) research team obtained electronic transaction data for the I-85 HOT lanes collected by the Georgia State Road and Tollway Authority that covered a 4-month period in fall 2012. The dataset included the date, duration, average speed, toll status, cost, and zip code for more than 1.6 million discrete transactions. Using this information, researchers could derive trip distance and toll cost per mile. Population and median income data for the relevant zip codes were gathered from the U.S. Census Bureau’s ACS. The transaction dataset was then filtered to analyze the core question of whether household income influences HOT lane use. Transaction records that were untolled, such as those with three-person carpools, registered transit vehicles, motorcycles and emergency vehicles, or that contained incomplete information were removed. Out-of-state zip codes or post office boxes were also removed. About 1.2 million transactions remained. To avoid weighting zip codes that contribute only a small number of transactions equally with those that account for the majority of trips, researchers excluded all zip codes that failed to exceed a minimum rate of per capita use. Only zip codes whose per capita toll lane usage over the 4-month study period was greater than 0.3 trips per capita (i.e., 300 trips per 1,000 residents) were included in the analysis. This screening left 1 million transactions but reduced the number of zip codes from more than 1,000 to 31 and accounted for about 78 percent of all transactions. Source: SELC, 2013 Measuring the Strength of Relationship between Median Income and Per Capita Usage The Southern Environmental Law Center (SELC) research team sought to measure the strength of the relationship between median household income and per capita usage of the managed lanes project using the Pearson Correlation Coefficient (R). The Pearson Correlation Coefficient, a standard measure of strength of a linear relationship between two variables, would indicate a value of “1,” if there is a perfect correlation, and a score of “0,” if there is no correlation. Two other additional statistical measures, Probability (p-value) and Coefficient of Determination (R2), were also used to examine this relationship. The p-value provides evidence that the relationship between two variables reflects the hypothesis rather than random chance. Thus, a value of 0.05 indicates a 5 percent likelihood that the results are a correlation due to random chance. The R2 mea- sures how accurately a model fits the data. Source: SELC, 2013

Figure 6. Maps of (a) average trip distance, (b) median income, and (c) average trip cost were prepared to support analysis of the role of distance in HOT lane usage in the I-85 corridor. (a) (b) (c)

262 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox can complicate the role of distance in an equity analysis; in this case, the spatial patterns of median income revealed that some of more distant zip codes from the city center had lower median incomes but were likely to have longer and higher priced trips (SELC, 2013). To further consider the effect of distance, the researchers segmented trips into four trip cost categories: less than $4, $4–$4.99, $5–$5.99, and more than $6. Results were mixed. The SELC researchers found a positive relationship between income and HOT lane use for trips under $4 as well as for trips costing between $5 and $5.99, but found no such relationship for trips from $4–$4.99 and over $6. The unexpected results could be partially explained by the fact that 94 percent of the trips cost less than $4, so more data were available for that category. When data were normalized to account for cost per mile, there was also not a strong relationship, suggest- ing that trip choices on tolled facilities are more influenced by the total cost of the trip than by incremental changes in the cost per mile (SELC, 2013). In their recommendations, the SELC researchers acknowledged the limitations of the scope of their research as well as with the level of detail available from the transactions dataset. Ideally, for example, a more detailed transaction dataset would include usage reporting to the address level (in the absence of individual records data). Income data could be compiled from census tract or block group, which provides more details than the high-level zip code data. Researchers also identified several additional variables the research did not consider that could affect HOT lane usage, including HOT lane entry points, likely destinations, availability of alternate routes, and the availability of transit services. These additional variables would likely require the use of travel modeling, GIS data, and other tools. The dataset was also limited in terms of the time period of the study. The SELC researchers concluded their report with recommendations for ensuring greater equity, including revenue recycling for funding parallel transit service, limiting state funding for premium lanes, setting the carpool level threshold for access to the managed lane back to two-persons rather than three or more persons, and providing a minimum level of “free access” for all registered users (e.g., through toll credits). These mitigation measures were proposed based on the evidence that suggests that the facilities result in disproportionate impacts for low-income drivers. Example 4: National Evaluation of Congestion Pricing Strategies. U.S. DOT funded the national evaluation studies of six UPA and CRD projects that were funded in 2007 and 2008. The selected sites were in Atlanta, Los Angeles, Miami, Minnesota, San Francisco, and Seattle– Lake Washington. To address urban congestion, each site was tasked with implementing a com- prehensive policy response reflecting four essential components known as the “4 Ts”: (1) tolling (or congestion pricing), (2) enhanced transit services, (3) telecommuting and other transportation demand management (TDM) strategies, and (4) advanced technology. Equity, one of several dimensions for assessment as part of the national evaluation framework, was examined at five of the six UPA and CRD sites (Zimmerman et al., 2015). Several methods of data collection and analysis were used to examine potential equity concerns of the UPA and CRD projects and to assess whether the positive or adverse effects of the projects fall dispropor- tionately on different user groups or geographic areas. The evaluation framework compared conditions prior to the start of each site’s congestion pric- ing project with conditions a year later. In general, a year of “before” data was compared against a year of “after” data at each site, although in some cases data for longer or shorter periods of time were used (Zimmerman et al., 2015). Local agencies were responsible for collecting the operational data for national evaluation. Data for the equity analysis were obtained from a number of differ- ent sources (see text box, Organizing Data to Support “Before-After” Evaluation, Atlanta Region). The National Evaluation Plan sets forth several questions to be addressed in the equity analysis for individual sites. One of the questions pertains to how the HOT lane or express lane facilities

Examining Spatial Patterns and Distribution of Users on Existing tolling Facilities 263 may affect different transportation user groups drawing data from the congestion analysis. Accord- ing to the final “lessons learned” report, most users benefited as a result of the pricing improve- ments. Increased costs to tolled users were offset by faster, more reliable travel, and transit users typically experienced a faster, higher quality trip on the tolled facility (Zimmerman et al., 2015). In Atlanta, Seattle, and Los Angeles, travel conditions in the free facility (i.e., alternate route or general purpose lane) either did not change or speeds decreased slightly, while travel improved in Minneapolis’ general purpose lanes. In Seattle, where tolling was deployed for all lanes on the SR 520 Bridge, lower-income groups eliminated a greater proportion of trips across Lake Wash- ington than other income groups. Table 2 presents an example from the National Evaluation Plan study of the I-85 Corridor in the Atlanta region showing how the before-after comparison by transportation user group was presented to examine changes in travel time and travel costs. In this example, it can be observed that general purpose lanes experience slightly slower travel during the morning and evening peaks after the express lanes opened. The general purpose lane travelers, however, did not experience additional costs borne by the express lane travelers. Transit riders had travel time savings with no change in their out-of-pocket costs, which were lower; express lane toll payers enjoyed travel time savings in comparison to the before condition in the general purpose lane. Measuring equity in terms of changes in travel times and costs as borne by transportation user groups is informative as to the overall performance of the corridor in achieving goals for conges- tion reduction. However, this framing of equity only indirectly provides information for making an assessment of the burdens or benefits borne by low-income and minority populations in an EJ assessment. More focused consideration can be given through surveys of travel behavior and opinions toward the fairness of pricing systems afforded by surveys that are designed to enable Organizing Data to Support “Before-After” Evaluation, Atlanta Region The Atlanta Congestion Reduction Demonstration (CRD) projects include the conversion of lanes for high occupancy vehicles (HOV) on approximately 16 miles of I-85 to high-occupancy toll (HOT) lanes, known as express lanes, along with the expansion and enhancement of transit service in that corridor, including new and expanded park and ride lots. Equity, one of several dimensions for assessment as part of the national evaluation framework, was examined for the I-85 corridor drawing on data from several other analyses performed in the national evaluation. To understand how different transportation user groups were affected by the I-85 corridor projects, travel times and vehicle miles traveled (VMT) based on traf- fic sensor data were obtained from the congestion analysis and findings from the Volpe Household Travel Survey; average toll rates were compiled from tolling trans- actions data based on Peach Pass toll tags; transit analysis from data on ridership, results of an on-board transit survey, and transit rates; and carpooler survey results came from a travel demand model (TDM) analysis, among other sources. The data from those parts of the national evaluation were supplemented with socioeconomic data from the U.S. Census Bureau and vehicle operating costs from the American Automobile Association (AAA). Source: Zimmerman et al., 2014

264 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox segmentation and analysis by income or race and ethnicity group. Toward this objective, a more detailed discussion of how survey efforts were designed and conducted in Atlanta (and Seattle) are described in a separate case example, “Conducting Pre-and Post-Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I-85 Corridor.” In the final “lessons learned” report, equity impacts, or the perceptions of fairness, especially for low-income travelers, were acknowledged to be a key factor in promoting the acceptance of transportation projects involving the introduction of pricing. The final “lessons learned” report (Zimmerman et al., 2015) as well as the site-specific evaluation test plans and final evalu- ation studies can serve as a useful resource for identifying data sources, methods, and analytical approaches for the performance measures and how they may be integrated to support an equity assessment. The resources section contains links to several of these studies. Table 2. Comparison of travel time and costs per trip on the I-85 corridor by user group in the morning and evening peak travel periods, Atlanta region. User Group Mean Travel Timea Costsb Before After Before After General Purpose Lane Travelers a.m. peak 16.1 16.9 $7.03 $7.03 p.m. peak 16.1 17.8 $7.13 $7.13 tolling a.m. peak 14.1 13.8 $7.03 $7.03 p.m. peak 14.3 13.8 $7.13 $7.13 Transit Riders a.m. peak 14.1 13.8 $5.00 $5.00 p.m. peak 14.3 13.8 $5.00 $5.00 Express Lane Toll Payers a.m. peak N/A 13.8 N/A $10.98c p.m. peak N/A 13.8 N/A $ 9.40c a Mean travel time in minutes obtained from Congestion Analysis. Times are for 11.75 miles northbound and 11.56 miles southbound where traffic sensor data were judged to be useable. b Vehicle operating cost per mile of 60.8 cents per mile for average sedan driven 15,000 miles per year from 2013 AAA “Your Driving Costs” multiplied by the northbound and southbound miles used in the travel time calculations. Transit cost is based on the cash fare for Xpress bus trip from the Sugar Mills park-and-ride lot to downtown or midtown. 2012 costs were used for both before and after to enable cross-mode comparisons. c Vehicle operating cost per mile of $7.03 for 11.56 miles southbound and $7.13 for 11.75 miles northbound plus average toll of $3.95 southbound and $2.27 northbound for the maximum trip distance in September 2012. Source: Zimmerman et al., 2014 Carpoolers, HOV2+ before and HOV3+ after

Examining Spatial Patterns and Distribution of Users on Existing tolling Facilities 265 What Are Its Limitations? The researchers acknowledged several of the limitations of the individual case examples (men- tioned in the case example discussions). Electronic toll collection datasets potentially offer highly disaggregate data about trip-taking, but confidentiality in reporting may lead to data reporting for only high levels of geography (e.g., zip code) rather than finer levels of geography (e.g., address, block, block group, or TAZ levels). The electronic toll account records are not likely to contain household income for the account holder, requiring the analyst to rely on general demographic characteristics from Census zip code geographies to serve as a proxy for specific information of the account holder. In the New Jersey Turnpike example, technical considerations related to translating census boundaries to zip codes led to some data loss. Electronic toll records may contain incomplete locational information because they only include data from where the user has a reported billing address for the account. Depending on the system, a fraction of drivers may pay tolls in cash and no payment will be traceable to a bill- ing address. When cash replenishment of a toll account is an option, a notable proportion may come from low-income, unbanked local residents, and less complete reporting of a permanent residence may result. A portion of users may also be out-of-town visitors, users of rental cars or car share services, especially near major metropolitan areas where rates of auto ownership are lower. Thus, the data collection misses forms of travel when payments are not linked to accounts. The absence of data from these users may bias the results, especially where income analysis is concerned (Peters and Gordon, 2008). In the research design, limiting assumptions about the market or geographic area are generally required to examine the relevant segment of toll road users. Thus, a study of the metropolitan area or corridor will likely exclude some super-commuters or occasional commuters to limit the amount of extraneous data caused by residents who are infrequent users, if users at all, of the tolled facilities (Peters and Gordon, 2008; SELC, 2013). In seeking to assess the relationship between income and toll facility usage, sorting through relevant factors such as distance may inform usage patterns. However, most of the research dis- cussed here made only limited use of travel network data, toll access points, or transit related information as independent variables, but these intervening factors may also influence the analy- sis of income and toll usage. In the National Evaluation Plan example, the “lessons learned” report mentioned several challenges in effectuating a “before-after” analysis, particularly as it may relate to comparing sites and generalizing findings. The report notes that measuring and comparing change was complicated by variations in the amount and quality of data from different collection methods. For example, roadway sensors needed for measuring traffic volumes and speeds were sometimes absent or not working properly. Implicit in the method is the challenge of distinguishing the effect of the implementation strategy from exogenous factors such as local construction projects and, in this case, the deep nationwide economic recession beginning in 2008. Because the six sites varied in their implementation schedules over the 2008 and 2013 period, the impact of the recession was likely felt differently among the sites and reflected in the data. What Types of Resources and Costs Are Required? Making use of available records such as electronic toll collection data or prior origin-destination or user surveys to assess user demographics and patterns requires considerable time from researchers familiar with statistical analysis and research techniques. Additionally, while using existing administrative data is generally more comprehensive than collecting data from sur- veys, the data must still be acquired through FOIL requests or other means.

266 Assessing the Environmental Justice Effects of toll Implementation or Rate Changes: Guidebook and toolbox Who Has Used It Successfully? Various approaches for using available data to analyze user demographics are described in the four examples presented in this section. The discussions highlight the techniques that were employed along with some potential limitations. More information can be found by examining the references presented in the resources section. Resources FHWA. 2015. Congestion Pricing Website. Urban Partnership Agreements. Retrieved from http://ops.fhwa.dot.gov/ congestionpricing/urb_partner_agree.htm. FHWA. 2015. Congestion Pricing Website. Congestion Reduction Demonstration. Retrieved from http://ops.fhwa. dot.gov/congestionpricing/urb_partner_agree.htm. Gordon, C. and Peters, J. 2011. “Measuring Toll Burdens: Applying Lorenz Curves to a Detailed Data Set of Users of Metropolitan Transportation Bridges in New York City.” Transportation Research Record: Journal of the Transportation Research Board, No. 2221. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 96–103. Patterson, T. and Levison, D. 2008. “Lexus Lanes or Corolla Lanes? Spatial Use and Equity Patterns on the I-394 MnPass Lanes.” Retrieved from http://www.nexus.umn.edu/papers/MnPASSEquity.pdf. Peters, J. and Gordon, C. 2008. “Measuring the Equity Burden in Public Service Provision: The Case of New Jersey Toll Roads.” Economic Papers, Vol. 27, No. 4, pp. 381–392. Peters, J. and Kramer, J. 2012. “Just Who Should Pay for What? Vertical Equity, Transit Subsidy and Road Pricing: The Case of New York City.” Journal of Public Transportation, Vol. 15, pp. 117–136. SELC. 2013. A Highway for All? Economic Use Patterns for Atlanta’s HOT Lanes. Atlanta. Retrieved from https:// www.southernenvironment.org/uploads/publications/selcatlantahotlanereport.pdf. Sullivan, E. 2002. “State Route 91 Value-Priced Express Lanes: Updated Observations.” Transportation Research Record: Journal of the Transportation Research Board, No. 1821. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 37–42. Zimmerman, C., Klein, R., Schroeder, J., Pessaro, B., Burris, M., Turnbull, K., Joy, B., and Schreffler, E. 2015. Con- temporary Approaches to Congestion Pricing: Lessons Learned from the National Evaluation of Congestion Pricing Strategies at Six Sites. FHWA-JPO-2015-217. Prepared for U.S. DOT. Zimmerman, C., Schroeder, J., Anderson, M., Braschayko, P., Balke, K., Burris, M., Songchitruksa, P., Goodin, G., Sug Park, E. S., Wood, N. Pessaro, B., Saunoi-Sandgren, E., Shreffler, E., Joy, B. 2014. Atlanta Congestion Reduction Demonstration: National Evaluation Report. FHWA-JPO-14-152. Prepared for U.S. DOT.

s e c t i o n 2 Case Examples

C o n t e n t s 271 Synopses of Case Examples 275 Case Example 1 Conducting Citizen Panels to Explore Key Issues of Value Pricing, Minneapolis– St. Paul Region, Minnesota 281 Case Example 2 Using an Environmental Justice Index to Identify Affected Populations, Dallas–Fort Worth Metro Region 287 Case Example 3 Mobilizing a Local Liaison to Recruit Community Leaders for Survey, Louisville– Southern Indiana Ohio River Bridges Project 294 Case Example 4 Targeting Local Grocery Stores to Administer Community Surveys, Louisville–Southern Indiana Ohio River Bridges Project 303 Case Example 5 Analyzing, Mitigating, and Monitoring Impacts on Low-Income Commuters, I-10 and I-110 ExpressLanes, Los Angeles County 315 Case Example 6 Mitigating Reduced Access via Toll Credits, Dallas–Fort Worth Metro Region 318 Case Example 7 Selecting a Design Alternative to Mitigate the Adverse Effects of Tolls on a Low-Income Community, St. Johns River Crossing, Clay and St. Johns Counties, Florida 323 Case Example 8 Conducting Pre- and Post-Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I-85 Corridor

271 Conducting Citizen Panels to Explore Key Issues of Value Pricing, Minneapolis–St. Paul Region, Minnesota The Minnesota Department of Transportation (DOT), with the support of the Humphrey Institute of Public Affairs at the University of Minnesota, initiated several laudable public engagement strategies in the early policy development stages for instituting value pricing in the Minneapolis–St. Paul region. Among these, the Citizen Jury process stands out as an approach for fostering meaningful public involve- ment to gauge public opinion on transportation funding alternatives. The Citizen Jury process provides a deliberative process to inform and amplify the voice of average citizens in a public policy debate. While the process is time-intensive, the intricacies of pricing solutions are complex and require adequate attention to educate representative members of an affected population so they can make informed recommendations. Through the citizens’ jury process, decision-makers and the broader public learn what people really think once they have the opportunity to study an issue closely. The jury process for the Minnesota value pricing policy debate was well received even by its skeptics. The seriousness in which the jurors approached the problems of both increased conges- tion and insufficient financial resources was particularly noteworthy. While equity and Environ- mental Justice (EJ) considerations were not the sole focus of the process, the issues presented to jurors led to concerns about the effects of pricing strategies on low-income families. Of particular benefit to the policy development process was that jury participants developed recommendations for future considerations. Toolbox Step: Step 4—Scope Approach to Measure and Address Impacts Using an EJ Index to Identify Affected Populations, Dallas–Fort Worth Metro Region The North Central Texas Council of Governments (NCTCOG), the region’s metro- politan planning organization, conducted two levels of EJ analysis as part of its Met- ropolitan Transportation Plan (MTP), Mobility 2035. One analysis was included in the plan itself, and the other was included as a supplement to the Regional Tolling Analysis (RTA). The goal of the RTA was to evaluate proposed expansion projects on the priced facility network and to understand their impacts on land use, air quality, and EJ populations. Both methods are easy to understand, implement, and replicate. They use data sources that are com- monly available and can be adapted to any region. Both analyses identify places with a higher concentration of low-income and minority populations and consider the relative accessibility of Synopses of Case Examples

272 Assessing the environmental Justice effects of toll Implementation or Rate Changes: Guidebook and toolbox these places to jobs, retail, and other important destinations. They also consider how this relative accessibility may change based on proposed transportation network changes. The methodologies developed for the EJ analysis, which included both an EJ index and four separate EJ categories, arose because of the demographic characteristics of the Dallas–Fort Worth region. Close to 50 percent of the population in the metro area is a minority. The team realized that if it were to use the methodology that classified a census geography as an EJ area when the minority population was greater than 50 percent, most of the Dallas area would qualify. This would largely defeat the purpose of the analysis. With this in mind, the team developed the two approaches used in the study. The lesson learned in this case is that EJ analysis methodologies are not “one size fits all.” Instead, they require approaches that are sensitive to the subject region’s demographic patterns. Toolbox Step: Step 5—Conduct Impact Analysis and Measurement Mobilizing a Local Liaison to Recruit Community Leaders for Survey, Louisville–Southern Indiana Ohio River Bridges Project A market research consultant was selected to undertake a community leader survey on behalf of the Kentucky Transportation Cabinet and the Indiana DOT to better understand the impacts of tolling of two new bridges from the perspective of the EJ communities. The consultants first assembled a potential target contact list of leaders, largely drawing from the Mayor’s Violence Prevention Work Group Task- force on the Kentucky side and a list furnished by the Indiana DOT on the Indiana side. Based on these two sources, the market research consultants had an initial list of about 48 leaders, split between Kentucky (32 leaders) and Indiana (16 leaders), from business, civic, and religious orga- nizations and neighborhoods. To help reach the leaders in these communities, the consultants contacted a local liaison known to them and the communities. This individual was a well-known African American resident of Louisville—an ordained deacon in one of the largest churches in the country and a radio personality—with deep ties to and trusted relationships in the EJ communi- ties. Identifying a local liaison who was trusted within the EJ communities on the Kentucky side of the river was paramount to the level of success achieved in recruiting community leaders to participate in the community leader survey. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 6—Identify and Assess Mitigation Strategies Targeting Local Grocery Stores to Administer Community Surveys, Louisville–Southern Indiana Ohio River Bridges Project A market research consultant was hired to survey minorities, low-income, and/or minority and low-income community members on their attitudes toward tolling and their willingness to consider using transponders, toll- free alternatives, and transit after tolling is implemented on two Ohio River bridge crossings. Using demographic information obtained from the Supplemental Final Environmental Impact Statement, six block groups were identified as areas with high concentrations of EJ populations. The consultant decided that the most appropriate venue to find and engage members of these potentially affected communities was at grocery stores within these census block groups. Store owners at four grocery stores located in the target communities allowed the consultant to con- duct on-site interviews with their customers in exchange for the consultant purchasing gift cards/vouchers from their stores. The consultant was allowed to approach store customers, offering the card/voucher to those who would view a short video on a digital tablet and complete a survey. Being able to offer incentives was an invaluable tool given the short turnaround time for the survey.

synopses of Case examples 273 The approach was noteworthy for its use of surveys to target the views of low-income and minor- ity populations in the design of tolling policies and possible mitigation. The purpose and scope of the community member survey solicited opinions, but did not focus on current travel behavior characteristics such as the type and number of trips taken, routes used, mode chosen, or time-of-day. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 6—Identify and Assess Mitigation Strategies Analyzing, Mitigating, and Monitoring Impacts on Low-Income Commuters, I-10 and I-110 ExpressLanes, Los Angeles County Prepared for the Los Angeles Metropolitan Transportation Authority (Metro), the Low-Income Assessment Report describes several modeling and market research activities that were undertaken for this Congestion Reduction Demonstration Program that converted high-occupancy vehicles to high-occupancy tolling lanes, among other initiatives. This case example describes various quantitative and qual- itative methods that were conducted for this Congestion Reduction Demonstration Project that are generally replicable and could be undertaken by other sponsoring agencies prepared to com- prehensively analyze the distributional effects of their tolling programs on low-income popula- tions. Metro’s commitment to developing mitigation strategies through investments in transit services and toll credit programs suggests innovative ways forward for other sponsoring agencies seeking to reduce potential adverse effects on low-income users. In addition, Metro’s adoption of a performance evaluation framework with equity considerations ensures that the agency is transparent in sharing information with interested stakeholders and remains open to exploring ways to improve its programs to serve its low-income users. Similarly, the agency has instituted multiple rounds of market research to explore strategies to build awareness of its Transit Rewards Program and its Low-Income Assistance Plan and tweak policies to improve their effectiveness. Toolbox Steps: Step 6—Identify and Assess Mitigation Strategies, Step 8—Conduct Post- Implementation Monitoring Mitigating Reduced Access via Toll Credits, Dallas–Fort Worth Metro Region In 2014, the North Central Texas Council of Governments (NCTCOG) proposed the alignment for the Chisholm Trail Parkway. The new parkway would operate as a toll road, running near the existing Old Granbury Road right-of-way in Fort Worth, Texas. The proposed toll route would prevent through traffic on Old Granbury Road, which was the existing route used by residents of St. Francis Village, a retirement community with a large number of low-income households near the alignment. Alternative non- tolled routes would more than double the residents’ travel distance and travel time. One of the proposed methods of mitigating this adverse impact involved the construction of an overpass to allow Old Granbury Road to remain a through street, therefore supplying a reason- able non-tolled alternative. However, this option was determined to be cost prohibitive because it would cost millions of dollars more than the original project. Instead an intervention was pro- posed that involved using government funds for the purchase of prepaid toll tags for the residents. NCTCOG’s Regional Transportation Council agreed to sponsor North Texas Tollway Authority’s TollTags as a method to mitigate the community’s reduced access. Mitigation reduced the costs for residents of accessing the toll road and enabled them to avoid a free alternative that would have nearly doubled their travel time. The program cost for this form of mitigation was substantially less than what it would have been to build a new road to supply the residents with access to an alternative non-tolled route. Toolbox Step: Step 6—Identify and Assess Mitigation Strategies

274 Assessing the environmental Justice effects of toll Implementation or Rate Changes: Guidebook and toolbox Selecting a Design Alternative to Mitigate the Adverse Effects of Tolls on a Low-Income Community, St. Johns River Crossing, Clay and St. Johns Counties, Florida The Florida DOT decided to design and implement a limited-access tolled facility from Branan Field-Chaffee Road to the 9B interchange with I-95 and construct a new bridge over the St. Johns River between Clay and St. Johns Counties in Florida. Under the preferred alternative, the Shands Bridge, a non-tolled facility, would be replaced by a toll bridge without a nearby alternate non-tolled route and would result in an impact that would adversely affect the low-income population in the study area. Because the alternatives did not provide viable non-tolled alternate routes, disproportionately high and adverse effects were expected to be borne by low-income populations in the project area. Therefore, Florida DOT decided to eliminate the bridge toll for the preferred southern alter- natives to avoid imposing these adverse impacts on local low-income drivers. This case example describes how the avoidance solution was developed as a form of mitigation so that low-income neighborhoods and communities near the bridge could avoid tolls for routine bridge crossings for local trips. The toll avoidance mitigation solution minimized adverse economic impacts to local populations and preserved access to local jobs and services under a condition in which non- tolled alternatives were not available. Toolbox Step: Step 6—Identify and Assess Mitigation Strategies Conducting Pre- and Post-Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I-85 Corridor This case example reports on survey research conducted for the I-85 cor- ridor, northeast of Atlanta. FHWA commissioned the survey research to document how the express lanes changed corridor travel behavior and attitudes toward tolling. The researchers designed a two-stage household travel diary survey, with surveys administered in two waves, one before the express lanes opened and the second after the lanes were in operation. In this longitudinal survey design, known as a “panel survey,” researchers contacted the same set of households for both waves of surveying. Participating households included people using a variety of modes and routes in the I-85 corridor: people who drove in the corridor (on I-85 or a parallel road), vanpooled, or rode transit. The Atlanta survey project provided valuable data to assess both the travel impacts of the tolled lanes on low-income and minority travelers, as well their opinions toward tolling. The survey researchers were able to present findings on many issues relevant to equity analysis, including: transponder ownership, use of the tolled express lanes, carpooling in the HOV lanes versus express lanes, and attitudes toward the tolled lanes before and after they opened, as well as the changes in attitudes from the pre- and post-implementation surveys. In each case, the researchers could com- pare the results for low-income and higher-income travelers and for minority and White travelers. Toolbox Steps: Step 4—Scope Approach to Measure and Address Impacts, Step 5—Conduct Impact Analysis and Measurement, and Step 8—Conduct Post-Implementation Monitoring

275 Conducting Citizen Panels to Explore Key Issues of Value Pricing, Minneapolis–St. Paul Region, Minnesota Background For more than 20 years, Minnesota transportation policy leaders have worked diligently to address public perception related to imple- menting value pricing projects in the Twin Cities region. In the mid- 1990s the Minnesota Legislature directed the Minnesota Department of Transportation (DOT) and the Metropolitan Council to explore value pricing as an option to manage congestion and pay for infrastructure improvements. Despite some initial setbacks, the persistence of the project sponsors and the robust outreach and education culminated in the implementation of two managed lane projects (I-394 and I-35W) with a third project opening in 2015 on I-35E. Minnesota DOT, with the support of the Humphrey Institute of Public Affairs, initiated several laudable public engagement initiative strategies in the early policy development stages. Among these, the citi- zens’ jury process stands out as an approach for fostering meaningful public involvement to gauge public opinion on transportation funding alternatives. The citizens’ jury process, according to the Jefferson Center, the organization that coordinated its use for value pricing and tolling in the Twin Cities region, provides an “opportunity for citizens to learn about an issue, deliberate together with a diverse group of their peers, and develop well-informed solutions to challenging public issues.” Through the citizens’ jury process, decision-makers and the broader public learn what people really think when they have the opportunity to study an issue closely. In this case, during the early planning years, a group of selected individuals came together as jurors to consider testimony representative of all facets of value pricing. The process led to recommendations related to the following key questions: • Can value pricing (previously called congestion pricing) be an effective strategy to address present and impending problems of traffic congestion and provide stable financing for surface transportation improvements? • If not value pricing, then what alternatives may be acceptable? c a s e e x a m p l e 1 Framework Step • Scope Approach to Measure and Address Impacts Stages in Decision-Making • Policy and Planning Tools and Techniques • Citizens’ Panel • Education Affected Populations • Jurors—cross section of residents in terms of age, race, education, geographic location, and commuting patterns Participants • Minnesota DOT • Metropolitan Council • University of Minnesota, Humphrey Institute of Public Affairs • Jefferson Center

276 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox While this informed citizen panel process did not exclusively focus on issues of equity or EJ, themes and information relevant to equity and EJ considerations were presented and deliberated during the five-day event. How It Was Done The citizens’ jury process was only part of the outreach and education activities undertaken during the public policy development stage. Other activities included citizen focus groups, tele- phone and interactive video surveys, and interviews with key opinion leaders. The citizens’ jury was held in St. Paul, Minnesota, for 5 days in early June 1995. The effort was undertaken to gather input for the technical studies evaluating the advantages and disadvantages of imple- menting value pricing in the Twin Cities and to better understand the public’s perceptions of value pricing and measure support or opposition. Identifying Participants. Diverse and balanced representation of the affected community is one of the most important considerations in empaneling a citizen’s jury. In this case, the region of study was defined to include the 13-county region to encapsulate the anticipated growth of the Twin Cities commuter-shed. Five demographic variables are typically used to achieve a balanced jury, including age, race, education attainment, gender, and geographic location. However, a sixth variable, commuting status, was also included by adding a screening question in the selec- tion of prospective jurors: “Do you commute to work or school three or more times a week?” The frequency in which someone experienced traffic congestion thus became a factor in the selection of jurors. A random telephone survey was taken of the 13-county region to recruit participants. The jury participants ranged in age from 16 to 73 years with the average age being 40.5 years. Both blue and white collar job occupations were represented, including a heavy equipment oper- ator, professor, teacher, caretaker, lawn care business owner, engineer, and bartender, among others. There was also a minister and a student in the selected jury pool. Witnesses were summoned to provide expert testimony. Twenty-two witnesses presented testimony on different aspects of value pricing, including potential problems and recommenda- tions for potential solutions. The following list provides job titles of these witnesses: • Assistant Professor of Communications, University of California • Deputy Commissioner, Chief Financial Officer, Minnesota DOT • Manager of the Office of Transportation and Transit Development, Metropolitan Council • Former Juror, Congressional Citizens’ Jury on Welfare Reform • Senior Policy Planner, Metropolitan Council • Planning Analyst, Metropolitan Council • Vice President of Government Relations, Minnesota Trucking Association • Transportation Consultant, Former Deputy Director, Office of Planning, U.S. DOT • Democratic State Senator representing Minneapolis • Principal, Strgar-Roscoe-Fausch, Inc. • Senior Vice President, Wilbur Smith Associates • Former Commissioner of Transportation, Director of the Center for Transportation Studies • President and CEO, AAA, Minnesota • Republican State Senator representing Fergus Falls • Professor Emeritus of Economics, University of Minnesota • Transportation Systems Engineer, Department of Public Works, City of Minneapolis • Democratic State Senator representing St. Paul • Division Engineer, Minnesota DOT • Senior Associate, Pacific Rim Resources • Legislative Analyst, House Research Department

conducting citizen panels to explore Key Issues of Value pricing, minneapolis–st. paul Region, minnesota 277 • Assistant Division Engineer, Metropolitan Division, Minnesota DOT • Vice President, Hughes Transportation Management Process. The first day was devoted to background information. Neutral witnesses provided details on the existing transportation system and challenges ahead, including traffic growth, con- gestion consequences, and financial implications. At the end of the day, the jurors were divided into small groups to discuss the problems. Expert witnesses, including advocates and opponents of value pricing, addressed the jurors over the next 2 days. At the close of each day, the jurors broke into small groups to discuss what they had learned from the testimony. On day 4, the jurors convened in four groups where they were assigned one model of value pricing. Each group dis- cussed the pros and cons of each option, which included discussion of how the approach could be structured to meet congestion and financial goals. The models are described below: • Cordon line. A fee would be charged for entering the service area within the metropolitan urban service area. • HOV “buy in.” Single occupant vehicles can pay to use the HOV lane. • Spot or facility pricing. Pricing-specific locations or bridges with flat-rate or variable pricing. • Road pricing. Although not a value pricing alternative, road pricing, including a statewide mileage-based tax, was discussed. On day 5, the jurors organized into a focus group forum to present their recommendations and entertain questions from transportation professionals. Eight jurors who had expressed greater concern for congestion broke off into a small group and their views were recorded by staff observers in a minority report. The project culminated with the full group giving its final recommendations. Results. The jurors provided the following answers to the initial key questions. In response to the question related to the current congestion problems in the 13-county area, only eight jurors thought there was an existing congestion problem. However, 22 voted that there would be a problem in the future. All 24 jurors believed that funding was a problem, but 17 jurors did not believe that value pricing would solve both the congestion and funding problems. Jurors offered several reasons for why they believed value pricing would not prove effective in solving the congestion and revenue problems, including: • Would not change travel behavior (11 jurors), • Inefficient way to raise revenue (19 jurors), • Would benefit wealthier travelers and disproportionately hurt low-income users (15 jurors), • Revenues generated could be used in the wrong way (5 jurors), and • Would violate personal freedoms (5 jurors). Jurors were also asked if they would want to see a pilot project on congestion pricing, funded by the federal government, tried in some portions of the metropolitan area. Eighteen jurors were in favor of this pilot initiative. Jurors in favor of the pilot thought it would be worthwhile for determining whether it would be an effective remedy for addressing congestion, generating revenues, and measuring cost effectiveness. While most of the jurors were opposed to a value pricing solution, eight jurors were in favor of value pricing under certain conditions and filed a minority report. They were particularly interested in severely congested corridors such as I-35W and I-494. Most jurors preferred to build their own solutions, which included raising gas taxes according to the following provisions: • Ensure Minnesotans receive the same percentage of the gas tax they now receive. • Recommend a 5-cent increase that increases to 15 cents over a 6-year period.

278 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox • Additional revenue should be spent as follows: 60 percent transit, 25 percent repair and improve roads, and 15 percent on new roads. • Provisions to be adopted over a 10-year period to avoid the need for a constitutional amendment. • Possible use of spot tolls to defray the cost of specific projects, including a sunset provision when a project is paid. Other recommendations included: • Expand the use of ramp metering to improve traffic management. • Encourage mode shift through education and training. • Set higher standards for new roads. • Create more bike paths separated from traffic. • Recycle existing rail routes for transit. • Design multi-modal approaches to transit. Resources and Costs The citizens’ jury process is relatively costly and time-intensive. However, it can provide valu- able insights and suggestions during a public policy and planning process. Each juror was paid $100 per day, including travel costs for those coming in from out of town. Government expert witnesses were not paid. Currently, the Jefferson Center is paying $150 per day and there is ongo- ing debate as to whether or not this is sufficient to attract a diverse range of jurors that represents the affected population. Depending on the geography being covered, the cost to run the entire process can range from $50,000 for a regional area to $80,000 for a statewide jury and even more for a national jury. The venue used for the event was free because it was a government facility. A considerable amount of lead time is required to organize the event; in this case, some 20 differ- ent project staff persons were needed at various times to recruit the jury participants, mobilize moderators, lead small group discussions, serve as expert witnesses, serve as technical resource persons and writers, provide staff assistance, and participate as steering committee members and oversight committee observers. Another important resource consideration is the need to ensure all facilitators have been properly trained in the citizens’ jury process. Benefits: Why It Was Effective The Jefferson Center created the citizens’ jury process to amplify the voice of average citizens in the public policy debate process. While the process is time-intensive, the intricacies of pricing solutions are complex and require adequate attention to educate the representative members of an affected population so they can make informed recommendations. The process provided attendees with an understanding of the difficult tradeoffs required between individual decisions and the common good. The power of this process is in the empowerment of the participants as jurors. The jury process for the Minnesota value pricing policy debate was well received even by skeptics of the process. The seriousness with which the jurors addressed the problems of both increased congestion and insufficient financial resources is particularly noteworthy. While equity and EJ considerations were not the sole focus of the process, the issues presented to the jurors led to concerns about the effects of pricing strategies on low-income families. Of particu- lar benefit to the policy development process is that jury participants developed recommenda- tions for future considerations. While value pricing was ultimately adopted a decade later by the Twin Cities as a viable solution to reduce congestion, the early opposition laid the groundwork for transportation policy leaders to incorporate the issues and concerns raised by the citizens’

conducting citizen panels to explore Key Issues of Value pricing, minneapolis–st. paul Region, minnesota 279 jury process into future studies and deliberations. This information led to better positioning of the value pricing solution in future debates. Challenges and Limitations The citizens jury’ process is intensive and unpredictable by design, which can pose risks to public agencies. While the drama of a jury can attract media attention, it can also create overreli- ance on the “verdict.” As such, managing the media is an important consideration. The media response for this citizens jury’ was mixed. Talk radio programs turned complex issues into sound bites that did not represent the complete story; however, print media was informative. Managing the media is something that must be factored into this process and can be done by emphasizing that the citizens’ jury process is but one of many techniques to gather public input and opinion. In addition, it is not the intent of the citizens’ jury to dictate public policy but instead to give transportation planners and decision-makers insights into public opinion in a representative and objective manner to inform public policy. It is particularly critical to ensure that the focus ques- tion posed to the jurors is properly framed to avoid bias and allows jurors to explore as many potential answers to the question without being constrained by political consequences. The pro- cess is powerful and can be readily combined with other outreach techniques to set the course of plan and policy development. One challenge with the process is recruiting a strong mix of neutral and advocacy witnesses (for and against) who are highly informed about transportation mobility, financing challenges and the economic theory, travel behavioral response, and equity and EJ considerations. One has to weigh expertise with neutrality when choosing witnesses. Project sponsors—who are likely to become invested in specific solutions—must be careful to avoid biasing the process toward pre- determined outcomes. There is no hard and fast rule to follow, but transparency and high com- petency are of utmost importance in the selection of witnesses and the topics of testimony given. In the subject case, equity considerations were raised in the testimony of an opposition wit- ness. It would be important to any future citizens’ jury process that the issues of equity and EJ as they relate to tolling be given full consideration, drawing on a growing body of empirical research and new projects. Testimony should include more attention to the particular challenges for the most vulnerable populations, particularly low-income households and the unbanked, to secure mobility and access to opportunities (i.e., jobs, education, and healthcare) and how they may be specifically affected where non-tolled options may be limited. In the current era, further discus- sion of the mitigation options and their effectiveness must be part of the testimony. Lessons Learned While the citizens’ jury process elicited a meaningful exercise in public discussion of value pricing, several issues surfaced that provided lessons to the Minnesota DOT and the Humphrey Institute on how to frame the issue of value pricing in future public engagement activities. These included the following insights: • It is critical to position value pricing within the context of other congestion relief strategies such as a gas tax increase or mileage-based tax. The pros and cons should be couched in the context of other financing options. • Discussions should include the role of alternative transportation options (i.e., transit demand and travel behavior). • It is not advisable to structure a public involvement process around two problem statements. Specifically, when offering value pricing as both a funding and congestion relief strategy, the

280 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox public tends to focus on the funding issue. It is difficult to combine both objectives because the public tends to focus on the financial perspective because it is easier to understand. There- fore, it is advisable to frame value pricing as a congestion relief strategy comparable to other strategies such as capacity expansion and demand management strategies. • It is critical to introduce participants to the basic economic theory behind value pricing. The concept of value of time, supply and demand, and marginal costs are paramount to an informed discourse. While economists may complicate these concepts with technical jargon, it is critical for communications’ experts to translate the concepts into everyday terms so that citizens can make well-informed decisions about tradeoffs. Resources Buckeye, K., and Munnich Jr, L. 2004. “Value Pricing Outreach and Education: Key Steps in Reaching High- Occupancy Toll Lane Consensus in Minnesota.” Transportation Research Record: Journal of the Transportation Research Board, No. 1864. Transportation Research Board of the National Academy of Sciences, Washington, D.C. pp. 16–24. Citizens Jury. 1995. Report on Traffic Congestion Pricing, June 5–9, 1995. Retrieved from http://www.hhh.umn. edu/centers/slp/transportation/pdf/CitizensJury1995.pdf. FHWA. 2011. Planning for Your Lane Pricing Project, Part 2, Presentation. Retrieved from http://www.charmeck.org/ fastlanes/PDFs/Meetings/FHWA_CongestionPricingWorkshopSession6.pdf. Hubert H. Humphrey School of Public Affairs, University of Minnesota. 1996. Buying Time Guidebook: Political and Institutional Issues of Congestion Pricing. Retrieved from http://lgi.umn.edu/centers/slp/transportation/ pdf/BuyingTime-FinalReport.pdf.

281 Background The Dallas–Fort Worth Metroplex spans 12 counties and contains a population approaching 9 million. It is the fourth largest metro area in the United States. The North Central Texas Council of Governments (NCTCOG), the region’s Metropolitan Planning Organization (MPO), conducted two levels of EJ analysis as part of its Metropolitan Transporta- tion Plan (MTP), Mobility 2035. One analysis is included in the plan itself, and the other is a supplement to NCTCOG’s Regional Tolling Analysis (RTA). The goal of the RTA is to evaluate proposed expansion projects on the priced facility network and to understand their impacts on land use, air quality, and EJ populations. Both analyses identify places with a higher concentration of low-income and minority populations and consider the relative accessibility of these places to jobs, retail, and other important destinations. They also consider how this relative accessibility may change based on proposed transportation network changes. How It Was Done NCTCOG used two different methodologies in identifying EJ populations—one for the assessment in MTP and the other for EJ assessment in RTA. Mobility 2035. MTP examines how various system development scenarios will affect EJ pop- ulations compared to non-EJ populations. An EJ index method is used to identify block groups as either protected or non-protected. For each block group, an EJ score is calculated based on three socioeconomic factors: percentage below poverty level, percentage of minority, and persons per square mile. Block groups with an EJ score of 10 or higher on a scale of 1 to 100 are designated as protected. The score is used as an initial filter in the EJ analysis to help determine if additional analysis needs to be conducted. Approximately 36 percent of all block groups had an EJ index score above 10. These results were consistent with the overall low-income and minority popula- tions of the region, which were 12 and 47 percent, respectively, at the time of the analysis. This method follows an approach described in the NCHRP Report 532: Effective Methods for Environmental Justice Assessment (see text box, Formula for Environmental Justice Index Calculation). The assigned block groups are applied to the regional transportation model’s traffic c a s e e x a m p l e 2 Using an Environmental Justice Index to Identify Affected Populations, Dallas–Fort Worth Metro Region Framework Step • Conduct Impact Analysis and Measurement Stages in Decision-Making • Statewide and Metropolitan Planning Tools and Techniques • Environmental Justice Index • Travel Demand Model Affected Populations • Low-Income • Minority Participants • North Central Texas Council of Governments

282 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox survey zones (TSZs) (similar to traffic analysis zones), which use census information to estimate overall travel demand and roadway usage (see Tables 1 and 2). The Regional Tolling Analysis. RTA accounts for the system level analysis and is more focused on the specific impact of tolling projects. It also applies the EJ index method for cat- egorizing TSZs. However, in the RTA, the identification of populations is more detailed. TSZs are classified into four categories versus two in the MTP evaluation: non-EJ TSZs, minority alone TSZs, low-income alone TSZs, and both minority and low-income TSZs. Minority TSZs were categorized based on the federal Council on Environmental Quality guidance document Environmental Justice: Guidance under the National Environmental Policy Act. Based on 2010 Formula for Environmental Justice Index Calculation EJ Index = DVPOP × DVMAV × DVECO Where: DVPOP = degree of vulnerability based on population density. DVMAV = degree of vulnerability based on presence of minority population, and DVECO = degree of vulnerability based on presence of low-income populations. Source: Forkenbrock and Sheeley, 2004 Population per Square Mile Score 0 0 > 0 and < 200 1 > 200 and < 1,000 2 > 1,000 and < 5,000 3 > 5,000 4 Source: Forkenbrock and Sheeley, 2004 Table 1. Degree of vulnerability based on population density. Percent Minority or Percent Low-Income Score < State average 1 > State average and < 1.33 times the state average 2 > 1.33 and < 1.66 times the state average 3 > 1.66 and < 2.0 times the state average 4 > 2.0 times the state average 5 Source: Forkenbrock and Sheeley, 2004 Note: NCTCOG’s analysis utilized a regional instead of a state average. Table 2. Degree of vulnerability based on presence of minority populations and low-income populations.

Using an environmental Justice Index to Identify affected populations, Dallas–Fort Worth metro Region 283 Source: NCTCOG, 2013 Figure 1. EJ Index map for the Dallas–Fort Worth Region. U.S. Census data, any TSZs with a population greater than 50 percent minority were identi- fied as a minority TSZ. A TSZ that had 50 percent or more of the TSZ population residing in a block group with a median household income below the 2009 poverty level was identified as a low-income TSZ. Accordingly, any TSZ that fell below both of these categories was identified as a minority and low-income TSZ. Of the 5,252 TSZs in the NCTCOG travel demand model, 2,274 were identified as EJ TSZs. Tables 1 and 2, show how the EJ index range (0–100) breaks down into its individual compo- nents (DVPOP × DVMAV × DVECO). Figure 1 shows the results of the analysis. The TSZs with affected populations above the threshold were identified. They were then used to assess and compare the level of impact that proposed tolling projects will have on accessibil- ity and mobility performance indicators for EJ populations (i.e., protected traffic survey zones) to non-EJ populations (i.e., non-protected traffic survey zones). At the regional scale, five key performance indicators were examined (see text box, Performance Indicators for North Central Texas Council of Governments Environmental Justice Analysis). This comparison of performance indicators was conducted for various network and demo- graphic scenarios (e.g., current network, future no-build network with future demographics, future build network with future demographics, and future no-build network without priced facilities and with future demographics).

284 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Resources and Costs Prior to 2008, NCTCOG used consultants for its tolling analysis. However, a lack of con- sistency in the methodologies used prompted NCTCOG to hire two National Environmental Policy Act (NEPA) specialists and develop NCTCOG’s current analysis system. This process evaluates the potential effects of projects from planning to construction and assesses the impacts of projects in the long-range (MTP) and in a regional system of priced facilities (RTA). NCTCOG performs all work in-house. Because internal staff performed the work, costs were not identifi- able for these tasks. Benefits: Why It Was Effective NCTCOG’s methods of protected population identification are easily replicable. They use data sources that are commonly available and can be adapted to any region. A series of demo- graphic maps are presented in the MTP report to lay the foundation for the EJ Index Map (see Figure 1). The approach is easy to understand by both practitioners and the general public. In the presentation on the findings of performance indicators in MTP, NCTCOG also pro- vides an instructive chart to assist in understanding the impacts of tolling scenarios as they may be borne by EJ (protected) and non-EJ (non-protected) populations. Figure 2 displays a side- by-side comparison of impacts on three key indicators and three network scenarios: current, future no-build, and future build. The initial instructions and the chart layout make it easy for the general public to observe that a disproportionate burden has not been placed on protected populations. For instance, in the build network for 2035, congestion (as measured by “peak hour volume capacity ratio”) increased by 12 percentage points for non-protected populations and by 10 percentage points for protected populations. Similarly, the four categories used in the RTA approach enable a closer and systematically layered investigation of the impacts of tolling projects in places with higher concentrations of “minority alone” or “low-income alone” populations versus places with higher concentrations of minority and low-income populations. An example of how NCTCOG presented and com- pared impacts on performance indicators for the RTA is presented in the tool, “Using Travel Demand Models for EJ Assessment.” Performance Indicators for North Central Texas Council of Governments Environmental Justice Analysis Accessibility • Number of jobs accessible within 30 minutes by auto • Number of jobs accessible within 60 minutes by transit • Population within 30 minutes to special generators (e.g., universities, regional shopping centers, hospitals) Mobility • Average level of congestion • Average travel time

Using an environmental Justice Index to Identify affected populations, Dallas–Fort Worth metro Region 285 Challenges and Limitations One of the limitations in this analysis is the statistical comparability of the data used in identifying populations. In this example, the U.S. Census data for 2010 did not include income information and had to be supplemented by the 2005–2009 American Community Survey (ACS) data. Therefore, the income data that was used did not represent the same time period as the census data, and it also did not represent a specific year. Additionally, the demographic data were on a different geographic scale than for the TSZ model. The trans- fer of demographic data from U.S. Census geographies to the TSZs means that there is less reliability in overall accuracy. Finally, the demographic projections to the future condition assume the same distribution of population types and do not account for potential shifts across the region. This approach is used, in part, in lieu of any available data about the race, ethnicity, and economic status of the users of priced facilities in the Dallas–Fort Worth region. At the time of this research, NCTCOG indicated that it had plans to conduct a survey of the users of priced facilities that will include questions about race, ethnicity, and economic status. It also cautions that its modeling for the purposes of trip distribution, mode choice, and traffic assignment treat all vehicle trips of the same type identically. As implemented, the MPO’s model was not capable of generating results that produce outputs that differentiate vehicle trips based on the economic characteristics of transportation system users. How to Read the Chart: Performance Measure Population Current Network 2035 Build No-Build Percent Change (Build vs. No-Build) Protected Non-Protected Total 1,703,380 5,074,821 6,778,201 2,068,901 7,764,477 9,833,378 2,068,901 7,764,477 9,833,378 Number of Jobs Accessible within 30 Minutes by Auto Protected Non-Protected Difference 981,839 554,399 427,440 1,238,172 557,689 680,483 867,244 363,927 503,317 42.8% 53.2% Number of Jobs Accessible within 60 Minutes by Transit Protected Non-Protected Difference 1,500,158 863,602 636,556 2,643,513 1,836,797 806,716 1,847,516 937,488 910,027 43.1% 95.9% Percent of Lane Miles Congested Protected Non-Protected Difference 44% 41% 3% 54% 53% 1% 59% 63% -4% -8.7% -15.9% Source: NCTCOG, 2014 This represents the percent change in congestion levels in the Build and No-Build Scenarios. A negative number indicates a reduction in congestion. Here the protected population will experience 8.7% more congestion in the No-Build Scenario than in the Build Scenario. This represents the total number of people that live in a zone that is considered protected. For example, if a zone has a percentage of low-income individuals that is greater than the regional average of 11.9%, then the entire population of the zone, both low-income and non-low-income individuals is considered protected. This represents the percent more jobs available in the 2035 Build Scenario for both the protected and non-protected populations. Here the protected population has access to 42.8% more jobs in the Build Scenario than the No-Build Scenario. Figure 2. NCTCOG chart to compare the impacts between protected and non-protected populations.

286 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Lessons Learned The methodologies developed for the EJ analysis (i.e., both the EJ Index and the four separate EJ categories) arose because of the demographic circumstances in the Dallas–Fort Worth region. Close to 50 percent of the population in the metro area is minority. The team realized that if it used the methodology that classifies a census geography as an EJ area when the minority popula- tion is greater than 50 percent, the majority of the Dallas area would qualify. This would greatly dilute the rigor of the analysis. With this in mind, the team developed the two approaches used in the study. The lesson learned in this case is that EJ analysis methodologies are not “one size fits all.” Instead, they require approaches that are sensitive to the subject region’s demographic patterns. Also, both approaches are relatively straightforward and can be implemented quickly. Resources Forkenbrock, D. J., and Sheeley, J. 2004. NCHRP Report 532: Effective Methods for Environmental Justice Assessment. Transportation Research Board of the National Academy of Sciences, Washington, D.C. NCTCOG. 2014. Regional Tolling Analysis for the Dallas-Fort Worth Metropolitan Planning Area Based on Mobility 2035–2013 Update. Retrieved from http://www.nctcog.org/trans/ej/documents/RTAJan2014.pdf. NCTCOG. 2013. Mobility 2035–2013 Update Appendix B: Social Considerations. Retrieved from http://www. nctcog.org/trans/mtp/2035/documents/AppendixB.pdf.

287 Background The Louisville–Southern Indiana Ohio River Bridges Project (Ohio River Bridges Project) is a construction and reconstruction project being undertaken to address long-term cross-river transportation needs in the Louisville metropolitan area. The project has been developed over more than 40 years in recognition of the need to improve cross-river mobility between Jefferson County, Kentucky, and Clark County, Indiana. In September 2003, FHWA issued a Record of Decision (ROD) that identified the preferred alternative in the Final Environmental Impact Statement (EIS) as two new Ohio River bridge crossings, connected approaches, and the reconstruction of the Kennedy Interchange. Shortly afterwards, a lawsuit was filed that delayed the planned bridge-crossing project for several years. This delay was followed by a decision to con- sider tolling as a potential way to finance construction of these bridges. The Commonwealth of Kentucky and the State of Indiana are jointly the project sponsors for the Ohio River Bridges Project. In support of this partnership, each state has agreed to take the lead in financ- ing and overseeing the construction of one-half of the project. Thus, Kentucky is responsible for financing and constructing the Downtown Crossing while Indiana is doing the same for the East End Crossing. In January 2011, the leaders of Kentucky and Indiana determined that the project was not financially feasible and directed the Louisville and Southern Indiana Bridges Authority (the Bridges Authority) to look for cost savings options, consider the use of tolls as part of the financ- ing options, and prepare a Supplemental Final Environmental Impact Statement (EIS). The Bridges Authority evaluated alternative delivery options for the project, including (1) a design-build construction approach, financed with tax-exempt toll revenue bonds and combined with a separate operations and maintenance contract following con- struction and (2) an availability payment public-private partnership (P3) model. In December 2011, the two governors and the Bridges Authority announced that the states would use both of the preferred delivery options. Kentucky elected to use a design-build contracting approach for the procurement of the Downtown Crossing and Indiana chose to use an availability payment P3 approach for delivering the East End Crossing. With environmental approvals in place, the project would become eligible for innovative financing, including Grant c a s e e x a m p l e 3 Mobilizing a Local Liaison to Recruit Community Leaders for Survey, Louisville–Southern Indiana Ohio River Bridges Project Framework Step • Scope Approach to Measure and Address Impacts • Identify and Assess Mitigation Strategies Stages in Decision-Making • Project Development/Environmental Review • Maintenance and Operations Tools and Techniques • Trusted Local Liaison • Community Leader List • Community Leader Survey Affected Populations • Low-Income • Minority Participants • Kentucky Transportation Cabinet • Indiana DOT • Louisville and Southern Indiana Bridges Authority • Market Research Firm

288 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Anticipation Revenue Vehicles (GARVEE) bonds and a Transportation Infrastructure Finance and Innovation Act (TIFIA) loan. The 2012 Supplemental Final EIS modified the preferred alternative to achieve cost savings and incorporate tolling to the project. The modified preferred alternative still included the two new Ohio River bridge crossings, connecting approaches, and the reconstruction of the Kennedy Interchange. On June 20, 2012, FHWA issued a revised ROD, approving the Modified Selected Alternative approach. In the revised ROD, FHWA found, based on the user cost analyses, that the project would have disproportionately high and adverse effects on minority and low-income populations. Although the project’s adverse effects would not be predominantly borne by EJ populations, the impacts would be appreciably more severe or greater in magnitude for these populations (see tool, “Assessing User Costs and Household Burden Effect”). Despite this finding, the revised ROD approved tolling on the Downtown and East End bridges based on the judgment that toll- ing was necessary to fund the project, which will deliver needed improvements in mobility. As a condition of approval, the revised ROD identified several measures to minimize and mitigate the effect of tolling on EJ populations, including: • The Kentucky Transportation Cabinet (KYTC) and the Indiana DOT committed to enhance bus service as part of the Modified Selected Alternative. • The two state agencies agreed to monitor changes in traffic patterns and to identify strategies to address any unanticipated disproportionately high and adverse effects, if any, resulting from the project’s implementation. • Prior to the implementation of the tolling, the two states were expected to adopt a “Tolling Policy” that is sensitive and responsive to low-income and minority populations. While devel- oping the Tolling Policy, KYTC and Indiana DOT were expected to (1) conduct a detailed assessment of the potential economic effects of tolls on low-income and minority popu- lations, using the latest available population data, traffic forecasts, and community input; (2) make the results of the study publicly available; and (3) identify and evaluate a range of measures for mitigating the effects. To advance the revised ROD mitigation commitment, FHWA required that KYTC and Indiana DOT, the joint sponsors of the Ohio River Crossing Project, obtain further information from the affected EJ communities that would be impacted by tolling the two proposed new bridges. To understand the impact of the proposed tolling from the perspective of the EJ communi- ties, a consultant was asked to interview leaders who represented, advocated for, or worked with low-income, minority, and/or low-income and minority residents in the Louisville–Southern Indiana region. The tolling policy’s development was to be informed, in part, by additional outreach and public involvement with EJ populations through several activities, including public meetings, a telephone interview opinion survey, open house events, an EJ Community Survey (see case example, “Targeting Local Grocery Stores to Administer Community Surveys, Louisville– Southern Indiana Ohio River Bridges Project”) and a community leader survey. How It Was Done In 2013, a local research consultant was selected to undertake a community leader survey on behalf of KYTC and Indiana DOT to better understand the impacts of tolling on two new toll bridges from the perspective of EJ communities. The consultants first assembled a potential target contact list of leaders, largely drawing from the Mayor’s Violence Prevention Work Group Task- force on the Kentucky side and an initial list furnished by Indiana DOT on the Indiana side.

mobilizing a local liaison to Recruit community leaders for survey, louisville–southern Indiana Ohio River Bridges project 289 Based on these two sources, the research consultant had an initial list of about 48 leaders split between Kentucky (32 leaders) and Indiana (16 leaders) coming from businesses, civic and reli- gious organizations, and affected neighborhoods. To help reach the leaders in these communities, the consultant contacted a local liaison known to them and the communities. This individual was a well-known African American resident of Louisville with deep ties to and trusted relationships with the EJ communities. His background included working as a communications specialist, community development professional, and public education administrator. In addition, he hosted a local weekly radio program that boasted significant listenership within the EJ communities, and he served as an ordained deacon in one of the 100 largest churches in the country. The liaison’s initial project responsibilities included supplementing the initial lists of EJ com- munity leaders, particularly on the Kentucky side of the Ohio River where he had his strongest relationships. He then introduced and endorsed the project and encouraged the leaders to make time to participate in the survey when the research consultants called. The consultants handled the logistics of scheduling calls with the participants. Following this approach, the liaison added 22 community leaders (19 Kentucky leaders and 3 Indiana leaders) to the list. These 70 leaders were African Americans, Hispanics, and Caucasians. The strategy of engaging a trusted liaison in the process of identifying local leaders, contacting them, and scheduling appointments was highlighted in FHWA’s 2011 publication, Environmental Justice Emerging Trends and Best Prac- tices Guidebook. Invitation and Scheduling Process. Once authorized, the local liaison began extending invitations to the identified 51 Kentucky leaders via email, phone call, or in face-to-face conver- sations. The names of those individuals that agreed to participate were given to the consultant to formally schedule an interview. Of the identified 19 Indiana leaders, 16 had been previously contacted by an Indiana DOT representative; therefore, the consultant contacted the Indiana list directly by phone call or email. In both cases, the research process was explained, and each leader was invited to schedule a phone call or in-person meeting at a time and location that was convenient for them. Over a one month period in the summer, the liaison and the consultant made at least three phone calls and/or email attempts to invite and involve each of the 70 leaders on the initial list. At the close of this period, an email was sent to those leaders that had not yet completed an interview, had not yet scheduled an interview, or had not declined to participate. The email encouraged them to share their thoughts online. The design and questions in the online survey were the same as those presented to the leaders in phone calls and at in-person interviews. As leaders completed surveys, they were asked if they could recommend other leaders that should be contacted. This information resulted in an additional 27 leaders being identified. Each new leader was then contacted one or two times by the consultant and asked to participate in an interview. Those individuals whose names were provided during the last full week of data col- lection were invited to offer feedback online. Interview Protocol. Before their scheduled interview, all leaders agreeing to participate in an interview were asked to review two educational pieces describing the Ohio River Bridges Project, the tolling assumptions, and the recommended mitigation strategies. The consultant emailed the leaders links to a 56-page report, Draft Assessment of Economic Effects of Tolling and Potential Strategies for Mitigating Effects of Tolling on Low-Income and Minority Populations dated June 24, 2013, and a 5-minute video posted on the project website and YouTube. Giving the leaders the opportunity to review the information before it was released to the general public was a way of conveying to the leaders that their participation mattered. Of the 97 leaders identified, three

290 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox indicated they did not have internet access or preferred to review a paper copy of the report. Two of these individuals were mailed paper copies of the documents to review and DVDs of the video. A third individual was mailed a paper copy of the document, but was unable to view the video as she did not have access to a DVD player. All information presented to the leaders was available in English and Spanish and the lead project manager on the consultant team was a Spanish speaker. Respondent Sample Size. By the close of the interview period, 38 of the 97 leaders com- pleted an interview (39.2%). Of the 38 leaders responding, 25 were from Kentucky and 13 were from Indiana. A total of 28 interviews were completed by phone call and 10 were submitted online. The population on the Kentucky side of the river was inhabited overwhelmingly by African Americans and very low-income Caucasians, whereas the populations on the Indiana side of the river included some Hispanics, as well as African Americans and Caucasians. His- torically there had been much more involvement from the African American community in Kentucky relative to EJ issues and transportation than on the Indiana side. Survey Questions and Results. The community leader survey interview questions explored the impact of tolling on community behavior and lifestyle, whether available free bridges or transit services would serve as travel options, strategies to increase transponder usage, and per- ceived effects of uniform tolling rates. The key themes, wording of the questions, and a summary of community leader responses are presented in Table 1 and described in the report, Assessment Theme/Questions Community Leader Survey Responses Tolling Impact on Commuting Behavior—How will the addition of tolls impact the community decisions of the individuals you represent? Change/impact No Change 11% Switch to non-tolled routes Yes 79% Reduce cross-river trips Yes 45% Carpool, use public transit Yes 18% Tolling Impact on Lifestyle—How will the addition of tolls impact the lifestyles of the individuals you represent? ` Change/impact No Change/Other 11% Switch jobs Yes 24% Change residence, doctors, schools Yes 21% Other ** Various Changes 68% Free Bridges as an Effective Option to Avoid Tolls—The materials indicated that the Sherman Minton Bridge (I-64) and the Clark Memorial Bridge (US 31/Second Street Bridge) will remain non-tolled as part of the Bridges project, meaning cross-river travelers will have two free river crossing alternatives. In your opinion, are these effective options for travelers who wish to avoid paying a toll? Yes 58% No 18% Unsure 21% Table 1. Questions and community leader responses.

mobilizing a local liaison to Recruit community leaders for survey, louisville–southern Indiana Ohio River Bridges project 291 Theme/Questions Community Leader Survey Responses Online Ordering 53% Low-Minimum Balance 45% Convenient Transponder Refills 71% Account Tied to Card or Bank Account 45% Minimum Transponder Balance—The report and video indicated that transponder accounts could require only a low-minimum account balance to be established. In your opinion, what amount of money would your constituents consider to be a low-minimum amount? Less than $20 70% Multi-Level Tolling Rate—The proposed mitigation measures do not include multi-level tolling rate which would give general motorists one rate and low- income residents or minorities another. In your opinion, how do you think this decision will impact individuals for whom you advocate? Lack of two-tier system would be a hardship 30% Tolling system is fine as proposed 16% Source: KYTC and Indiana DOT, 2014 Note: ** Leaders whose responses placed them in the "other" category provided comments, including that the burden of tolls will impact families in their travel activities and their ability to use their income for family and household expenses. Impact of Transit Authority of River City (TARC) Changes for All Travelers—Funds have been provided for TARC to buy more buses and vans, create more park-and- ride lots, and make other public transportation improvements. In your opinion, are these effective options for travelers who wish to avoid paying the toll? Yes 34% No 24% Unsure 42% Will People Change / Consider Using TARC—Given the proposed improvements to TARC's service, would the individuals you represent consider using public transportation to cross the bridge instead of driving? Yes 18% No 37% Maybe 42% Strategies Likely to Increase Transponder Usage—If the following conditions were met, would this increase, decrease or not impact your likelihood of using a transponder? Yes Overall 86% Lower Toll Rate 76% Free Transponder 87% Convenient Locations 79% Table 1. (Continued).

292 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations, issued in April 2015 by KYTC and Indiana DOT. In addition to the questions shown in Table 1, community leaders were also asked their views on other questions, including: • In your opinion, how will the Ohio River Bridges Project affect the neighborhoods where low-income and minority populations reside? To this question, the plurality of responses indicated that the project would result in increased traffic on non-tolled routes, but concerns were also expressed for economic development, quality of life, and financial burden on house- holds, among other issues. • What other investments and improvements, if any, could be made to mitigate the impact of tolling on low-income and minority residents and neighborhoods? To this question there were varied responses suggesting price reductions, relying on other taxes, expressing opposi- tion to tolls, giving low-income individuals jobs on the project, and needed investment in light rail transit, among others. Before wrapping up, community leaders were given an opportunity to make additional com- ments. These responses were varied, but 22 percent of the comments focused on the bridges’ direct financial effects on the EJ populations and emphasized the importance of having EJ com- munity members included in the conversation and for governments to understand their needs and how the cost of cross-river travel will affect them. Resources and Costs The local trusted liaison was paid $100 per hour and worked 13 hours to carry out his support activities. This work was accomplished in increments of 30 to 90 minutes each over the course of about 10 days. The actual surveys were conducted by the research consultant team. The full costs of the research consultant team’s efforts over the intensive six-week period were unavailable. Benefits: Why It Was Effective Identifying a trusted local liaison in the EJ communities on the Kentucky side of the river was paramount to the level of success achieved. The liaison was able to leverage long-held personal and professional relationships at many levels. Within 13 hours, 22 leaders were identified and contacted at a cost of only $1,300. Challenges and Limitations Construction on the two new toll bridges began prior to starting the community leader survey, and some of the face-to-face interviews were conducted within sight of ongoing construction. Because of this, there may have been some skepticism among leaders who believed that their participation was not being taken seriously because the construction was already underway. Leaders by definition are busy people with multiple responsibilities in many activities. Their time is precious and some wanted to know if participating would be worth their time or just an exercise that was not going to be taken seriously. There were several challenges in developing a list of leaders. Indiana had three separate small decentralized municipalities compared to Louisville, a large urban city. The more fragmented environment in Indiana made it challenging to identify leaders who spoke for a large section of those communities. In addition, the local liaison had more relationships in Kentucky than in Indiana.

mobilizing a local liaison to Recruit community leaders for survey, louisville–southern Indiana Ohio River Bridges project 293 Some of the leaders identified in the Louisville Mayor’s 2012 Violence Prevention Work Group felt that they did not have specific knowledge of the issues to be comfortable talking about the project. They also felt that their time would have been more effectively spent with causes with which they were more invested and more familiar. In short, using this group as a proxy for identifying EJ leaders was largely, but not entirely, effective. The consultant was also required to work under a fairly aggressive time constraint and had only 5 weeks to engage the local liaison, create a survey instrument and have it approved, iden- tify leaders, engage and schedule interviews with the leaders, get the leaders to review a 56-page report and video, and interview the leaders. Lessons Learned The effectiveness of a trusted local liaison is deeply connected to the social capital—the social, religious, and cultural relationships—built by the individual through their good works and exposure to various organizations and persons in their affected communities. Finding trusted local liaisons often depends on the prevalence and capacity of civic, community, and advocacy organizations and how interested these organizations and their membership may be with the particular transportation-related issues being presented for feedback discussions. Additionally, not all communities have individuals who are available or willing to take on the role of the trusted liaison. In this instance, the lead project manager had a long-standing relationship with this liaison and, as such, there was mutual trust between both persons that allowed the partner- ship to develop so quickly. Some of the leaders identified in the initial list had direct relationships with the local liaison and may have been more inclined to respond to his request as a result. Leaders on the second list may not have had a direct relationship with the local liaison and this degree of separation may have contributed to their not feeling as strong an obligation to participate. Not all the leaders had access to the internet. Some wanted paper copies and one leader did not have DVD access. Practitioners should remain open to the possibility that multiple forms of communication may be necessary for a successful outreach effort. Resources FHWA. 2011. “Trends in Public Involvement/Community Engagement.” Environmental Justice Emerging Trends and Best Practices Guidebook. FHWA-HEP-11-024. Retrieved from http://www.fhwa.dot.gov/environment/ environmental_justice/resources/guidebook/ejguidebook110111.pdf. KYTC and Indiana DOT. 2014. Appendix E4 EJ Community Leader Survey. Retrieved from http:// kyyinbridges.com/wp-content/uploads/Appendix-E4_IQS-EJ-Community-Leader-Survey-IQS.pdf. KYTC and Indiana DOT. 2015. Assessment of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations. Retrieved from http://kyinbridges.com/wp-content/ uploads/2015-04-24_EJ-Assessment-Plan.pdf.

294 Background In April 2012, a Supplemental Final EIS was signed for the Louisville– Southern Indiana Ohio River Bridges Project, followed in June 2012 by a revised ROD. As a post-revised ROD mitigation commitment, KYTC and the Indiana DOT were directed to obtain information from the EJ communities that would be impacted by constructing and tolling two proposed new bridges across the Ohio River. To understand the impacts from the perspective of the EJ communities, a consultant was charged with engaging and interviewing community members who were low-income, minority, and/or low-income and minority residents within the area of the project that would be impacted by bridge con- struction and tolling in the Louisville–Southern Indiana region. How It Was Done In 2013, a consultant was selected to survey low-income, minority, and/or low-income and minority community members on how the construction and tolling of two new bridge crossings of the Ohio River would impact them. Using demographic information obtained from the Supplemental Final EIS that identified EJ communities in the Louis- ville Metropolitan Planning Area, six EJ block groups were identified as areas of concern for direct and indirect impacts related to the proposed project (see Figure 1). The consultant decided that the most appropriate venue to engage members of these communities was at grocery stores within those cen- sus blocks. This decision was based on the consultant’s belief that these grocery stores were likely frequented by individuals living in the imme- diate vicinity of these stores. Using a series of online searches, the consultant generated a list of full-service Supplemental Nutrition Assistance Program (SNAP) gro- cery stores located in the identified block groups and made phone calls to the individuals with decision-making authority over these stores. Store owners at four grocery stores located in the target communities gave approval for the consultant to conduct on-site interviews with their customers in exchange for the consultant purchasing $20 gift cards/vouchers c a s e e x a m p l e 4 Targeting Local Grocery Stores to Administer Community Surveys, Louisville–Southern Indiana Ohio River Bridges Project Framework Step • Scope Approach to Measure and Address Impacts • Identify and Assess Mitigation Strategies Stages in Decision-Making • Project Design and National Environmental Policy Act • Implementation Tools and Techniques • Word of Mouth • Traveling Kiosks and Video • Participation Incentives • Survey Instrument Affected Populations • Low-Income • Minority Participants • Kentucky Transportation Cabinet • Indiana DOT • Louisville and Southern Indiana Bridges Authority • Market Research Firm • Community Grocery Stores

Targeting local Grocery stores to administer community surveys, louisville–southern Indiana Ohio River Bridges project 295 from their stores that were good only at their stores. The consultant was allowed to approach store customers, offering a $20 gift card/voucher to those who would view a short video and complete a 17-question survey. Site Identification. Participating grocery stores were identified in low-income and minor- ity communities in Louisville, including both locally owned and franchise-type establishments. One participating locally owned grocery store was situated in the Phoenix Hill neighborhood, just east of downtown Louisville directly across the street from Dosker Manor, a 685-unit low- income, disabled persons, and senior facility operated by the Louisville Metro Housing Author- ity. Very few other retail outlets were available in the vicinity. Another grocery site was a local franchise store located in the Portland neighborhood in Western Louisville that contained a number of Section 8 housing properties. Many of the stores near it were also discount stores. In Indiana, two local franchise discount grocery stores participated in the survey, including one in Jeffersonville, situated in a shopping center with other discount stores. The other was located in New Albany with a competing discount chain across the street. Both stores were located near mixed-income neighborhoods. Data Collection Techniques. At each of the four participating grocery stores, the consultant used a two-person team and a traveling kiosk to engage store customers. The traveling kiosk consisted of a wireless-enabled, digital tablet mounted on a platform that could be moved from store to store (see Figure 2). As a customer approached a grocery store, the consultant would ask the customer if he or she would view a short (5 minutes 14 seconds) educational video that ENVIRONMENTAL JUSTICE BLOCK GROUPS Source: U.S. Census Bureau, 2010 Note: A–F are environmental justice areas described above. Figure 1. EJ block groups in the area of concern.

296 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox outlined the project, tolling assumption, and recommended potential mitigation measures, and answer a 17-question survey. If the customer agreed to participate and completed the survey, the consultant gave the customer a $20 gift card/voucher to the grocery store. The video and survey were available in English and Spanish and team members conducting the study were Caucasian, African American, and Hispanic. Depending on the operator of the establishment, the kiosks were located just outside the entrance to the store or in the entryway to the store. All locations were highly visible and the consultant was able to approach customers with ease. For the occa- sion, consultant staff were dressed in business casual attire. As it turned out, the response was so great that the consultant really did not need to approach people; they stood in line for a chance to complete the survey (see Figure 3). Best Time to Reach Out to Customers. Once the consultant identified potentially avail- able grocery stores in the project area and reached agreements with them to purchase their gift cards/vouchers, their efforts turned to identifying the stores’ high-volume shopping hours. They decided to schedule events on or close to the first of the month so they could maximize inter action with as many Electronic Benefit Transfer (EBT) customers as possible. They were also interested in taking advantage of the long Fourth of July weekend, where possible, and the festivities sur- rounding it. Two of the store owners recommended conducting outreach on Friday, Saturday, and Sunday; one of the other store owners recommended Wednesday, Saturday and Sunday, while the other store thought outreach would be best on Friday and Sunday. Generally outreach was conducted on weekdays between mid-afternoon and early evening, on Saturdays between late morning and mid-afternoon, and on Sundays in the afternoon. Given that data collectors were standing on their feet during the entire outreach process, this schedule allowed each two- person teams to work for a manageable 4 hours per day. Outreach Schedule. The outreach process was set for a short and challenging, 11-day turn- around period in early July 2013. The consultant sought to visit as many stores as possible close Figure 2. Digital tablets were used at traveling kiosks to administer the survey.

Targeting local Grocery stores to administer community surveys, louisville–southern Indiana Ohio River Bridges project 297 to the first of the month—a period when customers were most likely to have received food and other social service benefits. The period selected also allowed the consultant to take advantage of the Fourth of July holiday and 2 weekends. Launching the process, the consultant piloted its data collection efforts at a single location. The slow start allowed them to be certain that the outreach plan, videos, and related technology were working properly before fielding it at multiple locations throughout the region simultane- ously. The consultant did not formally advertise the days or times that surveys were to be con- ducted, but it was clear after just one day of fielding that word-of-mouth would quickly spread the message. During the 2 weekends, the consultant completed 287 surveys, with 247 individuals identify- ing themselves as EJ community members. In some cases, shoppers waited in line for 2 to 3 hours to complete a survey. As it turned out, the consultant had to turn away so many people that peak times probably were not necessary to meet their engagement goals. Survey Questions and Results. The community survey questions explored the impact of tolling on community behavior and lifestyle, whether available free bridges or transit services would serve as travel options, strategies to increase transponder usage, and perceived effects of uniform tolling rates. The key themes, wording of the questions, and a summary of community responses are presented in Table 1 and described in report, Assessment of Economic Effects of Toll- ing and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations, issued in April 2015 by KYTC and Indiana DOT. The survey patterns reported are for the segment of the population that identified themselves as EJ community members (86% of respondents). Before wrapping up, community members were given an opportunity to make additional com- ments. These responses were varied, but the most common comments made were the desire for Figure 3. People lining up to take the Community Survey.

298 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Theme / Questions Community Member Individual Responses Tolling Impact on Commuting Behavior—How will the addition of tolls impact your commuting decisions? Change/impact No Change 31% Switch to non-tolled routes Yes 31% Reduce cross-river trips Yes 26% Carpool, use public transit Yes 18% Tolling Impact on Lifestyle—How will the addition of tolls impact your lifestyle? ` Change/impact No Change/Other 66% Switch jobs Yes 9% Change residence, doctors, schools Yes 10% Other ** Various Changes 9% Free Bridges as an Effective Option to Avoid Tolls—The video indicated that the Sherman Minton Bridge (I-64) and the Clark Memorial Bridge (US 31/Second Street Bridge) will remain non-tolled as part of the Bridges project, meaning cross-river travelers will have two free river crossing alternatives. In your opinion, are these effective options for travelers who wish to avoid paying a toll? Yes 70% No 14% Unsure - Impact of Transit Authority of River City (TARC) Changes for All Travelers—Funds have been provided for TARC to buy more buses and vans, create more park-and- ride lots, and make other public transportation improvements. In your opinion, are these effective options for travelers who wish to avoid paying the toll? Yes 63% No 17% Unsure - Will People Change / Consider Using TARC—Given the proposed improvements to TARC's service would you consider using public transportation to cross the bridge instead of driving? Yes 48% No 27% Maybe - Strategies Likely to Increase Transponder Usage—If the following conditions were met, would this increase, decrease or not impact your likelihood of using a transponder? Yes Overall 58% Lower Toll Rate 40% Free Transponder 58% Convenient Locations 48% Table 1. Community survey questions and community member responses.

Targeting local Grocery stores to administer community surveys, louisville–southern Indiana Ohio River Bridges project 299 having no tolls, keeping tolls affordable, stopping tolls once the bridge is paid for, and expanding bus service in southern Indiana. Resources and Costs The consultant spent $5,740 buying 287 gift cards valued at $20 each from four local grocery stores. Each customer who watched a short project video and completed a 17-question survey received the $20 gift card/voucher for use at that grocery store. Benefits: Why It Was Effective Being able to offer incentives was an invaluable tool given the short turnaround time. Because of the level of poverty, customers were willing to wait in line for several hours to participate in the process. This may have provided a degree of assurance to the consultant that they were effec- tively reaching the target customers of any race/ethnicity living at and below the poverty level. The delays, however, from a data collection standpoint were an unfortunate and unintended consequence and not by design. While giving away $20 gift cards/vouchers may at first seem too costly, it would have taken far longer for the consultant to try and engage the target audience without such incentives. It would have required more labor-hours from the consultant’s survey staff to try and engage the target audience, been more difficult to get store owner participation, and likely exceeded the consultant’s deadline for project completion. Theme / Questions Community Member Individual Responses Low-Minimum Balance 44% Convenient Transponder Refills 44% Account Tied to Card or Bank Account 43% Minimum Transponder Balance—The report and video indicated that transponder accounts could require only a low-minimum account balance to be established. In your opinion, what amount of money would you consider to be a low-minimum amount? Less than $20 70% Multi-Level Tolling Rate—The proposed mitigation measures do not include multi-level tolling rate which would give general motorists one rate and low- income residents or minorities another. In your opinion, how do you think this decision will impact you? Lack of two-tier system will have a negative impact 13% Will not impact me 43% Source: KYTC and Indiana DOT, 2014b Note: ** Leaders whose responses placed them in the "other" category provided comments, including that the burden of tolls will impact families in their travel activities and their ability to use their income for family and household expenses. Online Ordering 45% Table 1. (Continued).

300 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Challenges and Limitations This example highlights the approach to undertaking a convenience sample, one of the main types of non-probability sampling methods. A convenience sample is made up of people who are easy to reach. A relatively comparable opinion survey of community leaders was also under- taken (see case example, “Mobilizing Local Liaison to Recruit Community Leaders for Survey, Louisville–Southern Indiana Ohio River Bridges Project”). The similarities and differences reg- istered by the two populations were compared in the Economic Impact Assessment Report and in other supporting documentation prepared by the market research firm. While informative in the design of a tolling policy, the community member survey sought to solicit an opinion and did not include current travel behavior characteristics such as the type and number of trips taken, routes used, mode chosen, or time-of-day. Thus, the survey was limited in what it could suggest about probable travel behavior impacts before and after the implementation of a tolling deci- sion. The survey itself was administered over a short period of time in summer to a small sample size, which is particularly important if the data were to be stratified by geography of origin and destination and trip purposes. Despite these limitations, this case example conveys an effective convenience sampling approach for securing feedback in low-income communities. It highlights several creative, replicable strategies and navigates some challenges related to fielding surveys. Given that the geographic target area was very narrowly defined and the options for finding grocery stores in typically underserved areas were low, the consultant was perhaps fortunate to find four grocery stores willing to participate on such short notice. During this period, the market research con- sultant had approached regional and national chains, but these organizations did not respond to their requests in time to include them in the process. The consultant generally found it productive to work directly with store owners who had decision-making authority rather than store managers. The managers would have had to run the request up the corporate ladder, a process not conducive to the consultant’s deadline. Ultimately, when the owner of one of the franchise operations agreed to participate, the con- sultant found that they could effectively parlay the terms of that agreement in their dealings with the other participating chains as well as with the locally owned establishment. The value of the gift card was decided during a discussion with the first store owner who agreed to participate. The owner indicated that $20 was the lowest denomination of gift card that they had. As a result $20 was used with each of the other three stores. The franchise grocery operations used gift cards, while the locally owned establishments used paper vouchers. In both cases, it took quite some time to activate 287 gift cards/vouchers. Each $20 gift card had to be individually activated at a cash register and then checked by the consul- tant to make sure the $20 amount had registered on the card. This effectively shut down one checkout lane for some time. The paper vouchers each had to be individually signed by the store owner, and while not a particularly efficient method, could be done without shutting down a checkout lane. For those considering a similar approach, it is helpful to consider some of the other challenges that the survey research team faced in fielding the survey. These challenges are listed below. • Safety concerns can become a challenge. In this case, two customers, unrelated to the survey implementation process, got into a knife fight and were injured at one of the grocery stores where the consultant’s survey takers were working. This incident caused a member of the consultant’s survey team to quit over concerns for safety. • Notice of unusual credit card activity. One of the consultant’s staff attempted to buy a large number of the gift cards using the company credit card. Initially, the firm’s bank denied the

Targeting local Grocery stores to administer community surveys, louisville–southern Indiana Ohio River Bridges project 301 charge because it was an unusual purchase for the research firm. Management at the firm had to call the bank to verify the charges and arrange for future similar purchases to be accepted. This cost the consultant time, and to expedite matters, the staff member responded to the initial declined charge by using a personal card to handle the gift card purchase, which was processed without question. • Avoiding glare with use of the electronic, hand-held tablet. The consultants were able to use their hand-held tablet outside to show the video and provide a touch screen for conducting the survey because they operated under a covered walkway or inside the store to avoid any potential problems with glare. • Working with low literacy populations. There were some challenges with literacy and the basics of using a touch screen. Each two-person team member was trained to provide consis- tent, non-leading clarifications for questions. • Short shifts for survey takers proved humane and effective. Because the two-person teams were always on their feet, they were asked to work only four-hour shifts. Also, these short shifts eliminated the need for the consultant to ask the store owners if they could use a store’s bathroom facilities. Lessons Learned There are some valuable lessons to take away from implementing the approach, including the following: • Offering to buy gift cards/vouchers from grocery store owners and requiring those receiving the gift cards/voucher to use them at that same store was a win-win for the store owners and customers alike. As such, it increased the store owners’ willingness to participate. • Once the store owners and staff realized that the consultant would not be hassling their cus- tomers, they became more comfortable and accommodating. The employees working in the stores were also generally interested and intrigued by what the consultant was doing. It helped that the consultants were buying a large order of gift cards/vouchers. Store owners and staff saw the consultants as an interesting profit generator and largely treated them well. • Offering community members a gift card to their local grocery store allowed them to purchase more groceries than they had anticipated. The value of the gift card increased willingness for community members to participate. • The approach created efficiencies for survey-taking. The approach allowed the consultant to take advantage of the fact that people were already going to be at the grocery store. This allowed the consultant to stay in one place and conduct surveys rather than having to go out looking for members of the target population. • Working at independent locations and locally owned franchises made it easier to ramp up the survey-taking events. Because store owners had the authority to make decisions, consultants found that independent locations and franchises that were locally owned were more respon- sive than the regional or national chains. • Having more than one traveling kiosk available at each grocery store would have reduced the wait time for customers. It was challenging for the consultant team to watch individuals wait for hours in line for the opportunity to receive a $20 gift card; this was a palpable demonstra- tion of the difficult economic circumstances of the respondents. Resources FHWA. 2011. “Trends in Public Involvement/Community Engagement.” Environmental Justice Emerging Trends and Best Practices Guidebook. FHWA-HEP-11-024. Retrieved from http://www.fhwa.dot.gov/environment/ environmental_justice/resources/guidebook/ejguidebook110111.pdf.

302 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox KYTC and Indiana DOT. 2014a. Appendix E3, Louisville-Southern Indiana Ohio River Bridges Project EJ Commu- nity Survey Populations. Retrieved from http://kyinbridges.com/wp-content/uploads/Appendix-E3_IQS- EJ-Community-Survey-IQS.pdf. KYTC and Indiana DOT. 2014b. Appendix E5, EJ Comparison of Surveys Conducted to Identify EJ Populations’ Perceptions of Tolling Options and Potential Mitigation for Disproportionate Adverse Effects of Tolling on Low-Income Commuters and Travelers in the Project Area. Retrieved from http://kyinbridges.com/wp-content/ uploads/Appendix-E5_Qk4-Comparison-of-Surveys-IQS-Community-and-Leader-2.pdf. KYTC and Indiana DOT. 2015. Assessment of Economic Effects of Tolling and Strategies for Mitigating Effects of Tolling on Environmental Justice Populations. Retrieved from http://kyinbridges.com/wp- content/uploads/2015-04-24_EJ-Assessment-Plan.pdf. U.S. Census Bureau. 2010. 2006–2010 ACS. Environmental Justice Block Groups. Map.

303 Background The Los Angeles Congestion Reduction Demonstration (CRD) Express Lanes Program resulted in the conversion of HOV lanes on I-10 El Monte Freeway (from I-605 to Alameda Street) and I-110 (from Adams Boule- vard to Harbor Gateway Transit Center/182nd Street) to HOT lanes. This project was part of a larger CRD program funded by the U.S. DOT in 2007. The Los Angeles CRD partners include the Los Angeles County Metro politan Transportation Authority (Metro) and the California DOT (Caltrans), District 7. Metro was the designated lead agency for the Los Angeles CRD project and responsible for overseeing the policy, planning, and design of tolling and transit operations, and served as the coordinating body for all of the local partners [Gardena Transit (GTrans), the Los Angeles DOT, Metrolink, and Torrance Transit] and enforcement partners (California Highway Patrol). Caltrans owns the facilities and maintained an oversight and advisory role during the planning and implementation stages of the Los Angeles CRD. Metro employed four complementary strategies, including tolling, transit, telecommuting/travel demand management, and technology. For the toll lane conversion to be implemented, state enabling legis- lation (SB 1422, 2008), which contained an important provision on potentially affected low-income commuters, was enacted. The legisla- tion states that: The LACMTA [Metro] shall assess the impacts of the program on commuters of low income and shall provide mitigation to those impacted commuters. Mitigation measures may include, but are not limited to, reduced toll charges and toll credits for transit users. To address the requirements of SB 1422, Metro retained the services of a consulting firm to evaluate the program effects from I-10 and I-110 ExpressLane conversion on low-income commuters. In advance of the HOV to HOT conversion, Metro also agreed to make substantial investments in transit and vanpool services in both corridors using available federal grant funds. Various key themes, analytical consid- erations, and methods for impact assessment of these program elements are detailed in the Metro ExpressLanes Project: Draft Final Low-Income Assessment, LA ExpressLanes Program (Metro, 2010). c a s e e x a m p l e 5 Analyzing, Mitigating, and Monitoring Impacts on Low-Income Commuters, I-10 and I-110 ExpressLanes, Los Angeles County Framework Step • Identify and Assess Mitigation Strategies • Conduct Post-Implementation Monitoring Stages in Decision-Making • Project Design and National Environmental Policy Act • Implementation Tools and Techniques • Estimation of Low-Income Demand • Low-Income Assistance Program • Transit Rewards Program • Revenue Recycling for Transit • Performance Measures • Program Monitoring Affected Populations • Low-Income • Minority Participants • Los Angeles County Metropolitan Transportation Authority

304 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox This report provided a foundation for Metro to establish its Equity Plan Account [currently called the Low-Income Assistance Plan (LIAP)] for low-income commuters, offering eligible Los Angeles County residents a one-time $25 toll credit and an automatic waiver of the $1 monthly account maintenance fee. LIAP account holders must be residents of Los Angeles County and have an annual household income of less than twice the federal poverty level. Currently, this amounts to slightly more than $39,000 per household. How It Was Done In framing the objectives of the low-income assessment report (Metro, 2010), the Metro con- sultant briefly introduced various dimensions of equity often explored in transportation research, including individual, group, geography, market, opportunity, and outcome (Taylor and Norton, 2009). The report explained that its primary focus was on “group equity”—ensuring that low- income commuters as a group are not being disadvantaged or “priced out” of using the toll lanes or, if they are, that mitigation is provided. Additionally, the report noted its concern for “market equity,” which refers to ensuring that shares of benefit are in proportion to taxes or other charges paid because the financial burden of tolls should not exceed the value of travel time savings. With this broad frame in mind, the report estimated low-income commuter demand and assessed potential impacts on low-income users from the CRD program implementation, exploring the potential effectiveness of select mitigation solutions that Metro was considering. The study’s primary focus was on travel behavior impacts rather than assessing a wider range of socio-cultural, air, noise, and other natural environmental impacts of the CRD implementation program that might be expected topics of an environmental impact study. The report first described how the term “low-income” was defined and using four distinct methods, estimated the potential demand for the ExpressLane corridors by low-income com- muters. These methods, Metro’s assessment of the potential impacts on low-income com- muters, and potential forms of mitigation are described in greater detail below. The role that various performance measures could play in monitoring the effectiveness of the mitigation strategies is also discussed. Defining Low-Income. The authorizing legislation (SB 1422) explicitly mandated that eligi- bility requirements for “low-income” toll credits should be set at a level no lower than five other referenced state and local programs serving the needs of low-income populations. In response to this requirement, Metro’s consultant compared existing eligibility thresholds set by these pro- grams (see Table 1) as well as benchmarked other Los Angeles County programs, planned or in use, such as the Metro Rider Relief Program for low-income transit users. Following this review, the report recommended a threshold of $35,000 (in 2009 dollars) based on an annual income for a household of three persons, which was close to the Los Angeles County average. In providing a rationale for the threshold, one that was double the federal poverty level, the report compared the recommendation against other recent studies, programs, and modeling tools (see text box, Defining Low-Income for Program Definition and Assessment). Estimating Low-Income Users of ExpressLane Corridors. With no direct way to deter- mine the number of low-income users of the I-10 and I-110 ExpressLanes, four different data sources and methods were examined to make an estimate of low-income user demand. U.S. Census Data. The first method examined information prepared by the Southern Cali- fornia Associations of Governments (SCAG) on commuting modes by income group derived from the 2000 U.S. Census. This census analysis showed that approximately 22 percent of the region’s commuters who drive alone are in the lowest two-fifths of the income distribution while 27.5 percent of carpoolers and 49 percent of transit users come from this income segment.

analyzing, mitigating, and monitoring Impacts on low-Income commuters, I-10 and I-110 expresslanes, los angeles county 305 Defining Low-Income for Program Definition and Assessment The low-income assessment report (Metro, 2010) benchmarked eligibility levels for several state and local programs and examined assumptions used in the region’s travel demand modeling to recommend a low-income Metro of $35,000 (in 2009 dollars), noting: • It was higher than any of the legislatively prescribed eligibility levels listed in Table 1. • It was higher than the County Supervisors’ recently adopted eligibility levels for State Assistance Programs (e.g., eligibility for a household of four is $27,564 for food stamps) • It was higher than Metro’s Rider Relief eligibility thresholds for households of three persons or less (i.e., eligibility for a household of three is $33,300). • It was approximately equivalent to 200 percent of the federal poverty thresh- old as calculated by SCAG based on the 2000 U.S. Census data and 200 percent of the federal poverty threshold was the low-income threshold used by a 2005 Fast and Intertwined Regular (FAIR) lanes study in Alameda County, California. • It was slightly higher than the income stratification in the travel demand model used for the project. • It was equal to an income stratification used in the Metro HOT lanes license plate survey completed June 2009. • It fell between two income levels used in the Metro HOT lanes stated preference survey ($30,000 and $50,000). • It was below the levels used in the 2008 General Public and Environmental Justice surveys for the Metro HOT lanes project. Table 1. Annual income eligibility requirements for other assistance programs. Household Size SB 1422 Referenced Programs Other California Work Opportunity and Responsibility to Kids Act (CalWORKS) County Aid and Relief to Indigents (Los Angeles County) State Supplementary Assistance Programs for Aged, Blind, Disabled Food Stamps and Food Assistance Metro Rider Relief Program 1 $6,384 $7,440 $10,680 $13,524 $25,000 2 $10,464 $9,726 $18,528 $18,204 $29,600 3 $12,960 $10,188 $22,884 $33,300 4 $15,384 $12,396 $27,564 $37,000 Year Dollars 2009 2009 2009 2009 2008 Source: Network Public Affairs, 2009

306 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox MPO Survey Data. The second data source was SCAG’s 2005 State of the Commute survey, a telephone survey of full-time workers in 2,800 households throughout the SCAG region. The survey indicated that 25 percent of the respondents fell below the $35,000 recommended eligibility requirement for households. The survey reached regular users of the SCAG region’s transportation system, but was not specific to the catchment region for users of the ExpressLanes corridors. License Plate Survey. The third source of data came from a license plate survey targeted to the users of the ExpressLanes corridors. The survey concluded that 18 percent of users had household incomes below $35,000 and 30 percent had household incomes below $50,000 per household. However, the sample population was deliberately oversampled with peak-period and HOV users to capture the opinions of key user segments and their potential behavioral changes. Thus, the sample was designed to have about 70 percent peak users and 30 percent off-peak users, split evenly between HOV and general purpose lane users. The sampling method was also vulnerable to non-response bias. Some indication that respondents may not have been fully representative can be found by tracking the language of the returned surveys; only 5 percent of the completed surveys used the Spanish-language version, while 95 percent of the returned surveys were submit- ted using the English-language version. Travel Demand Model. SCAG’s regional travel demand model was a final source used to estimate low-income users for the project. Trip origins by travel analysis zone (TAZ) during the morning peak period were overlain with SCAG demographic data on household income at the TAZ level to estimate the incomes levels of corridor users. However, as noted by the Metro consultant, this approach is susceptible to the “ecological fallacy” criticism—drawing general- izations about individuals in an area based on the overall characteristics of the area. In fact, the percentage of low-income users by this method was estimated to be higher than the other three prior methods, potentially demarcating an upper bound. Table 2 compares the corridor-specific results using the license plate survey and the travel demand model approach that focuses on TAZ origin demographics. Each of the four methods produced a different estimate of low-income users. After considering the pros and cons, the license plate survey was determined to be the most useful data source and method for estimating the likely percentage of corridor users who would qualify as low-income. Data Collection Method and Thresholds I-10 General Purpose I-10 HOV Total I-110 General Purpose I-110 HOV Total License Plate Survey Below $35,000 17% 21% 11% 25% Below $50,000 30% 33% 24% 34% Travel Demand Model TAZ Origins Up to 200% of Poverty Level * 36% 34% 35% 41% 37% 40% Two Lowest-Income Quintiles ** 45% 43% 45% 50% 46% 49% Source: Metro, 2010 Notes: *Corresponds to the proposed $35,000 low-income threshold; **Corresponds to the higher potential threshold of $46,513. Table 2. Estimates of low-income users through license plate and travel demand model methods.

analyzing, mitigating, and monitoring Impacts on low-Income commuters, I-10 and I-110 expresslanes, los angeles county 307 While vulnerable to previously discussed sampling biases, the license plate survey exhibited pat- terns generally consistent with SCAG’s Census data analysis of commuting patterns by income group. Additionally, the attributes of the survey enabled the differentiation by corridor and between solo and carpool drivers. Potential “Low-Income” Driver Demand for Toll Credits. Metro also wanted to understand the potential size of the pool of ExpressLane corridor users who might be eligible for a “toll credit”—a one-time setup fee waiver of $25 for the acquisition of a transponder—if this form of mitigation were made available to low-income drivers. To make this estimate, Metro’s consult- ing team examined data from existing user accounts for qualifying carpools and paying vehicles. Drawing on the license plate survey results, the demographic characteristics of HOV lane users and general purpose lane users were then applied to the carpool numbers and the paying vehicle numbers, respectively (see Table 3). Potential Transit Demand for Toll Credits. At the time of the study, Metro was also consid- ering whether to extend toll credits to transit riders who frequently use transit services along the subject corridors. Under the contemplated concept, transit users taking 16 round trips in 60 days would be rewarded with a $5 credit that could be used for transit fares or a toll credit. While eli- gibility would extend to all transit users regardless of income, low-income users were expected to disproportionately benefit based on the demographics of the transit user pool. While not shown here, Metro’s consultant made an estimate of the monetary value of the transit credits based on the existing transit user levels. The results from the cost-benefit analysis revealed that the toll cred- its would reduce the overall net benefit by less than 10 percent, and the program would still yield a positive benefit-cost ratio, confirming the potential viability of implementing the toll credit. Impact on Low-Income Commuters. The likelihood that low-income drivers would choose to use the ExpressLanes was also investigated. In theory, travel behavior can be inferred by con- sidering a driver’s value of time. Economic theory suggests that drivers will use the ExpressLane if they perceive the price of the ExpressLane to be less than their marginal value of time (VOT). After reviewing the transportation literature and guidance, the Metro consultant set the appro- priate VOT at 60 percent of the wage rate for low-income households ($35,000 annual income), leading to a $10.10/hour rate for operating a single-occupancy vehicle. For carpoolers the value of time was assumed to be 1.82 times that value for a total of $18.38/hour. Roadway and Lane Group Estimated Unique Accounts Pool 200% of Poverty Pool Q1+Q2* I-10: General Purpose Lanes 107,000 18,190 32,100 HOV Lane 84,000 17,640 27,720 Total All Lanes 191,000 35,830 59,820 I-110 : General Purpose Lanes 77,000 8,470 18,480 HOV Lane 154,000 38,500 52,360 Total All Lanes 231,000 46,970 70,840 Source: Metro, 2010 Note: *Corresponds to household income below $50,000 per survey category, slightly higher than the sum of two lowest-income quintiles ($46,513). Table 3. Estimate of low-income drivers potentially eligible for toll credits.

308 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox These values were then compared to vehicle marginal VOT as estimated by a toll optimization model (TOM). The TOM calculated vehicle marginal VOT for small segments of the roadway and weighted them based on vehicle miles traveled on each segment. The vehicle marginal VOT ranged from as low as $24.95 to as high as $67.78 for a single-occupancy vehicle but illustrated that as priced there was no situation in which the ExpressLane would be desirable to a low-income single-occupancy vehicle at the $10.10/hour or even at the $18.38/hour VOT for carpoolers. While grounded in mathematical rigor, the Metro consultant noted that the method, as presented, did not fully account for specific circumstances in which a low-income driver’s VOT might increase significantly, leading to a willingness to take the tolled ExpressLane (e.g., being on time for work, picking up a child from daycare, emergency healthcare). In exploring the sensitivity of this toll payment choice, it was determined that it would take a “toll credit” and a much higher VOT (i.e., 120% of the wage rate rather than the 60% level in the base case example, perhaps suggestive of an urgent situation for the low-income driver) to find a price point at which the HOT lane would be attractive. Transponder Issues. To inform the setting of transponder-related policies, the Metro consul- tant noted several potential barriers reported in the transportation research literature to the usage of transponders by low-income users, including the cost of a transponder, setting up an account, methods of payment that favor credit card over cash replenishment, minimum required account balances, and minimum monthly charges. National and regional studies showing the relation- ship between lower family incomes and the use of bank accounts and credit cards were presented to further frame the analysis. Select policies of other operating toll systems in California at the time of the study (e.g., I-15 and SR-91) were also examined to help formulate recommendations for designing a toll system responsive to low-income commuters that waives or reduces specific charges, among other strategies. Outreach Processes. In addition to the quantitative assessment of impacts on low-income commuters, a robust outreach effort was undertaken to better understand the transportation needs of the low-income population. The first and most important goal of the outreach effort was to identify the modes of travel by low-income population groups. Input generated during the outreach process suggested that low-income individuals use the I-10 and I-110 corridors, but most carpool to reduce transportation costs. Consequently, most low-income commuters asked for better transit options that drove many of the CRD program elements, including significant investment in capital and operational improvements to transit. Low-income users consulted during the outreach process also requested that the upfront cost to enter the program be reduced to facilitate use of the corridors when absolutely essential. Mitigation Strategies. Informed by quantitative analyses suggesting that low-income com- muters are effectively priced out of the express lanes as well as low-income commuter travel behavior and attitudes toward the toll lanes and transponder access, Metro decided to implement innovative programs to mitigate the upfront costs of using the ExpressLanes and provide credits for regular transit users along the corridor. The LIAP allows qualifying Los Angeles County residents a one-time $25 credit when applying to be an ExpressLane user. Eligibility for the program follows the income threshold levels by household size shown in Table 4. The credit can be applied to either the transponder deposit or prepaid toll deposit. In addition, Metro waived the $1 per month maintenance fee for qualifying participants. The qualifying criteria make it easy for participants eligible for other low-income programs to qualify. Applicants are required to present a paycheck stub, current tax return, or proof of enrollment in one of several existing low-income programs (see text box, Low-Income Assistance Plan). Individuals can apply for LIAP with cash, checks, or reloadable Visa cards at two service locations as well as online. The LIAP program is marketed through numerous venues, including billboards, food courts at shopping malls, radio programs, local papers, and cable TV geared to limited English proficiency populations.

analyzing, mitigating, and monitoring Impacts on low-Income commuters, I-10 and I-110 expresslanes, los angeles county 309 Originally, the Transit Rewards Program (TRP) was instituted as a way to reward its transit users with a $5 toll credit for every 32 one-way trips. Any individual with a Transit Access Pro- gram (TAP) card can register, but the program was expected to be particularly valuable to the low-income commuters who use the transit system more often. Following a recent review of the program that suggested low usage, the threshold number of one-way trips for rewards eligibility was reduced by half to 16 to earn the $5 toll credit. In addition, Metro is currently in the process of implementing a new policy to allow the credit to be for both the ExpressLanes and transit usage. Recommended Performance Measures. The Metro consultant recommended several oper- ational performance measures that could be used to monitor the effectiveness of the tolling program during and after the demonstration period, explaining the purpose, data collection methods, and timeliness of the measures (see text box, Recommended Performance Measures). Metro continues to collect data for most of the recommended performance measures to support program review. This information is acquired from the toll vendor (Xerox), ridership data from the transit agency, and data collected by the toll consultant on operational measures, including Household Size Income Threshold** 1 $23,540 2 $31,860 3 $40,180 4 $48,500 5 $56,820 6 $65,140 7 $73,460 8 $81,780 For each additional person, add $8,320 Source: Metro, 2010 Note: ** This threshold is twice the 2015 Federal Poverty Level. Table 4. Low-income assistance plan eligibility. Low-Income Assistance Plan The Low-Income Assistance Plan (LIAP) requires that participants be residents of Los Angeles County. Residents have to meet the income threshold requirements by household size. Several documents may be submitted along with an applica- tion to provide a proof of income-eligibility, including: • Paycheck stub • Public benefit verification • MediCal participation • Tax return • Lifeline user • LAUSD lunch recipient • Electronic benefits transfer card Source: Metro, 2010

310 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox travel speeds. Supplementing these measures, the Metro consultant also recommended the use of targeted surveys to registered low-income corridor users during and following the demonstration period to gauge perceptions of congestion levels as well as the perceived ease of use of accounts and payment options, among other issues. The use of surveys and focus groups are part of the overall market research efforts undertaken by Metro to explore how user perceptions may differ from operational data and metrics. Recommended Performance Measures Several potential performance measures for considering effects on low-income users were proposed by Metro’s consultant, including: 1. Number of low-income commuters [including percentage of Transit Access Program (TAP) users] who sign up for a transponder. Data can be collected from toll operators as well as from Metro’s collection of transponder distribu- tion and from TAP program records. 2. Number of peak-period low-income users of HOT lanes (and percentage of overall HOT lane users). Data can be tracked from transponder records col- lected by the toll vendor. 3. Usage of HOT lane credits for low-income drivers (credit redemptions). Data can be collected from TAP records and transponder account records since these are linked. 4. Mode choice of low-income drivers (carpool versus single-occupant vehicle), compared with mode choice before the project is implemented. Data can be collected from transponder account records before and after the ExpressLanes are implemented to see if there are mode shifts. 5. Performance of transit service in the ExpressLanes corridors during the dem- onstration period. Average speed, trip time, time savings, and trip reliability of transit buses in both corridors can be tracked using transit operations data. 6. General purpose lane speeds during the demonstration period. Operational data from Metro and Caltrans can be collected to evaluate time savings and trip reliability of the general purpose lanes because low-income commuters would still benefit from a reduction in congestion levels in the general pur- pose lanes with the implementation of the HOT lane. 7. Account balance problems of low-income commuters compared with non- low-income. Data can be collected through Metro’s toll operator to show how many registered low-income commuters encounter problems maintain- ing a required minimum balance in their transponder accounts. 8. Share of time savings by low-income ExpressLanes drivers compared with the share of tolls and transponder costs they pay. Data could be collected for low-income drivers at entry and exit points for vehicles with a transponder. The travel time savings can be calculated and compared as a percentage of toll costs. 9. Trends in trip distance and trip time by low-income commuters compared with non-low-income. Data can be collected from operational data for transponders of low-income drivers to non-low-income drivers. 10. Toll revenue investment. Data can be collected through Metro accounting records and reports to show where revenues are being invested in the corridor. Source: Metro, 2010

analyzing, mitigating, and monitoring Impacts on low-Income commuters, I-10 and I-110 expresslanes, los angeles county 311 Resources and Costs The resources required to prepare the low-income assessment report (Metro, 2010) are not significant or time-consuming in light of possible equity-related impacts and the importance of addressing equity considerations for sustaining popular support for the HOT lane conversion. The Metro consultant was able to make extensive use of regional and corridor-specific travel demand models, travel survey data, and a tolling optimization model that were developed in planning the project for purposes of exploring effects on low-income users. However, a similar assessment may be difficult for some smaller urban areas to replicate if this type of data and models are not readily available. The resources required to maintain the LIAP and TRP appear relatively small (less than a full- time position). To keep costs down and encourage participation, Metro accepted the eligibility of participants already qualified for other low-income programs. Moreover, qualification for program eligibility to date has only been verified at the time of the initial application process rather than on an annual or periodic basis. Benefits: Why It Was Effective The Low-Income Assessment Report—drawing on other modeling and market research— suggests a scope of upfront research considerations and subsequent post-deployment commit- ments to performance evaluation that can be employed to identify and address travel behavior differences by income groups. The quantitative and qualitative methods are generally replicable and could be undertaken by other sponsoring agencies prepared to comprehensively analyze the equity distributional effects of tolling programs. Metro’s commitment to developing mitigation strategies through transit service investments and toll credit programs also suggests ways for- ward for other sponsoring agencies seeking to reduce the potential effects on low-income users. Metro’s adoption of a performance evaluation framework with equity considerations ensures that the agency is transparent in sharing information with interested stakeholders and exploring possible ways to adjust the program to serve its low-income users. Challenges and Limitations Several methodological challenges and limitations of prospectively assessing the impact of the HOT lane conversion on low-income populations have already been described in previous sections. Metro, in implementing the HOT lane conversion, also accepted the challenge of insti- tuting a post-implementation data collection protocol to monitor the effectiveness of its pro- gram, including its low-income mitigation strategies. Specifically, the federal grant required formal evaluation of the pilot period based on 12 months of data collection; the state law also required a performance report of the demonstration period. After the opening of the Metro ExpressLanes on the I-110 freeway in November 2012 and the I-10 in late February 2013, data collected from Metro’s tolling vendor and other sources began to accumulate to support the formal evaluation of the CRD program. One of the challenges that Metro has experienced is keeping a dedicated person on staff to track data for performance reporting. The data and resulting reports suggest ways in which program implementation effects can be shared and assessed retrospectively by sponsoring agencies and affected stakeholders with due consideration to its low-income users, among other issues. For example, Metro tracks several data points, including the number of individuals signed up since program inception. One lingering question to date has been whether or not some 8,882 plan accounts enrolled in the LIAP (end of 2015) was a “good” number of enrolled participants in light of the estimated 1.349 million poten- tial eligible low-income households in Los Angeles County in early 2014. Metro has used market research studies in the form of surveys, focus groups, and intercept surveys (see text box, Metro

312 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Metro ExpressLanes Market Research Studies Metro has employed a three-pronged approach to evaluate and improve the effectiveness of the Low-Income Assistance Plan (LIAP), including a survey of LIAP users, two rounds of focus groups, and intercept surveys. In 2013 Metro conducted an online survey of all LIAP account holders distributed via email (83% of account holders), offering a $10 toll credit to respondents, and received back responses from 21% of its account holders. The survey consisted of 18 multiple choice questions as well as open-ended and demographic questions. The survey explored the respondents’ perceptions of the accessibility of the program, their knowledge of the pro- gram, and overall satisfaction. The typical respondent was between 25 and 44 years, with household incomes below $35,000. Most had three or more people in the household with two cars. A vast majority (87%) rated their experience as good or excellent and said they would recommend the program to family and friends. Most respondents learned about the program through the Metro website and signed up primarily to get the toll credit. In the spring of 2014 and 2015, Metro conducted two sets of focus groups to understand customer attitudes of the LIAP, including benefits, accessibility, and importance. Four one-hour focus group sessions were held with 4 to 6 individuals (selected from LIAP survey respondents). The general results of the focus group sessions were similar for both the 2014 and 2015 sessions, and included the recommendation to advertise the program within the community especially at locations frequently visited such as gas stations, churches, schools, and social service outlets. Participants included a mix of carpoolers and single-occupancy drivers who use I-10 more than other corridors. Most participants were employed, rented their housing, had at least a high school educa- tion, and represented different backgrounds and professions. Most participants learned about the program through the Metro website and believed the program benefits should be better advertised through social media outlets. They appreciated the $25 toll credit and the availability of the cash option and waived main- tenance fee. An interesting result of the 2014 focus group input was the change in name from Equity Plan (used originally) to LIAP because users believed the term equity was confusing and not indicative of the intent of the program. Intercept Surveys were fielded in October 2015 to understand the perceptions of non-users primarily. In addition, the survey was designed to better understand how low-income populations think about travel and included questions about attitudes and lifestyle trends. The targeted impact areas for field collection were chosen from areas that revealed low usage of LIAP. Zip codes along the 11-mile ExpressLanes corridors with current LIAP users were mapped against low-income populations to compare areas with the highest potential usage of low-income commuters to LIAP account holders. Census-based race and ethnicity information was collected for these areas to identify target response rates proportionate to population demographics. Surveys were admin- istered in English, Spanish, and Mandarin over a four-day period with most respondents choosing to fill out a hard copy rather than a web-based survey. Survey participants received a $10 gift card for responding, and a total of 450 surveys were completed (22% African American, 22% Asian, 11% white, and approximately half Spanish-speaking). Most of the respondents were female with a high school education and some college with incomes ranging from $10,000 to $51,000 with one or more cars in the household. Most lived in a home with at least four or more persons and rented their homes with the largest percentage of household income spent on housing. Some of the most interesting survey findings include: • Respondents recommended using TV as a means to disseminate information on the Metro ExpressLanes. • To learn more about products and services, respondents indicated that they were most likely to perform online searches, followed by TV, then family and friends. • More than half of the respondents agreed with the statement, “I can’t afford to take the ExpressLanes.”

analyzing, mitigating, and monitoring Impacts on low-Income commuters, I-10 and I-110 expresslanes, los angeles county 313 ExpressLanes Market Research Studies) to help evaluate the effectiveness of the LIAP. While the number of LIAP accounts has doubled in 2 years, Metro is committed to ongoing efforts through additional market research studies to encourage more use of the ExpressLanes. Metro’s market research studies have indicated a relatively low awareness of the low-income programs among its customers; however, information collected from the market studies continues to guide policy changes such as lowering the number of transit trips from 32 to 16 to receive a $5 toll credit and allowing the credit to be used for transit. Metro’s market research program also included an Equity Plan Survey of its low-income account holders at the close of its first year of operations. The survey explored customer satis- faction with the program and confirmed that the Equity Plan—now known as the LIAP—was important for more than 82 percent of the respondents in making the decision to get a trans- ponder account to use the ExpressLanes. The spatial distribution of transponders (FasTraks) in the Los Angeles metropolitan area was closely examined at the zip code level as part of the formal evaluation of the pilot period (see Figure 1). The analysis found that there were fewer transponder accounts in areas with low median household incomes and high rates of poverty—a pattern that the researchers concluded might be expected given lower rates of car ownership in those areas. A high cor- relation was also observed between proximity to the ExpressLanes and the concentration of accounts, suggesting that the observed distribution of accounts could not be explained solely as a function of income. Figure 1. Transponder accounts were mapped (shown here) along with LIAP accounts and overlain with income and poverty maps to evaluate usage patterns.

314 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox While not shown here, the research also found a higher percentage of LIAP accounts in areas with lower median household incomes and higher rates of poverty. In many cases, the areas with higher percentages of equity plans were in a lower-income area where fewer individuals obtained a transponder. Lessons Learned One significant lesson of the Metro ExpressLanes project is that a quantitative analysis of low-income user impacts can be valuably informed by implementing a range of survey and focus group methods. Surveys and focus groups offer a means for more direct measurement of attitudes toward tolling and travel behavior and, when properly designed in terms of sample and questions, support assessment of socioeconomic differences along income, race, language- proficiency, and other factors. These tools can be used before and after the implementation of tolling changes to gauge the awareness of pricing and transit alternative options, usage levels, and mitigation strategies and to assess whether policies are being responsive to all segments of the population that travel along the corridor. The Metro project is noteworthy, in part, for the agency’s exploration of relevant performance measures and its commitment to program monitoring after implementation to mitigate equity- related impacts. The Metro project was funded with a federal CRD grant and was required to address equity considerations, but the agency’s actions were also informed by state legislation that sought to focus on the effectiveness of its mitigation strategies. The performance evaluation framework and the agency’s ongoing market research program have brought continuing atten- tion on equity effects and the potential remedies for its low-income users. The thoughtful approach of defining the threshold for low-income households is worthy of replication by other transportation agencies. The outreach program that Metro uses is also replicable and could be highly effective at identifying the challenges faced by low-income com- muters in accessing HOT lanes. Resources Battelle. 2015. Los Angeles Congestion Reduction Demonstration (LA CRD) ExpressLanes Program. National Evaluation: Technical Memorandum on Congestion, Tolling, Transit, and Equity Results. FHWA-JPO-14-126. Prepared for the U.S. DOT. California Bill Number, SB 1422, as amended. Retrieved from ftp://www.leginfo.ca.gov/pub/07-08/bill/sen/ sb_1401-1450/sb_1422_bill_20080825_amended_asm_v96.html. Metro. 2010. Metro ExpressLanes Project: Draft Final Low-Income Assessment, LA ExpressLanes Program. Metro. 2014. Metro ExpressLanes: Equity Plan Survey Analysis. Metro. 2014. LA County Congestion Reduction Demonstration Program: ExpressLanes Equity Plan Account Income Threshold. Memorandum of the Ad Hoc Congestion Pricing Committee. Network Public Affairs. 2009. Low-Income Impact Assessment, Methodology Memorandum, Los Angeles Congestion Reduction Demonstration Project. Prepared for Los Angeles County Metropolitan Transportation Authority. Noble Insights. 2015. Low-Income Research Report. Prepared for Los Angeles Metropolitan Transportation Authority. Noble Insights. 2015. Metro ExpressLanes: Low Income Field Surveys. Prepared for Los Angeles Metropolitan Transportation Authority. Noble Insights. 2014. Equity Research Report. Prepared for Los Angeles Metropolitan Transportation Authority. Taylor, B. D., and Norton, A. T. 2009. “Paying for Transportation: What’s a Fair Price?” Journal of Planning Literature, Vol. 24, No. 1, pp. 22–36.

315 Background In 2014, the North Central Texas Council of Governments (NCTCOG) proposed an alignment for the Chisholm Trail Parkway. The new parkway would operate as a toll road, running near the exist- ing Old Granbury Road right-of-way in Fort Worth, Texas. The pro- posed toll route, shown in Figure 1, would prevent through traffic on Old Granbury Road, which was the existing route used by residents of St. Francis Village, a retirement community with a large number of low- income households near the alignment. Alternative non-tolled routes would more than double the residents’ travel distance and travel time. In NCTCOG’s long-range planning process and regional tolling analy- sis, it was determined that siting the new toll road would adversely affect roadway access for residents of the St. Francis Village. As part of Texas state law, a reasonable free alternative route must be available in addition to a toll route. While drivers from other origins would be allowed a reason- able free alternative via the newly constructed Summer Creek Drive and Cleburne Road, the location of St. Francis Village would require additional drive time along McPherson Boulevard prior to reaching the free route. One of the proposed alternatives for mitigating this adverse impact involved the construction of an overpass to allow Old Granbury Road to remain a through street, supplying a reasonable non-tolled alterna- tive. However, this option was determined to be cost prohibitive because it would cost millions of dollars more than the original project. Instead, an intervention was proposed that involved using government funds for the purchase of prepaid toll tags for the residents. In 2014, NCTCOG’s Regional Transportation Council (RTC) agreed to sponsor North Texas Tollway Authority’s (NTTA) TollTags as a means to mitigate the community’s reduced access. How It Was Done Every household was supplied with a TollTag pre-loaded with a credit of $250. Sponsorship was limited to one TollTag per household. To receive the credit, each household had to self- certify their financial need to receive a sponsored tag. The funds added to the card were only eligible for use on toll roads in the Dallas–Fort Worth region. If the household already had a TollTag, a credit of $250 was simply added to the user’s account. Additionally, residents were able to link a credit card/debit card to the account so that the TollTag could be used for parking at the Dallas–Fort Worth International Airport or Dallas Love Field. c a s e e x a m p l e 6 Mitigating Reduced Access via Toll Credits, Dallas–Fort Worth Metro Region Framework Step • Identify and Assess Mitigation Strategies Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act Tools and Techniques • One-Time Toll Credit Affected Populations • Low-Income Participants • North Central Texas Council of Governments • North Texas Tollway Authority • Tarrant County • St. Francis Village Retirement Community Residents

316 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Outreach activities were an integral element in the development of the mitigation approach. In April 2014, representatives from Tarrant County, NCTCOG, and NTTA met with the residents of the community to present information on how the program would work and answer questions. NTTA and NCTCOG staff were also present a week later to provide residents a second opportunity to sign up. Table 1 summarizes the number of sponsored TollTags requested and processed. Resources and Costs The total budget allotted for this mitigation approach was approximately $100,000. The cost of providing the pre-loaded TollTags to the residents ended up totaling $67,750, excluding administrative, planning, and project development costs. Figure 1. Chisholm Trail Parkway and St. Francis Village. Source: NCTCOG, 2014a

mitigating Reduced access via Toll credits, Dallas–Fort Worth metro Region 317 Benefits: Why It Was Effective Mitigation reduced the costs for residents accessing the toll road and enabled them to avoid a free alternative that would have nearly doubled their travel time. The program only cost $67,750 versus the millions of dollars that would have been required to build a new road to supply the residents with access to an alternative non-tolled route. Challenges and Limitations One of the limitations of the mitigation approach was that residents were required to sign up for a TollTag to receive free toll road access. Reimbursements in the form of cash or checks were not given for prior toll usage. To be eligible, residents were given a period of only about 3 weeks to sign up for the one-time, $250 value, non-recurring TollTag credit program. While it may be extended, the program currently has a time limit of 3 years to draw down toll credit funds. The rationale for setting the credit at $250 in the mitigation development process, or whether it was sufficient to satisfactorily mitigate impacts for its affected residents, has not yet been fully assessed. Lessons Learned Even though this particular solution was triggered by considerations and adverse impact determinations established under specific state law, this example suggests a form of mitigation that has a relatively low cost to implement and maintain. It is a method of economic mitiga- tion that enables specific low-income neighborhoods or communities—burdened by the barrier effects or isolated by tolling facilities—to have the same level of access to roads, and therefore to jobs and services, as those of greater means. Resources NCTCOG. 2014a. “Sponsorship of NTTA Toll Tags as an Abatement of Reduced Access.” Retrieved from http:// www.nctcog.org/trans/committees/rtc/Item_4.rtc010914.pdf. NCTCOG. 2014b. “Sponsorship of NTTA Toll Tags as an Abatement of Reduced Access: Finalizing the Partner- ship with Tarrant County.” Retrieved from http://www.nctcog.org/trans/committees/rtc/documents/ Item_7.rtc111314.pdf. Table 1. Number of sponsored TollTags created by mitigation. TollTag Outreach and Support Activities New TollTag Accounts Existing TollTag Accounts New TollTag Accounts that Attached a Credit/Debit Card Totals April 3 and 10 meetings 216 25 19 260 After the meetings (via mail or calls) 10 1 0 11 Total 226 26 19 271 Source: NCTCOG, 2014a

318 Background The Florida DOT identified the need for a new limited-access high- way corridor and bridge-crossing for the St. Johns River between Clay and St. Johns counties in Florida. The St. Johns River separates the two counties that are only connected, within the defined project area, by the Shands Bridge as shown on Figure 1. Florida DOT identified ten build alternatives with the ability to provide additional capacity, promote and support employment and economic development, and improve emer- gency evacuation. All build alternatives include a new bridge across the St. Johns River. The black and purple alternatives would cross the river north of Green Cove Springs leaving the Shands Bridge in place. The eight southern alternatives would cross the river near the existing Shands Bridge, which would be removed once the construction of the new bridge was com- pleted (see Figure 1). The build alternatives presented in the environmental impact study were informed by prior planning level, corridor level, public scoping- related studies, and desktop and screening studies. The build alternative began as early as 2002 to refine the alternatives and clarify key issues and decisions. In August 2006, Florida DOT decided to design and imple- ment the project as a limited-access tolled facility from Branan Field- Chaffee Road to the 9B interchange with I-95, including the construction of a new bridge over the St. Johns River. Most of the residents of southern Clay County are low-income and low-income communities are located within or adjacent to the study area of the southern alternatives (Brown 1 and 2, Green 1 and 2, Orange 1 and 2, and Pink 1 and 2 Alternatives) as shown in Figure 1. No low-income or minority populations in the study area are located along the black and purple alternatives. How It Was Done The potential effects of tolling, among other issues, on low-income populations were assessed in the EJ analysis of EIS. Florida DOT evaluated the impacts of three key aspects: (1) tolling tech- nology, (2) toll rates, and (3) the availability of reasonable, non-tolled alternate routes. c a s e e x a m p l e 7 Selecting a Design Alternative to Mitigate the Adverse Effects of Tolls on a Low-Income Community, St. Johns River Crossing, Clay and St. Johns Counties, Florida Framework Step • Identify and Assess Mitigation Strategies Stages in Decision-Making • Policy and Planning • Project Design and the National Environmental Policy Act Tools and Techniques • Exempt Local Traffic from Toll on New Bridge • Cash Replenishment Option Affected Populations • Low-Income Participants • FHWA • Florida DOT

selecting a Design alternative to mitigate the adverse effects of Tolls, st. Johns River crossing, clay and st. Johns counties, Florida 319 Tolling Technology. The St. Johns River Crossing toll road, as proposed, will operate as a fully electronic toll facility. Vehicles will be required to have transponders to use the facil- ity, and toll gantries placed along the corridor will read the transponders to identify where the vehicle entered and exited the toll facility. This information will be used to determine the toll to be charged for the trip. Each transponder will be registered to a specific user whose account will be charged for the toll. On behalf of Florida DOT, a contractor will be hired to manage the daily operations of the tolling facility. The EJ analysis found that maintaining a prepaid account balance on the toll account and the need for a credit card to replenish the account could have disproportionate impacts for low-income populations. Toll Rates and Alternate Routes. The Final EIS noted that the actual toll rates were unknown but the financial advisors estimated that the toll rates for the project would be between $0.15 and $0.20 per mile in the opening year. Assuming the higher rate of $0.20, the estimated cost would be $2.30 per trip, including a $1.00 surcharge for the new bridge-crossing. The Final EIS also disclosed that the actual toll rate would be established at the time of bid by the potential private partners (i.e., concessionaires). The conditions of bidding documents have not been set but will Figure 1. Environmental justice areas. Source: FHWA and Florida DOT, 2013

320 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox generally specify the requirements of the concessionaire to design, build, operate, maintain, and finance the project for a term of 50 to 75 years (FHWA and Florida DOT, 2013). The EJ analysis found that, in general, tolling has the potential to create a higher economic impact on low-income users because the cost of paying tolls will represent a higher percentage of household income. Under some of the subject alternatives, drivers could choose to reduce the user costs and the financial burden of tolls by using existing non-tolled roads. For example, under the black and purple alternatives, drivers could use the existing Shands Bridge which is currently a non-tolled crossing. The non-tolled alternate travel routes would not add substantial travel distances; there- fore, these alternatives were not found to be significantly high and adverse for local low-income populations. However, this choice would not be available under the southern alternatives because the Shands Bridge would be replaced by a toll bridge without a nearby alternate non-tolled route. In fact, under the southern alternatives, the alternate route to cross the St. Johns River without pay- ing a toll would be approximately 37 miles longer than the existing route, which was considered to be unreasonable and would result in an impact that would adversely affect the low-income populations in the study area. Proposed Mitigation. Because the southern alternatives did not provide viable non-tolled alternate routes, disproportionately high and adverse effects on low-income populations in the project area were anticipated. To avoid these impacts, Florida DOT made the decision to elimi- nate the bridge-crossing toll for the southern alternatives for trips using the toll road solely to cross the St. John’s River. Florida DOT made the determination to eliminate the bridge toll for the southern alternatives for the segment between the US 17 interchange and the CR 16A inter- change in either direction. Thus, drivers would need to exit the route at SR 16 or SR 17 to avoid paying a toll. The proposed non-tolled local trip route is shown on Figure 2. Florida DOT also made a commitment to allow drivers to set up and maintain a transponder account without use of a credit card. These mitigation commitments were introduced to minimize impacts to what was deemed a less than significant level (FHWA and Florida DOT, 2014). Preferred Alternative. After public input, agency coordination, and extensive engineer- ing and environmental analysis, the recommended alternative for the St. Johns River Crossing Project was the Pink 1 Alternative (see Figure 1). The Pink 1 Alternative (one of the proposed southern alternatives) proposed the construction of a new 31-mile limited-access highway from SR 21/Branan Field-Chaffee Road interchange in Clay County to I-95 in St. Johns County. This alternative includes the construction of a new bridge parallel and to the south of the existing Shands Bridge, which will be removed once the construction of the new bridge is completed. Resources and Costs In 2006, Florida DOT conducted a preliminary toll revenue and feasibility study for the proj- ect that determined that the St. Johns River Crossing Project was not toll-feasible as a stand- alone project under any of the build alternatives. After these findings, Florida DOT decided to combine this project (for tolling purposes) with the Branan Field-Chaffee Road (SR 23) Project. The combination of the two independent projects makes the St. Johns River Crossing Project toll-feasible. Florida DOT developed project cost estimates for construction, engineering and inspection, design, right-of-way acquisition, and wetland mitigation for each of the final build alternatives. The total estimated costs ranged from $2.5 billion for the black alternative to $1.9 billion for

selecting a Design alternative to mitigate the adverse effects of Tolls, st. Johns River crossing, clay and st. Johns counties, Florida 321 the preferred alternative (Pink 1 Alternative), including the demolition of the existing Shands Bridge. Florida DOT concluded that the project could not be implemented using traditional methods of finance, so it determined that it needed to design and implement the project as a tolled facility. Benefits: Why It Was Effective To avoid a disproportionately high and adverse financial burden on low-income popula- tions from tolling, mitigation in the form of a non-tolled bridge was recognized as a means for reducing disruption in mobility across the St. Johns River that could be caused by charging low- income residents for a local trip. The approach thus preserved and maintained the economic and social linkages that the crossing provided for local residents. In addition to reducing the potential additional costs for local residents in making local trips, by avoiding toll charges to cross the St. Johns River, it eliminated the need for residents to consider taking a substantially longer non-tolled alternative route. In addition, this type of targeted local mitigation recognizes Source: FHWA and Florida DOT, 2013 Figure 2. Toll-free local trip route.

322 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox and reduces the potential for recurring economic impacts on low-income communities within the study area. Challenges and Limitations The focus of the analysis and mitigation was on the replacement of an existing non-tolled bridge with a tolled bridge and no reasonably alternate non-tolled route. In this case, the alter- nate non-tolled route would have added 37 miles and was found to be unreasonable. However, there is no established standard for defining what is “reasonable” in terms of the additional distance required to be traveled or the additional time required to be spent in travel. Therefore, making the determination of whether a shorter non-tolled route (e.g., 7 or 10 miles) should be considered viable may prove challenging to determine in a different context for those preparing analyses of tolling or tolling diversion related impacts. Establishing a toll-free bridge-crossing to mitigate the financial burden associated with the local trips crossing the St. Johns River is a highly localized strategy to avoid disproportionately high and adverse effects of tolling on the low-income populations living near the existing Shands Bridge. The scope of the approach adopted here was limited to addressing impacts on locally affected communities; any potential user cost or mobility and access-related impacts for low- income populations that are more distant from the existing Shands Bridge were not addressed by the mitigation strategy. The segment of the St. Johns River Crossing project with the new bridge-crossing is in the preliminary design phase with permitting and right-of-way acquisition activities expected to con- tinue through 2020. With the construction phase not yet scheduled, it will be some time before it is operational (Florida DOT, 2015). As a result, it is not yet possible to evaluate facility use and the potential revenue loss from the non-tolled bridge to compare with other mitigation strategies such as issuing toll credits. Understanding the relationship between long-term revenue from the tolled facility and the impact of the non-tolled portion is important to determine whether this avoidance as mitigation strategy could be feasibly implemented on other projects. Lessons Learned This example uses the design of the toll collection points and avoidance as mitigation. Low- income neighborhoods and communities near the bridge are able to avoid tolls for routine bridge crossings for local trips. They are also able to avoid the much longer detour that would be required to avoid tolls. The mitigation preserves access to roads, and therefore to jobs and services. Resources Florida DOT. 2015. First Coast Expressway: About the Project. Retrieved from: http://firstcoastexpressway.com/ about-the-project.shtml. FHWA and Florida DOT. 2013. St. Johns River Crossing. Final Environmental Impact Statement. Retrieved from: http://firstcoastexpressway.com/documents/2013-10-11-st-johns-river-crossing-final-environmental- impact-statement.pdf. FHWA and Florida DOT. 2014. St. Johns River Crossing Project in Clay and St. Johns Counties, Florida: Record of Decision. Retrieved from: http://firstcoastexpressway.com/documents/2014-04-07-st-johns-river-crossing- record-of-decision.pdf. Wilbur Smith Associates. 2010. First Coast Outer Beltway Planning Level Traffic and Revenue Analysis Draft Report— Toll Traffic Forecast with Independent Economist Data. Retrieved from: http://firstcoastexpressway.com/ documents/Client%20FCOB-S2-S3A-S3B-S4%20Report%20041510.pdf.

323 c a s e e x a m p l e 8 Background One tool for determining whether toll roads impact EJ populations differently from other population subgroups is to determine two pieces of information that can be collected from surveys: 1. What do people in the EJ populations themselves think about toll- ing? Surveys can probe variants of this question such as: Do you think [the tolled facility] is a good idea? Do you think it will benefit the community? Do you think it would benefit you/your family? Do you think it is fair? 2. In what ways and to what extent does tolling impact travel behavior choices for members of EJ groups? Do they use the tolled facility at all? Did a new toll or increased toll rate cause them to stop traveling in the corridor? (Or, if the facility has only been proposed, do they think they would use the facility?) This case example reports on survey research conducted for the I-85 Corridor northeast of Atlanta. FHWA commissioned the survey research to document how the express lanes changed corridor travel behavior and attitudes toward tolling. The researchers designed a two-stage household travel diary survey, with surveys administered in two waves, one before the express lanes opened and the second after the lanes were in operation. In this longi- tudinal survey design, known as a “panel survey,” researchers contacted the same set of households for both waves of surveying. Participating households represented people using a variety of modes and routes in the I-85 corridor: people who drove in the corridor (on I-85 or a paral- lel road), vanpooled, or rode transit. Both the Atlanta survey and a partner survey of Seattle commuters traveling in the corridor of a newly tolled bridge exemplify a unique survey research approach for monitoring toll equity. These studies are “the first diary-based study of the impacts of congestion pricing at the household level, including equity impacts” (Ray et al., 2014). Conducting Pre- and Post-Implementation Surveys of Traveler Behavior and Opinions, Atlanta Region, I-85 Corridor Framework Step • Scope Approach to Measure and Address Impacts • Conduct Impact Analysis and Measurement • Conduct Post-Implementation Monitoring Stages in Decision-Making • Statewide and Metropolitan Planning • National Environmental Policy Act and Project Development • Maintenance and Operations Tools and Techniques • Before-and-After Surveys: Travel Behavior and Attitudes Affected Populations • Low-Income • Minority Participants • FHWA • U.S. DOT—Volpe National Transportation Systems Center • Georgia DOT • State Road and Tollway Authority • Georgia Regional Transportation Authority • Transportation Survey Research Firm

324 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox How It Was Done The following sub-sections describe the tolling project under study, research participants, study objectives, study methods, and key findings related to equity for low-income and minority travelers in the I-85 corridor. Description of the I-85 Tolling Project. The I-85 project converted 16 miles of HOV lanes requiring at minimum two persons (HOV 2) near Atlanta into so-called “express lanes” that are free to carpools of three or more persons (HOV 3+) and available for a toll to solo drivers and two-person carpools. The toll varies according to the level of congestion, a so-called “dynami- cally priced” rate. The tolling project changed the experience for carpoolers in two ways. First, eligible carpools use of the lanes went from a minimum of two to three passengers. Second, three plus carpools must have a Peach Pass account and transponder, something that carpoolers did not have to use prior to the implementation of the express lanes. Drivers choose between two types of transponder accounts: the “Peach Pass” and “Pay n GO Peach Pass.” To use a Peach Pass, drivers must link the account with a credit card or bank account that will be used to replenish the account. For drivers who cannot or do not wish to link their account with a credit card or bank account, the Pay n GO Peach Pass can be purchased and replenished with cash. Pay n GO Peach Pass account holders cannot switch between the non-toll (carpool) and tolled modes, but Peach Pass account holders can. However, Peach Pass account holders must plan ahead and have access to an internet-connected device to switch between the toll-exempt (carpool) and toll-paying modes. Using a website or mobile device app, drivers can change the transponder setting from paying to toll-exempt status. Drivers can set up regular periods when they are in paying or toll-exempt status, but when they wish to change their status, they must do so at least 15 minutes before entering the express lanes. Survey Research Participants. FHWA commissioned and funded the survey project through its Urban Partnership Agreement and CRD programs, which both supported the express lane proj- ect. The Georgia DOT, State Road and Tollway Authority (SRTA), and Georgia Regional Transpor- tation Authority (GRTA) were the Georgia partners leading the express lanes project. Researchers from the Volpe National Transportation Systems Center designed and led the research, and a transportation survey research consulting firm conducted the field work to implement the survey. Research Objectives. FHWA commissioned the survey research to document how the express lanes changed corridor travelers’ travel behavior and attitudes toward tolling. Specific travel behaviors of interest included mode choice, route choice, and trip timing. Attitudes tested also included a question about whether tolling was fair. Equity was one of the study objectives in the implementation of the survey. As explained in the Volpe report that presented the findings from the Atlanta surveys, “Another important goal of the household survey is to understand the implications of congestion pricing for socioeconomic and geographic equity, by analyzing the impacts on household budgets, time allocation, and trip making behavior across groups of households” (Petrella et al., 2014). The surveys were intended for a national evaluation of congestion pricing, not just for Atlanta, and included equity as one of several performance measures (Zimmerman et al., 2011). Sampling and Recruitment Survey Design. The researchers designed a two-stage house- hold travel diary survey, with surveys administered in two waves, one before the express lanes

conducting pre- and post-Implementation surveys of Traveler Behavior and Opinions, atlanta Region, I-85 corridor 325 opened and the second after the lanes were in operation. In this longitudinal survey design, known as a “panel survey,” researchers contacted the same set of households for both waves of surveying. A total of 1,655 households participated in both survey waves. The total number of respondents was 3,126 because each household adult completed the survey (Petrella et al., 2014). Further details about the sampling approach, timing of implementation, and use of incentives to increase participation are discussed in the accompanying text box (see text box, Survey Design Considerations: I-85 Corridor, Atlanta). Survey Design Considerations: I-85 Corridor, Atlanta The survey design for the Atlanta I-85 project required careful attention to the sampling approach, the timing of the survey waves, and the use of incentives. More details can be found in Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the I-85 Corridor in Atlanta (Petrella et al., 2014), which is excerpted and paraphrased below. The Sampling Approach. The survey was designed to capture a representative sample of people traveling in the I-85 corridor during commute hours. The study targeted its sampling plan to people traveling by three modes: driving, taking the transit, and vanpooling: • Drivers on the corridor were identified via license-plate capture photography on sections of I-85 and Buford Highway, which runs parallel to I-85. Buford Highway was selected because of its proximity to I-85, and the fact that it runs parallel to I-85 for the entire length of the corridor, thus offering a good alternative to I-85. Limited resources did not allow the sampling of other parallel arterials. License plate collection was focused on peak and shoulder periods (6:00 to 10:00 a.m. and 3:00 to 7:00 p.m.) because these periods were expected to be most affected by the tolling project and additional transit service. • Transit riders were intercepted by survey staff at the Park and Ride facilities and at transit stations in the corridor. At the park and ride lots and Metropolitan Atlanta Rapid Transit Authority (MARTA) stations, survey staff engaged with transit riders as they waited for their bus, described the survey effort and answered questions, and distributed invitation postcards. • Members of Georgia Regional Transport Authority-organized vanpools received an email solicitation to participate; those who indicated interest provided their contact information on a survey website and were mailed a survey packet inviting them to participate. Timing of the Survey Waves. Travelers received the Wave 1 surveys, the “before” data, in the spring of 2011, a few months before the express lanes opened. Wave 2 surveys went out in the spring of 2012. Researchers chose the date for two reasons: it was long enough after the express lanes opened that travelers were likely to have settled into new habits, and the two waves would occur at the same season of the year (as travel patterns often vary by season). Survey Completion Incentive. Respondents received gift cards after completing each survey wave as an incentive to complete the questionnaires. Survey researchers commonly use incentives when asking respondents to participate in surveys that require considerable time to complete. The Atlanta survey researchers explained that they felt that the incentives were particularly important to achieve an adequate number of responses from low-income travelers. Final Panel Composition. Completed one year after the initial wave, the final panel included more than 1,500 households, meeting the target for participation, and resulting in a retention rate of 58 percent across all income groups. The retention rate was defined as households where every adult member answered every single question in both Year 1 and Year 2 surveys. Year 2 partial surveys and partial households were not included in the retention rate.

326 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Trip Diary and Questionnaires. Each adult in the participating households was asked to fill out a two-day travel diary, a demographic questionnaire, and a supplemental questionnaire that asked additional travel behavior questions and attitudinal questions. One section of the survey asked respondents basic socio-demographic questions. The details most relevant to the equity analysis were race, Hispanic ethnicity, and income. Income data was collected by asking respondents to choose one of 10 different income ranges. Other socio- demographic questions, such as educational attainment, were also asked. In the travel diary, respondents reported on every trip they made during an assigned 48-hour period. For each trip, they recorded the type of details common to household travel diary studies, including the origin and destination, mode, number of people in the vehicle, and trip purpose. The supplemental attitudinal survey asked respondents questions about their general travel patterns, details of their regular commuting behavior, and their attitudes related to their travel and to the express lanes. The two attitudinal questions most directly related to equity analysis asked respondents whether or not they agreed with the following statements (Petrella et al., 2014): • I will use a toll route if the tolls are reasonable and I will save time. • Highway tolls are unfair to people with limited incomes. Respondents had the option to complete the surveys online (which most did) or by phone. Also, respondents had the option to read the survey project materials in English or Spanish. Survey Findings. The most detailed survey findings about equity were reported in two separate documents: a detailed project report for the I-85 Corridor (Petrella et al., 2014) and a separate report that compared equity findings from both the Atlanta surveys and a parallel panel survey conducted with people in Seattle who traveled in a corridor with a newly tolled bridge (Ray et al., 2014). The two reports treated the equity issues differently in several respects. First, Petrella et al. reported some findings by race, but Ray et al. did not. Second, the two reports grouped house- holds by income using different metrics. Petrella et al. defined as “low-income” any household where respondents said they earned less than $50,000 a year. By contrast, Ray et al. developed four income groups, with the poorest group, Group 1, representing households that earned up to 3 times the national poverty level. This metric considered the number of adult and child resi- dents in the household when determining income status. Group 1 had a mean annual income of about $37,000, but the range fluctuated from $0 to $99,999. This definition of low-income differs from the FHWA Order 6640.23A, which focuses on households below the poverty line. The authors explained that they had hoped to look at just those households below the poverty line, but there were too few participating in the survey to do so. The discussion that follows presents some key findings by income and race, looking at tran- sponder ownership, travel behavior, and opinions related to equity. Neither report presented any findings by Hispanic ethnicity or those who completed the surveys in Spanish. Demographics. Among all individual survey respondents, 25 percent reported a race other than white and 6 percent reported being of Hispanic ethnicity. With respect to income, 13 per- cent of households reported earning less than $50,000 (Petrella et al., 2014). Transponder Ownership. Low-income households were less likely to have a Peach Pass than wealthier households. Among households earning less than $50,000, 20 percent had a Peach Pass, compared to 48 percent of the highest-income households (Petrella et al., 2014). The reasons that people gave when asked why they did not have a Peach Pass, if that was the case, did not

conducting pre- and post-Implementation surveys of Traveler Behavior and Opinions, atlanta Region, I-85 corridor 327 vary by income. The most frequent responses among people in all income groups were “Tolls are too expensive,” “Don’t use toll roads enough,” and “Against tolling, in general” (Ray et al., 2014). Neither report presented findings on how Peach Pass ownership varied by race. Travel Behavior. Low-income and minority respondents were modestly less likely than Whites to use the express lanes, and these groups also used the express lanes somewhat less than they did the HOV lanes. More specifically: • Income: In Wave 1, people from the lowest-income households were less likely to use the (free) HOV lanes. Similarly, in Wave 2, people from the lowest-income households were less likely to be “regular” express lane users, defined as people who used the lanes at least once per week. Lower-income respondents made up 9 percent of regular express lane users but 14 per- cent of the full sample. Comparing Wave 1 to Wave 2, the lowest-income group’s use of the express lanes compared to the HOV lanes fell by 2 percentage points (Petrella et al., 2014). • Race: Among both Asian and Black respondents, regular express lane usage was 7 percentage points lower than the regular HOV lane usage (Petrella et al., 2014). The report by Ray et al. looked at travel behavior slightly differently but came to similar con- clusions when they examined the travel diary data: In Wave 1, adjusted income was related to corridor route and mode choices for the panel sample . . . [L]ower income groups were somewhat less likely to use HOV lanes (particularly in comparison to the wealthi- est group) or to take transit on the corridor. In Wave 2 . . . use of the Express Lanes (vs. the HOV Lanes) rose across the board, with use being only somewhat greater among those with higher incomes compared to those with lower incomes. The share of trips on I-85 increased for all income groups, with the exception of the highest income group, who had a greater share of trips on other roads in the corridor (Ray et al., 2014). Ray et al. concluded that: “In Atlanta the income equity impacts were relatively small, which makes sense considering the availability of a free alternative adjacent to the new toll lane. While the lowest-income group used the priced facility less than higher-income groups, the differ- ences were not large, and we did not see a greater cancellation in trips among the lower income” (Ray et al., 2014). Opinions about Toll Roads. The two attitudinal survey questions most directly related to equity asked respondents if they would use the tolled express lanes and if they thought that high- way tolls are unfair to people with limited income. Petrella et al. summarize their findings about the Wave 2 opinions as follows, finding significant differences in opinions by income and race: • Income. “By income, those in the lowest income bracket were significantly more likely to disagree than agree that they would use a toll route (45% disagreed vs. 35% agreed); whereas the reverse was true among the highest income bracket, with agreement outweighing dis- agreement (52% vs. 34%). . . . With regard to equity, the highest income group stands out as somewhat anomalous, as less than one-half agreed that tolls are unfair to those with limited income and one-third disagreed. . . . For all other income groups, a majority agreed that tolls are unfair. . . . The lowest income group was the most concerned about equity.” Among this group, 62% agreed that the tolls were unfair (Petrella et al., 2014). • Race. “On each of the measures there are significant differences by race. Blacks were signifi- cantly less likely to agree that they would use a toll route (33% vs. 42% for Whites and 41% for Asians). . . . In addition Blacks and Asians were more likely than Whites to be concerned about equity (the difference between percent agree and disagree was significantly larger for Blacks (49 percentage points) and Asians (42 percentage points) than for Whites (22 percent- age points)” (Petrella et al., 2014). Ray et al. looked at how opinions on these topics changed between the two waves. Looking at the lowest-income group, they found that slightly fewer thought tolls are unfair in Wave 2,

328 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox but that far fewer people thought they were likely to use the toll lanes in Wave 2. More specifi- cally, the authors explained: [T]he proportion agreeing that “tolls are unfair to those with limited income” was fairly similar across all the income groups in Wave 1, with 67% to 75% agreeing with the statement. In Wave 2, agreement dropped across the board, though the lowest income group had the smallest shift in opinion. In Wave 1, 67% of the lowest income group agreed that tolls are unfair, and in Wave 2 61% did so. . . . Among the wealthiest group, however, agreement dropped from 71% to 44%, a decline of 27 percentage points. . . . A majority of all income groups agreed that they would use a toll route to save time in Wave 1, with little difference in opinion by income group. However, all groups became significantly more negative towards tolling in Wave 2. Among Group 1, the percent agreeing they would use a toll route decreased from 61% in Wave 1 to 38% in Wave 2. Groups 2 and 3 saw a similar shift in opinion. While the wealthiest group was also less likely to say they would use a toll route in Wave 2 (52% vs. 64% in Wave 1), the shift was signifi- cantly smaller than found for all other income groups (Ray et al., 2014). Summary of Equity Findings. The two reports suggest a number of important findings related to equity: • Low-income and racial minority respondents were a small portion of the overall sample. However, a comparison of the proportion of low versus high-income respondents and White versus minority respondents showed that the groups of special concern were only modestly less likely to use the express lanes on a regular basis. • Of greater concern from an equity perspective, low-income respondents were considerably less likely to own a transponder. (Neither report compared transponder ownership by race.) • In Wave 2 of the survey, considerably fewer low-income respondents predicted that they would use the express lanes than did high-income respondents. Lower-income respondents were also more likely to say that tolls are unfair to those on limited incomes. Similar patterns held when comparing Asians and Blacks to Whites. • Among low-income respondents, slightly fewer said that tolls were unfair to low-income trav- elers in Wave 2 than in Wave 1. Resources and Costs Conducting two waves of a large-sample-size survey with a complex questionnaire is invari- ably expensive and time-consuming. Project sponsors who are committed to the survey activity using this methodology will need to budget for the work and begin the planning processing at least a year or two before the tolling project opens. Cost. The market research firm selected for the Atlanta survey was also deployed for the Seattle survey. The combined cost for its services, including a transit-intercept survey in both regions, was reportedly on the order of $900,000. It is also challenging to maintain a panel across multiple waves as described here. While only a small item in the total cost, the research team employed the use of gift cards valued at $15 and $30 and sent out quarterly reminders of their availability to promote participation by the panelists in the ongoing research effort. Time required. The original plan for the Atlanta survey anticipated over two years for the pro- cess from drafting the survey methods plan to production of the final dataset (Zimmerman et al., 2011). Agencies should anticipate time for developing a request for proposals and hiring a survey research consultant in addition to time at the end of the process for data analysis and report writing. Benefits: Why It Was Effective The Atlanta survey project provided valuable data to assess both the practical impacts of the tolled lanes on low-income and minority travelers, as well their opinions. The survey was able to provide a very rich dataset because it included pre- and post-implementation waves of sur- veying, a travel diary and additional survey questions, and a large enough sample size to be able

conducting pre- and post-Implementation surveys of Traveler Behavior and Opinions, atlanta Region, I-85 corridor 329 to draw conclusions with some confidence about particular population groups of interest for equity analysis. More specifically, the researchers were able to present findings on many issues relevant to equity analysis, including the following factors: • Transponder ownership (both waves) • Use of the tolled express lanes (Wave 2) • Carpooling in the HOV lanes (Wave 1) versus express lanes (Wave 2) • Attitudes toward the tolled lanes before and after they opened, as well as the change in atti- tudes from Wave 1 to Wave 2 In each case, the researchers could compare the results for lower-income and higher-income travelers and for minority and White travelers. Challenges and Limitations While the Atlanta survey project provided invaluable information for understanding the equity implications of the tolling project, there were some aspects of the data and reporting that could be done differently in future projects to provide even more useful results: 1. The reports documenting the results did not always report the findings broken down by income groups and racial groups, and none of the documents did so by Hispanic ethnicity or limited English proficiency status. It would be useful to consistently report findings for all these groups in all reports, if there will not be a separate “equity” report that does so. By con- trast, if a primary project report is not designed to cover equity issues, then a separate report doing so should report on all populations of special concern. 2. One reason that the survey reports did not always present the findings for all populations of special concern may be that there were too few responses from people in these groups to make it possible to report out results for just these social groups. (People from these groups may be less likely to respond to surveys, and similar to the Atlanta project, they may simply be less likely to travel in the tolled corridor in question, especially during peak hours.) One solution to this problem is to design a sampling plan that deliberately over-samples people in these groups. 3. The license plate survey method is appropriate to capture people traveling on a specific cor- ridor (as opposed to people who live or work in a specific location). However, the researchers appropriately point out that: “One important limitation of this license plate capture approach is that survey materials can be sent only to the addresses of the registered owners of vehicles that are photographed. This has the effect of excluding from the sample those travelers in the corridor who are passengers in a vehicle owned by someone else (or in some cases, as the driver of someone else’s vehicle). Likewise, official vanpool vehicles cannot be matched to the lead driver’s address, much less to the vanpool passengers” (Petrella et al., 2014). 4. The participants recruited through the license plate survey all traveled during the peak hour, when travelers would face the highest tolls, which is a good rationale for focusing on this group. However, the sampling strategy also creates a less diverse pool of respondents. As Ray et al. explain, “Due to the sampling method, the sample has a greater share of employed com- muters, with correspondingly higher incomes and greater representation from the middle age brackets.” The timing of the two waves of the survey was ideal for determining how the open- ing of the express lanes changed travel behavior, the goal of the Atlanta project. However, the Wave 1 survey results came too late to provide data to inform the project’s preliminary design phase. Agencies designing their projects in the early stages would find it helpful to know how many low-income, minority, or limited English proficiency people travel in the corridor and by what mode (especially during peak hours, if the tolling might be variable), as well as their opinions about the proposed project. Therefore, project sponsors may wish to add an addi- tional survey wave (or a different survey) earlier in the project development process. 5. One challenge that both reports faced was how to define “low-income” in a meaningful way. Ray et al. (2014) normalized household incomes to family size and compared them

330 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox to the federal poverty level, which is informative as well as sensitive to the region’s higher cost of living. However, U.S. DOT Order 5610.2(a) and FHWA Order 6640.23A establish narrower definitions of low-income—a person whose median household income is at or below the Department of Health and Human Services poverty guidelines. The poverty guidelines do not vary by region of the country in the lower 48 states—a potential limita- tion to their usefulness. Future researchers will likely continue to investigate appropriate definitions of low-income that account for differences in the cost of living locally but agen- cies and practitioners should also consider oversampling methods to satisfactorily address the criteria embedded in the U.S. DOT and FHWA orders. Lessons Learned The Atlanta survey project suggests a number of lessons for future project sponsors in survey design implementation as summarized below and in its practical implementation (see text box, Longitudinal Household Travel Diary Surveys: Some Lessons and Recommendations): 1. Large-sample-size panel surveys are costly but they provide an invaluable set of data to docu- ment actual usage and opinions among populations of special concern. Longitudinal Household Travel Diary Surveys: Some Lessons and Recommendations The longitudinal household travel diary survey approach employed for the Atlanta and Seattle regions provides sufficient demographic detail to support equity assessments but is challenging in many ways. The Volpe-led research team prepared a poster (Peirce and Petrella, 2014) that shared some practical advice for panel retention throughout the life of the project, excerpted and paraphrased below. Share basic results with your panel members • You are giving back to them (not asking them to take another survey) • They see they are part of a panel, know what other panel members think, and learn more about why the project is important • But, be mindful of what you share so you do not impact/bias future surveys Panel maintenance via emails: restrain yourself! • Over multi-year projects—keep electronic contacts (e.g., emails) to less than monthly • Ask about communication preferences (do they want text message reminders?) • Carefully design and test emails for all environments (Gmail, Outlook) • Mix and clearly differentiate emails that share information, invite optional participation, and request participation • Avoid contact during school vacation and holiday periods Continually convey support and appreciation for their panel membership • Every contact with panel members invites free-form feedback for doing better • Every contact with panel members reiterates why project is important • Re-issue gift certificate electronically to those who have not cashed it yet—as a means of staying in touch and reminding panel members to continue participating • Give long-lead times—such as a month for providing information about changes to household information (new vehicles, new child) • Provide reminders for assigned travel dates and stagger emails by time of day, day of week, and looking at employment information

conducting pre- and post-Implementation surveys of Traveler Behavior and Opinions, atlanta Region, I-85 corridor 331 2. When planning sample size, it is important to account for the fact that people’s travel patterns will change over time (e.g., they get a new job, move within the region, start working from home some days). This makes the pre-post comparison challenging. When planning sample size, it is important to account for people dropping out both because they lose interest and also because they may move out of the area. 3. Reports documenting survey findings should be required to clearly outline the findings for all populations of special concern compared to other population subgroups. 4. Surveys should ask about both travel behavior and opinions to present a rich picture of both actual and perceived impacts from tolling. 5. Surveys should make sure to explore transponder ownership rates and perceptions about the devices and accounts because the need for a transponder can prove a significant barrier to toll facility access. Resources Peirce, S., Puckett, S., Petrella, M., Minnice, P., and Lappin, J. 2014. Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the SF-520 Corridor in Seattle. Volpe National Transporta- tion Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54063/UPA_Panel_Survey_Seattle_ Final_Report_Volpe.pdf. Peirce, S., Petrella, M., Puckett, S., Minnice, P., and Lappin, J. 2014. Urban Partnership Agreement and Conges- tion Reduction Demonstration Programs: Lessons Learned on Congestion Pricing from the Seattle and Atlanta Household Travel Behavior Surveys. Volpe National Transportation Systems Center. Retrieved from http:// ntl.bts.gov/lib/54000/54000/54065/UPA-CRD_Panel_Survey_Lessons_Learned_Final_Report_Volpe.pdf. Peirce, S., and Petrella, M. 2014. 2010–2012 Longitudinal Household Travel Diary Study: Seattle & Atlanta. Poster. Retrieved from http://static.tti.tamu.edu/conferences/tss12/posters/14.pdf. Petrella, M., Puckett, S., Peirce, S., Minnice, P., and Lappin, J. 2014. Effects of Full-Facility Variable Tolling on Traveler Behavior: Evidence from a Panel Study of the I-85 Corridor in Atlanta. Volpe National Transportation Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54062/CRD_Panel_Survey_Atlanta_ Final_Report_Volpe.pdf. Ray, R., Petrella, M., Peirce, S., Minnice, P., Puckett, S., and Lappin, J. 2014. Exploring the Equity Impacts of Two Road Pricing Implementations Using a Traveler Behavior Panel Survey: Full Facility Pricing on SR 520 in Seattle and the I-85 HOT-2 to HOT-3 Conversion in Atlanta. Volpe National Transportation Systems Center. Retrieved from http://ntl.bts.gov/lib/54000/54000/54064/UPA-CRD_Panel_Survey_Equity_Final_Report_ Volpe.pdf. Zimmerman, C., Gopalakrishna, D., Pessaro, B., Goodin, G., and Saunoi-Sangren, E. 2011. Atlanta Congestion Reduction Demonstration: National Evaluation: Surveys and Interviews Test Plan. Retrieved from http://ntl. bts.gov/lib/51000/51600/51687/11-104.pdf.

s e c t i o n 3 Reference Tables

C o n t e n t s 337 Reference Table 2.1 Applicable Requirements Governing Tolling Projects 339 Reference Tables 3.1, 3.2, and 3.3 Decision-Makers and Stakeholders: Actions, Decisions, and Concerns 350 Reference Table 4.1 Examples for Resource Topic Considerations Added by Tolling 353 Reference Table 5.1 Environmental Justice Assessment Methods by Resource Topic Area 358 Reference Table 6.1 Qualitative versus Quantitative Evaluations of Resource Topics

337 Reference Table 2.1 (see Step 2 in the Guidebook) provides a list of relevant EJ or non-dis- crimination related regulations or policies that should be considered in tolling projects. The table provides a “criteria” column with the key information that should be considered dur- ing the evaluation process to determine if the regulation or policy applies to the project under consideration. R e f e R e n C e t a b l e 2 . 1 Applicable Requirements Governing Tolling Projects Stage of Decision-Making Regulation/Policy Criteria Policy and Planning Title VI of the Civil Rights Act of 1964/EO 12898/U.S. DOT Order 5610.2(a) EJ populations provided opportunities for involvement and input in the creation of the LRTP and STIP/TIP alternatives 23 CFR 450.216 (fiscal restraint requirements prior to determination of CE, FONSI, or ROD) LRTP identifies tolling as a goal, objective, or strategy LRTP financial plan and STIP/TIP reflect anticipation of funding from tolled projects Specific tolling projects are included in the STIP or TIP 23 CFR 771.111 and 23 U.S.C. 139(b)(3) (early coordination, public involvement, and project development) LRTP analyzes impacts of tolling alternatives FAST Act and MAP 21 - Toll Pilot Programs - Require MOU with FHWA and annual audits Project utilizes state bonds, private activity bonds (PABs), or other U.S. DOT-approved financing instruments. FAST Act Section 1411 (tolling, HOV facilities, Interstate reconstruction and rehabilitation) Allows states to consider the use of tolls for the reconstruction and rehabilitation of the Interstate System in their respective states. Clean Air Act* (NAAQS) Consistent with regional air quality goals as determined in the LRTP Reference Table 2.1. Applicable requirements governing tolling projects. (continued on next page)

338 assessing the environmental Justice effects of toll Implementation or Rate Changes: Guidebook and toolbox Stage of Decision-Making Regulation/Policy Criteria 23 CFR 771 (FHWA NEPA Regulations) EJ populations provided opportunities for involvement and input in the development of the project (§771.105, §771.109) EO 12898 Directs federal agencies to identify and address the disproportionately high and adverse human health or environmental effects of their actions on minority and low-income populations EO 13166 Required federal agencies to develop and implement a system to provide those services so limited English proficiency persons can have meaningful access to services Clean Air Act* (NAAQS) Requires "hot spot" analysis due to non- attainment and a potential significant impact on the human environment Requires quantities of MSATs evaluation 23 U.S.C. 327: Surface Transportation Project Delivery Program State DOT is delegated review authority of EJ analysis U.S. DOT maintains review authority of EJ analysis Project Design and Implementation NEPA, 40 CFR 1500–1508 (NEPA Regulations) NEPA review Qualifies for CE Requires EA Results in FONSI Requires EIS Requires ROD Reference Table 2.1. (Continued).

339 Decision-Makers and Stakeholders: Actions, Decisions, and Concerns Reference Tables 3.1, 3.2, and 3.3 (see Step 3 in the Guidebook) identify the various types of agencies and stakeholders likely to be encountered at various stages of decision-making. For each of these entities, the primary roles and responsibilities, key actions, and concerns as they relate to tolling decisions and the assessment of EJ impacts are highlighted and differentiated by the various decision-making stages. Collectively, the tables reinforce the important point that EJ is an ongoing responsibility of sponsoring agencies in a tolling context at all stages of decision-making. Recalling the fundamental principles of EJ, it is entirely appropriate for practitioners to develop strategies for reaching and engaging with a variety of stakeholders, including EJ populations, at various stages of a toll implementation to foster informed and productive working relationships with stakeholders and communities. R e f e R e n c e T a b l e s 3 . 1 , 3 . 2 , a n d 3 . 3 Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens U.S. DOT and FHWA Sets policy on general tolling (Section 129) and HOV/HOT lanes (Section 166), including linkages between funding and congestion reduction. Establishes toll pilot programs, e.g., Express Lane Demonstration Program, Interstate System Construction Toll Pilot Program, Interstate System Reconstruction and Rehabilitation, Value Pricing Pilot Program. The U.S. DOT Secretary is responsible for decisions on the allocation of private activity bonds, often used as a component for long-term financing of toll facilities operated under P3s. U.S. DOT is also responsible for evaluating and issuing loans and credit enhancement under the Transportation Infrastructure Finance and Innovation Act (TIFIA) program, which is often a component in the long-term financing of toll roads operated by state and local agencies or procured through P3s. Produces materials that review best practices surrounding EJ analysis, public outreach, and EJ mitigation for toll road projects. Reviews MPO and State DOT plans and procedures to determine if Title VI, EJ, and public participation are adequately addressed. Reference Table 3.1. Decision-makers and stakeholders: actions, decisions, and concerns for Stage 1 (policy and planning). (continued on next page)

340 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox State Legislature/State Government State legislatures establish the authority of state and local agencies to engage in toll road development and have the power to create tolling authorities and establish the legal powers, required oversight, and mechanisms for funding and finance. Legislatures in many jurisdictions establish laws and standards for the application and collection of tolls and the procurement, delivery, and operation of toll road facilities. Legislatures may hold hearings to examine policy and rate changes or examine feasibili- ty of remedies to assist low-income constituents. Legislative actions could include authorization of toll exemptions for public transportation agencies, discounts or exemptions for eligible low-income users, or low-income tax credits. State DOTs Identifies need and funding/finance resources in coordination with local agencies through the statewide long-range transportation programming of toll facilities through the State Transportation Improvement Program (STIP). Responsible, in some cases, for entering into memoranda of understanding with FHWA division offices to establish eligibility for tolling on a federal-aid highway and to facilitate annual toll revenue and compliance audits under Section 129. Addresses the benefits and adverse effects of transportation investments among different population groups and develops appropriate procedures, goals, and performance measures. Ensures that the STIP, findings of statewide planning compliance, and NEPA activities satisfy the intent of Title VI requirements and EJ directives. Implements public involvement activities to ensure the meaningful participation of low- income and minority populations in the planning process. Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Public Tolling Authorities Responsible, in some cases, for entering into memoranda of understanding with FHWA division offices to establish eligibility for tolling on a federal-aid highway and to facilitate annual toll revenue and compliance audits under Section 129. In some jurisdic- tions, public tolling authorities take on all responsibility for planning, design, project delivery, and operations/maintenance for toll facilities. Toll road development and operations may also be procured by public authorities through P3s in some jurisdictions, with options ranging from Design Build through Design Build Operate Finance Maintain through long-term lease transfers with private concessionaires. Public tolling authorities can act as applicants and borrowers under the TIFIA program, which is often used as a compo- nent in the long-term financing of toll road facilities operated by state agencies. Public tolling authorities also often have the authority to issue municipal bonds and other financial instruments secured by a pledge of future toll revenues. Public tolling authorities are typically not responsible for overseeing transportation planning and policy development, instead the responsibility falls to the relevant MPO or state DOT. Reference Table 3.1. (Continued).

decision-Makers and stakeholders: actions, decisions, and concerns 341 Public Transportation Agencies No role in most cases at this stage. Public transportation agencies are often called on to mitigate impacts on EJ populations for tolling projects; therefore, it may be helpful for transit agencies to understand the breadth of EJ populations affected in the planned tolling corridor to ensure future transit planning studies are aware of possible impacts and that no changes are needed to their policy or planning documents that guide future investments. Public-Private Partnerships and Private Sponsors Facilities that are delivered through P3s or long-term lease/concession agreements are established through a planning process by federal, state, and local agencies as described above. Private sponsors in these arrangements may have responsibility for the delivery and operation of the project, including the annual setting and collection of tolls, consistent with parameters specified in a comprehensive development agreement or lease. No role in most cases at this stage; however, there may be concerns with risks associated with possible project halts or eliminations because of community or public opinion concerns. Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Local Government Legislatures also establish mechanisms for the funding and financing of government and public authority operated toll road facilities, including the creation of State Infrastructure Banks, and bond issuing entities. Legislatures in many jurisdictions have enacted statutes enabling P3s and establishing standards for comprehensive development agreements, long-term lease agreements, toll collection agreements, and related mechanisms. Legislatures are responsible for establishing authorities for the issuance of private activity bonds on behalf of private toll road sponsors. Local governments work with metropolitan and state agencies to identify need and funding/finance resources for the long-range transportation planning process and TIP/STIP process. In some jurisdictions, local agencies can take on all the responsibility for the study, plan, design, project delivery, and operations and maintenance of toll facilities, including project delivery and operations through P3s. Local agencies can act as applicants and borrowers under the TIFIA program, issue bonds for the financing of toll facilities, and in some jurisdictions, act as conduits for the issuance of private activity bonds. Legislatures could authorize the use of electronic benefit transfer cards for replenishment of toll accounts. Public opinion of local governments can be affected during the policy and planning stage if community members are not amenable to the future tolling changes and by their perceived impact on EJ populations, which could, in turn, affect future local elections. Reference Table 3.1. (Continued). (continued on next page)

342 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Reference Table 3.1. (Continued). Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Private Investors and Bond Holders Private bank lenders and large institutional bond investors may conduct due diligence prior to investment but have little role in the planning for a facility or in toll rate setting. Private equity investors may work with P3s or private sponsors to contribute to concept planning, or toll-setting within the framework of a comprehensive development agreement or long-term lease. No role in most cases at this stage. Private bank lenders, bond investors, and private equity investors may consider public opinion, community concerns, and EJ impacts and environmental performance goals identified during project planning in due diligence to evaluate political risk, project delivery risk, and uncertainty in project cost, schedule, and operations. Permit Agencies No role in most cases at this stage. No role in most cases at this stage. Research Institutions May provide research or advisory services to State DOTs, MPOs, local governments, and toll agencies, among others, in the development of demand-related or feasibility assessments of transportation funding and toll road alternatives, including traffic and revenues, travel survey, impact assessment, market research, focus groups, survey implementation, public involvement, and other research or technical assistance activities. May convene public events to bring wider attention to transportation funding challenges or provide subject matter expertise. May provide research or advisory services to State DOTs, MPOs, local governments, and toll agencies, among others, in the development of EJ-related policies and procedures. May conduct research into socioeconomic impacts, perform market research, focus groups, and travel surveys and facilitate public involvement, among other research or technical assistance activities. Social Service, Community, and Advocacy-Based Organizations No role in most cases at this stage. Social service, community, and advocacy- based organizations may provide facilities, expertise, and other forms of assistance to enable affected populations to increase their engagement in the policy and planning stage. They may hire experts to review and comment on technical documents on behalf of the community, represent or voice concerns/need of client communities, and provide perspectives on possible impacts on community cohesion, community services, and client populations and ways to minimize or mitigate effects. Communities of Concern/Affected Populations Stakeholders have the opportunity to participate in the planning and toll development process through the long- range planning process, the TIP/STIP process, and processes for comment mandated for public toll agencies, public comment processes mandated for public finance agencies, and the legislative process for enabling toll road development, P3s, and funding and finance mechanisms. A community's influence on planning is typically limited to public meetings and surveys, although focus groups, workshops and other formal and informal events can be held to provide information and invite dialogue. Concerns may be wide-ranging and could include the economic cost of paying a regular toll, job access, access to services, and depending on the location of the project, the environmental and social impacts of the physical project.

decision-Makers and stakeholders: actions, decisions, and concerns 343 Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens U.S. DOT and FHWA Oversees State DOT role described below to ensure compliance with EJ guidance and orders (unless state is acting as FHWA for environmental review purposes through NEPA assignment). For states without NEPA assignment, FHWA provides project-level oversight and review/approval authority over the EJ analysis process during project design. For certain CE-projects, more authority may be delegated to the state DOTs through programmatic agreements. For states with full NEPA assignment (23 U.S.C. 327), FHWA oversees EJ review at a program level (ensuring the state DOT has appropriate EJ analysis procedures in place and consistently follows the procedures), but is no longer involved in project-specific decisions (with certain exceptions including tribal consultation). For states with CE assignment (but without full assignment), FHWA oversees the treatment of EJ in CEs as a program, and provides project-level oversight of EJ issues in EA/EIS documents. State DOTs State DOTs, in consultation with FHWA, determine the NEPA class of action; make NEPA process decisions on whether to include tolling in the purpose and need statement; the range of tolling/non-tolling alternatives to be considered; and determine the scope of EJ analysis related studies such as traffic, noise, and air quality. Conduct preliminary identification of EJ concerns during scoping; determine whether an EJ analysis is necessary and the scope of the analysis; evaluate alternatives to avoid disproportionately high and adverse effects; implement public outreach strategies to provide opportunity for meaningful community input; and document all of the above in CEs, EAs, and EISs, as appropriate. For states with NEPA assignment, state DOTs act as FHWA in ensuring compliance with FHWA EJ guidance and orders during project development (23 U.S.C. 326, for CEs only; 23 U.S.C. 327, full assignment). MPOs Based on recent FAST Act legislation (P.L. 114-94), MPOs must be consulted on the placement and amount of tolls if the tolling facility is located on an HOV facility on an interstate in a metropolitan area. In some cases, as in the case of the San Francisco area’s MPO, the MPOs may be responsible for conducting the NEPA review process for specified projects. Public Tolling Authorities See role description under Policy and Planning in Table 3.1. No role in most cases at this stage. State Legislature/State Government See role description under Policy and Planning in Table 3.1. Also, it is in the best interest of state governments to review the EJ analyses related to traffic, air quality, and noise to ensure the methodologies fully examine possible impacts on constituents. State governments should also carefully review P3 contracts when they are implemented. State legislators have the ability to establish additional environmental requirements in the project design and review stage. This can include requirements under a State Environmental Policy Act (SEPA) or individual laws that address tolling projects and EJ. An example of one such law is Texas's requirement that a reasonable non-tolled alternative route must be available for a tolled route to be created. Reference Table 3.2. Decision-makers and stakeholders: actions, decisions, and concerns for Stage 2 (project design including NEPA). (continued on next page)

344 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Local Government In the case of local assistance projects, local government may make certain project definition decisions that are described above for state DOTs. See role description under Policy and Planning in Table 3.1. It is also in the best interest of local governments to review the EJ analyses related to traffic, air quality, and noise to ensure the methodologies fully examine possible impacts on local constituents, as well as examine P3 contracts. Public Transportation Agencies If toll facilities are being proposed where public transportation services are not exempt from tolls, the agencies should advocate for toll-exempt alternatives and policies. Public transportation agencies must also ensure public transportation alternatives are not limited to protecting private investments from competition, specifically in the development of contracts between private and public sectors. Where transit brings value to the project objectives, public transportation agencies should seek to be members of P3 deals and require full usage of data of future tolling facilities at no cost to facilitate future planning. Inclusion of public transportation agencies and their services in tolling projects should be considered as mitigation for toll road project impacts as well as a means to garner the support of EJ and other stakeholder communities. Public Private Partnerships and Private Sponsors Project sponsors (concessionaires) may be responsible for conducting data collection and analysis, public participation, and contract administration and other activities associated with NEPA and project development subject to the terms of a comprehensive development agreement. It is the responsibility of the FHWA Division office to ensure that a P3 project meets EJ requirements under Title VI and EO 12898. The FHWA Division office reviews the level to which a project’s documentation meets the Title VI requirements and provides comments to that effect. Private Investors and Bond Holders No role in most cases at this stage. Private equity investors may work with P3s or private sponsors to contribute to project design alternatives and toll-setting plans within the framework of a comprehensive development agreement or long-term lease. No role in most cases at this stage. Private bank lenders, bond investors, and private equity investors may consider public opinion, community concerns, and EJ impacts and environmental performance commitments identified during NEPA as part of due diligence to evaluate political risk; project delivery risk; and uncertainty in project cost, schedule, and operations. Permit Agencies No role in most cases at this stage. In general, participating agencies in the NEPA process may be involved in agency outreach activities (e.g., such as workshops on the scope of environmental analysis, reviewing draft versions of documents). U.S. EPA frequently reviews and comments on EJ analysis in FHWA NEPA documents. Other permitting agencies such as U.S. Army Corps of Engineers (USACE) or U.S. Fish and Wildlife Service (USFWS) may be the federal lead agency for certain projects with no FHWA involvement and would be responsible for following EJ executive orders. Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Reference Table 3.2. (Continued)

decision-Makers and stakeholders: actions, decisions, and concerns 345 Reference Table 3.2. (Continued) Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Research Institutions May provide research or advisory services to State DOTs, MPOs, local governments, and toll agencies, among others, in the development of demand-related or feasibility assessments of transportation funding and toll road alternatives, including traffic and revenues, travel surveys, impact assessments, market research, focus groups, survey implementation, public involvement, and other research or technical assistance activities. May convene public events to bring wider attention to transportation funding challenges or provide subject matter expertise. May provide research or advisory services to State DOTs, MPOs, local governments, and toll agencies, among others, in the development of EJ related policies and procedures. May conduct research into the socioeconomic impact distribution of tolling effects on communities of concern, perform market research, focus groups, travel surveys, and facilitate public involvement, among other research or technical assistance activities. Social Service, Community, and Advocacy-Based Organizations These organizations can connect state DOTs, MPOs, local governments, and toll agencies to community stakeholders for input on surveys and to encourage participation at public meetings. Social service, community and advocacy based organizations may provide facilities, expertise, and other forms of assistance to enable affected populations to increase their engagement in the NEPA and project development stages. They may hire experts to review and comment on technical documents on behalf of the community; represent or voice concerns/needs of client communities; and explain possible impacts on community cohesion, community services, and client populations. Communities of Concern/Affected Populations At the project stage, the community's influence is typically either through surveys and/or public meetings. Generally, they have limited power to prevent the development of the overall tolling project. Participate in outreach activities and submit comments during evaluation of alternatives, scoping, and on the EJ analysis presented in the draft environmental documents. Participate in involvement processes and possibly initiate litigation. The general public can be involved in the same manner as communities of concern/affected populations, although they may not be targeted by outreach efforts if they are located outside of the project area of influence.

346 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Reference Table 3.3. Decision-makers and stakeholders: actions, decisions, and concerns for Stage 3 (implementation). Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens U.S. DOT and FHWA The original federal standard for setting tolls was that tolls be “just and reasonable.” This standard was set in 1906; however, the U.S. DOT’s oversight authority to enforce this was repealed in 1987. Although the standard still exists, under the 1987 Transportation Act, it is now under the authority of the courts to determine. U.S. DOT decides what toll funds can be used for when toll roads are enacted under federal toll pilot programs, P3s, or with the use of other federal financing tools. Oversees state DOT consideration of EJ impacts of existing tolling operations/proposed changes in tolling. State DOTs Division offices responsible for annual tolling revenue and compliance audits on tolled federal-aid facilities for review and approval by the Secretary (Section 129). In some jurisdictions, state DOTs can be responsible for the setting of toll rates, the collection of tolls, and the ongoing operations of a toll facility. Evaluates EJ implications of existing tolling and proposed toll changes. MPOs In most jurisdictions, MPOs have no role in ongoing operations or toll setting, other than the responsibility for including facilities in forecast models maintained for air quality conformity and long-range planning. However, transportation officials in the San Francisco Bay area have seen a growing need to include policies on pricing in long-range plans to create processes and plans that are consistent in the application of policy, allocation of revenues, and the design and technology applied. May support analysis of EJ implications with regional models and data.

decision-Makers and stakeholders: actions, decisions, and concerns 347 Reference Table 3.3. (Continued). Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Public Tolling Authorities Can be responsible for the setting of toll rates, the collection of tolls, and the ongoing operations of a toll facility. Toll rates and the use of toll funds are often decided by the operating boards of toll authorities. These organizations have some accountability to the state governments, and depending on whether it is an appointed or an elected position, some accountability to the public. Toll authorities also make decisions that may influence the difficulty of using the tollway. For instance, some toll authorities charge extra when a toll is paid in cash or paid by a method other than a transponder. Toll authorities could implement a marketing strategy to create an outreach campaign directed at the EJ communities to increase transponder usage. The toll system would record statistics related to transponder use rates for EJ communities and for the area as a whole. These statistics could be documented and provided to the tolling body as part of an annual progress report. If it is discovered that there is a statistically significant difference between the EJ populations’ transponder use rate and the use rate of the population as a whole, the state or entity would, to the extent it is practicable, make adjustments to the plan to increase transponder usage among the EJ populations. State Legislature/State Government In some jurisdictions, state legislatures define the standards for toll setting and operations by public agencies and private concessions. Constraints are placed on the use of tolling funds. In this way, lawmakers can influence how equitable the function of toll roads are by requiring that funds be used on items like public transportation. Local Government See role description under Policy and Planning in Table 3.1. See role description under Policy and Planning in Table 3.1. Public Transportation Agencies See role description under Project Design including NEPA in Table 3.2. Public transportation agencies may have a role in implementing mitigation for EJ impacts by providing additional transit services. (continued on next page)

348 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Public Private Partnerships and Private Sponsors Operations are typically overseen by a private concessionaire. However, the private entity functions under regulations set by state and federal governments; tolls are imposed with limitations for how toll funds may be used. This is typically limited to the cost of operations, maintenance, and financing. In some situations, additional tolls may be permitted to fund enhancements in transit operations, usually as mitigation for EJ adverse impacts. Private Investors and Bond Holders In some P3s or long-term concession arrangements, lenders and bond holders have the right to step into project construction, toll setting, and operations in certain instances of default or non- performance to help ensure repayment of debt obligations. The step-in rights and obligations are defined in and governed by the comprehensive development agreement or long-term lease. In instances of default or non-performance where lenders and bond-holders exercise step-in rights, they will act to make any changes in project construction, operations, or tolls permitted within the comprehensive development agreement or long-term lease that would maximize or accelerate the repayment of debt. These stakeholders will abide by any EJ or environmental performance commitments specified in the comprehensive development agreement or long-term lease. Permit Agencies No role in most cases at this stage. Permitting agencies may be involved with post- NEPA monitoring and reporting on project implementation (such as documentation that mitigation measures committed to in the NEPA process in fact were implemented). This monitoring (i.e., water quality, air quality, or noise monitoring during construction) is typically related to other non-EJ regulatory requirements, but may indirectly touch on EJ populations. Research Institutions May provide research or advisory services to state DOTs, MPOs, local governments and toll agencies, among others, in review of toll pricing impacts, including traffic and revenues, before- after studies on travelers, impact assessments, market research, focus groups, survey implementation, and public involvement, and other research or technical assistance activities. May convene public events to bring wider attention to transportation funding challenges or provide subject matter expertise. May provide research or advisory services to state DOTs, MPOs, local governments and toll agencies, among others, in the review of toll facility equity effects on low-income and minority populations through before-after studies, market research, focus groups, travel surveys, public involvement, and other research or technical assistance activities. Reference Table 3.3. (Continued).

decision-Makers and stakeholders: actions, decisions, and concerns 349 Decision-Makers and Stakeholders Tolling Lens Environmental Justice Lens Social Service, Community, and Advocacy-Based Organizations Community and advocacy-based organizations may be involved in Implementation Task Forces, as in the case of Minneapolis/St. Paul’s I-394 MnPASS HOT Lane project, where the task force was formed to guide the department through project implementation. This was part of a larger extensive consensus building and stakeholder involvement to build support for tolling projects (FHWA, 2011). In some states, organizations or interested groups could organize voter initiatives to constitutionally prohibit tolling implementation on certain types of facilities, as groups in Arizona attempted in 2013 (Fischer, 2013). Social service, community, and advocacy-based organizations may provide community facilities, technical expertise, and other forms of technical assistance to speak on their behalf and/or enable affected populations to voice their concerns with impacts or needs pertaining to toll pricing changes or operations. They may provide a forum to explain transponder replenishment policies or barriers to usage. Social service, community, and advocacy-based organizations may provide on-the-ground verification of mitigation implementation, or alert appropriate authorities if tolling authorities, P3s, state governments, or other entities are not fulfilling their agreements from previous stages. Communities of Concern/Affected Populations See role description under Social Service, Community, and Advocacy-Based Organizations in Table 3.3. Communities concerned about toll impacts that they deem unfair or unjust can be expressed and responded to via the public outreach process; through separate technical research conducted by advocacy or academic researchers; or via recourse through the legal system, typically either through the courts or the state legislature. Reference Table 3.3. (Continued).

350 Reference Table 4.1 (see Step 4 in the Guidebook) categorizes the resource topics that are typically considered in an environmental analysis from an environmental justice perspective. The table poses questions that the practitioner may wish to consider for each of the major topics areas—mobility, access, and safety; social and economic; physical and environmental; and cultural and historical resources. This same framework carries forward to conduct the analysis in Step 5 and in the companion Reference Table 5.1. As noted in the Guidebook and the tables, the Guidebook and the Toolbox focus on the unique impacts related specifically to tolling. Other potential environmental jus- tice impacts that relate to non-toll aspects of the project, e.g., construction impacts or cultural resources, are to be evaluated by subject matter experts applying the standard methods they would use on any roadway project. R e f e R e n c e T a b l e 4 . 1 Examples of Resource Topic Considerations Added by Tolling Reference Table 4.1. Examples of resource topic considerations added by tolling. Categories Resource Topics Questions Mobility, Access, and Safety Mobility and access— accessibility Will the toll or the toll pricing policies result in low- income drivers being “priced out” or excluded from certain trips? What reasonable alternative transportation modes are available to those who cannot afford the toll? Are there adequate non-tolled corridors to serve as alternative roads? If pricing yields travel-time savings, are they experienced by all users? Route, mode, and trip time— transportation choice Will the additional travel costs result in lower-income users choosing to use less desirable (to them) modes or routes to meet their mobility needs? Will low-income commuters change their travel times to avoid tolls or to avoid congestion on non-toll routes? Will the toll facility impact transit (e.g., altered bus routes, transit times/schedules)? Safety Will the toll facility divert a substantial amount of traffic through an EJ community? Will diverted traffic through EJ communities impose a higher safety risk to local pedestrians and bicyclists?

examples of Resource Topic considerations added by Tolling 351 (continued on next page) Categories Resource Topics Questions Social and Economic Community cohesion Will the toll facility alter social relationships through changes in design, access, or displacement? Social isolation (subset of cohesion) Will the facility pricing, design, property value effects, and changes in mobility and access isolate communities or particular populations? Displacement (subset of cohesion) Will the toll facility displace a larger number of residents and businesses compared to the non-toll roads? Household cost burden (subset of economic impacts) How will the pricing change affect the travel costs of the low-income user in relationship to their household budget? Will the non-toll alternatives be equitable in terms of travel time or distance? Community facilities and services (subset of economic impacts) How will the toll facility impact the access of EJ communities to work, schools, hospitals, and recreation? Business access (subset of economic impacts) How will the toll facility impact business access for both customers and deliveries? Noise Will the toll facility divert a substantial amount of traffic through an EJ community? If yes, will there be additional noise impacts attributable to shifts in free-flow and delay? Visual How will the toll facility visually affect those who have to look at it and those who use it? Land use and property values Are there foreseeable changes in housing or commercial rents attributable to changes in access to opportunities? Will the toll facility impact land use and property values through changes in access and reliability (i.e., commercial versus residential)? Reference Table 4.1. (Continued).

352 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Reference Table 4.1. (Continued). Note: The table has been adapted to include additional resource topics such as water, quality, and drainage and visual quality to more closely conform to FHWA topics and NCHRP Report 532: Effective Methods for Environmental Justice Assessment. Some resource topics have been moved among categories and subset clarifications have been added to more closely conform to resource topics, analysis methods, and checklists used through the rest of the Guidebook. Categories Resource Topics Questions Cultural and Historic Resources Cultural Will the toll facility benefit or discourage access to cultural and recreational resources? Historic Will the toll facility benefit or discourage access to historic sites and landmarks? (Prozzi et al., 2006). Physical and Environmental Air quality Hazardous materials Water quality and drainage Health Will the toll facility divert a substantial amount of traffic through an EJ community? If yes, will air quality be affected by shifts in free-flow and delay? Will the toll facility increase the number of vehicles carrying hazardous materials through or near particular populations, either on the toll facility or through diversions due to the toll road? Will the toll facility design and implementation improve or worsen water quality and water quantity conditions for particular populations? Will the toll facility benefit or discourage access to health facilities? Will diversion of traffic through EJ communities impose a higher risk of health issues (e.g., asthma and other pulmonary chronic diseases)? Source: Adapted from Guidebook for Identifying, Measuring and Mitigating Environmental Justice Impacts on Toll Roads

353 Reference Table 5.1 (see Step 5 in the Guidebook) summarizes assessment methods consistent with the resource topic areas introduced in Step 4 in the Guidebook and Reference Table 4.1. • Column 1 carries forward the categories of resource topics from Reference Table 4.1. • Column 2 provides a summary of standard analysis methods, referring (with slight modi- fications) to NCHRP Report 532: Effective Methods for Environmental Justice Assessment (Forkenbrock and Sheeley, 2004). This report includes detailed method descriptions for all resource areas (presented in a different order in the report). Practitioners are expected to follow state and FHWA guidance for non-toll areas, but this additional resource may provide helpful information. • Column 3 summarizes modifications to standard analyses that may be required to address specific toll-project implications in an analysis. • Column 4 summarizes the tools in the Toolbox to assist in such analyses. As noted in the Guidebook and the tables, the Guidebook and the Toolbox focus on the unique impacts related specifically to tolling. Other potential environmental justice impacts that relate to non-toll aspects of the project, e.g., construction impacts or cultural resources, are to be evaluated by subject matter experts applying the standard methods they would use on any roadway project. R e f e R e n c e T a b l e 5 . 1 Environmental Justice Assessment Methods by Resource Topic Area

354 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox Reference Table 5.1. Environmental justice assessment methods by resource topic area. Mobility, Access, and Safety Impacts Mobility, Access, and Safety Impacts (also referred to as Transportation User Effects) Transportation Accessibility 1. Unmodified transportation demand models 2. Adapted transportation demand models 3. Advanced adapted transportation models 4. Activity-based travel simulation A wide range of possible adaptations to and considerations for toll projects exists. Transportation Choice 1. Modal quality assessment 2. User demand and evaluation surveys 3. Improved transportation surveys and models Modifications to the methodologies would include greater specificity as to modal (e.g., HOV-2 to HOV-3) and transit use of the toll facility; VOT and cost factors. Transportation Safety Effects 1. Analyze national data 2. Comparison approach 3. Regression analysis 4. Bicycle safety index 5. Bicycle compatibility index 6. Pedestrian street crossings 7. Pedestrian danger index No specific differences in approach . If traffic diverts from a newly tolled facility to a parallel collector or arterial streets, there might be additional traffic on those streets. Travel demand models and surveys can assess. Social and Economic Impacts Community Cohesion 1. Focus groups to identify interaction patterns 2. Personal interviews 3. Deliberative polling 4. Travel demand models with GIS 5. Stop watch and distance wheel A toll facility may have impacts on cohesion related to accessibility and mobility. In-depth public involvement as described in the Guidebook and the Toolbox will identify this potential impact, supplemented by modeling if appropriate. Public Involvement Plan; Conducting Surveys; Travel Demand Models; VOT/Willing to Pay; Select Link Analysis; Using PUMS; Using NHTS; plus local knowledge. Public Involvement Plan; Conducting Surveys; Travel Demand Models; VOT/Willing to Pay; Select Link Analysis; Using PUMS; Using NHTS; plus local knowledge. Refer to NCHRP Report 532 and local, regional, or state subject matter experts (SMEs) for assessment guidance and assistance. Public Involvement Plan; Conducting Surveys; Focus Groups; Travel Demand Models. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. EJ Assessment Resource Topic Area Basic EJ Assessment Methods Special Considerations for Toll Projects Guidebook Tools to Assess EJ Toll Considerations

environmental Justice assessment Methods by Resource Topic area 355 Reference Table 5.1. (Continued). EJ Assessment Resource Topic Area Basic EJ Assessment Methods Special Considerations for Toll Projects Guidebook Tools to Assess EJ Toll Considerations Economic Development and Other Economic Impacts 1. Map and GIS assessment 2. Surveys or focus groups 3. Gravity models For toll projects, assess added transportation costs as a household burden. Detailed toll facility usage and revenue analyses (e.g., investment grade analyses) will likely be commissioned if private investors or owner/operators will have a stake in the project. For all projects, take care to NOT double count project benefits. User Costs, HH Burden. Travel Demand Models if applicable. Investment grade analyses usually include income segmentation, origin and destination, and value of time analyses, which are useful for discerning burdens and benefits as well as proposed pricing strategies. Noise 1. Initial evaluation screening 2. Highway project noise analysis 3. Transit project noise analysis No notable differences in approach from other highway projects. Construction and operation noise effects will need to be evaluated if there are nearby noise receptors. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. Visual Quality 1. Existing condition photos 2. Illustrative plans, sketches 3. GIS view-shed analysis 4. Photo simulation/ montage 5. Computer imaging 6. Computer animation 7. 3-D physical models 8. Videos No notable differences in approach from other highway projects. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. Land Prices and Property Values 1. Market studies and expert opinion 2. Property comparisons and appraiser opinion 3. Hedonic regression No notable differences in approach from other highway projects. Care must be taken to NOT double count beneficial or adverse impacts (e.g., noise, access). Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. (continued on next page)

356 assessing the environmental Justice effects of Toll Implementation or Rate changes: Guidebook and Toolbox EJ Assessment Resource Topic Area Basic EJ Assessment Methods Special Considerations for Toll Projects Guidebook Tools to Assess EJ Toll Considerations Physical and Environmental Impacts Air Quality 1. General air quality review 2. Detailed microscale analysis 3. Detailed regional analysis 4. Analysis using pollution surfaces No notable differences in approach from other highway projects. Some projects may produce regional air quality benefits while some may produce localized harm or benefits. See text box, Mobile Source Air Toxics on page 58. No special tools for EJ analysis, although most air quality models are closely tied to or interdependent with regional travel demand models. Water Quality and Drainage 1. Land acquisition checklist 2. Visual quality checklist 3. Accessibility checklist 4. Groundwater quality checklist 5. Surface water quality checklist No notable differences in approach from other highway projects. Identify the scope of proposed water quality and drainage improvements—are there impacts on protected populations? If so, modify. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. Hazardous Materials 1. Phase 1 desktop assessment 2. Phase 1 computer- based assessment 3. Hazardous materials transport screening study 4. Hazardous materials transport- probability modeling No notable differences in approach from other highway projects. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. Reference Table 5.1. (Continued).

environmental Justice assessment Methods by Resource Topic area 357 EJ Assessment Resource Topic Area Basic EJ Assessment Methods Special Considerations for Toll Projects Guidebook Tools to Assess EJ Toll Considerations Cultural and Historical Resources Cultural and Historical Resources 1. Multilevel impact valuation 2. Site visit and survey with a community leader 3. Stakeholder and expert charrette No notable differences in approach from other highway projects. In-depth public involvement as described in the Guidebook and the Toolbox may be needed. Public Involvement Plan, other public involvement tools. Refer to NCHRP Report 532 and SMEs for assessment guidance and assistance. Note: Analysis Methods (column 2) refers (with slight modifications) to NCHRP Report 532: Effective Methods for Environmental Justice Assessment (Forkenbrock and Sheeley, 2004) that includes detailed method descriptions for all resource areas (present- ed in a different order in the report). EJ Resource Topic Areas (Column 1) also adhere (with slight modifications) to the NCHRP Report 532 taxonomy. Reference Table 5.1. (Continued).

358 Reference Table 6.1 (see Step 6 in the Guidebook) summarizes typical analysis frameworks for the resource topic areas (as previously identified in Steps 4 and 5 in the Guidebook). The right hand column summarizes some of the toll implementation or rate change project considerations that might contribute to a finding of disproportionately high and adverse effects for EJ popula- tions or a finding of a disproportionately less beneficial effect for EJ populations compared with non-EJ populations. The analysis and findings for each resource topic area for a specific project will be particular to that project and to the interactions and consultations with EJ populations throughout the project development. R e f e R e n c e T a b l e 6 . 1 Qualitative versus Quantitative Evaluations of Resource Topics

Qualitative versus Quantitative evaluations of Resource Topics 359 Reference Table 6.1. Qualitative versus quantitative evaluations of resource topics (refer to comparable resource topic Table 5.1 for analysis methods and references to tools). Assessment Resource EJ Topic Area Types of Analysis Special Considerations for Disproportionate Impact Analysis for Toll Projects Mobility, Access, and Safety Impacts (also referred to as Transportation User Impacts) Transportation Accessibility Quantitative and qualitative Compare travel times and costs, EJ versus non-EJ, project versus non-project; project and system-wide, and over time (cumulative and indirect effects). Compare toll account policies and transponder ownership issues as borne by those without credit access or unbanked populations vis-à-vis other populations, as identified in the tools, case examples, and EJ population outreach throughout process. Transportation Choice Quantitative and qualitative Modal and transit use by EJ versus non-EJ populations of the toll facility compared to non-toll facilities. Can relate to travel time and cost in accessibility analysis. Continue to consult with EJ populations. Transportation Safety Effects Quantitative and qualitative No specific differences in approach. Be aware of potential traffic diversion to parallel collector or arterial streets. Social and Economic Impacts Community Cohesion Quantitative and qualitative Related to accessibility and mobility, but not identical. EJ input essential. Economic Development and Other Economic Impacts Quantitative and qualitative Assess added transportation costs as a household burden. Compare burdens and benefits to EJ versus non-EJ populations. Note that disproportionate provision of benefits to non-EJ populations can be as problematic as disproportionate adverse effects to EJ populations. Noise Quantitative No notable differences in approach from other highway projects. Visual Quality Qualitative No notable differences in approach from other highway projects. Land Prices and Property Values Quantitative and qualitative No notable differences in approach from other highway projects. Do not double count benefits or adverse impacts. Increased property values due to improved accessibility may be an adverse impact to low-income persons if rents go up. Physical and Environmental Impacts Air Quality Quantitative No notable differences in approach from other highway projects. Be alert to localized harm versus regional benefits (hot spots). Hazardous Materials Quantitative and qualitative No notable differences in approach from other highway projects. Water Quality and Drainage Quantitative and qualitative No notable differences in approach from other highway projects. Cultural and Historical Resources Cultural and Historical Resources Qualitative No notable differences in approach from other highway projects.

s e c t i o n 4 Checklists

C o n t e n t s 365 Checklist 1 Checklists for Framing the Project 366 Checklist 1a. Tolling Actions and the Impact-Causing Aspects of Tolling Actions 368 Checklist 1b. Potential Direct and Indirect Effects 370 Checklist 1c. Context Considerations at Different Project Stages and Phases 373 Checklist 2 Suggested Instructions for the Documentation Checklist 374 Checklist 2. EJ Impact Documentation Checklist

365 Purpose. The checklists provide a self-assessment tool to help the practitioner identify the type of action, the anticipated impact-causing aspects, potential direct and indirect effects, and the context for assessing the degree and extent to which there could be disproportionately high and adverse effects from the action on minority and low-income populations. How to Use. Add comments and questions to the checklists as appropriate describing what has been initially discovered about the project. These reminders can assist in developing the technical memorandum as described in Step 1 in the Guidebook and summarized below, for convenience. The practitioner prepares a technical memorandum that describes the approach, methods, data, and maps used to identify potential issues related to adverse effects on minority and low- income populations. The narrative describes the availability of previous studies and relevant outreach conducted to date and is supported by the checklists and reference tables completed by the practitioner to identify the type of action, the anticipated impact-causing aspects, poten- tial direct and indirect effects, and the context for assessing the degree and extent to which there could be disproportionately high and adverse effects from the action on minority and low- income populations. Checklists 1a, 1b, and 1c contain several items to assist in documentation: • Checklist 1a summarizes the eight types of tolling actions and the impact-causing aspects of the tolling actions from Steps 1.1 and 1.2. The checklist can be used as a reminder of possible interactions to consider for projects. • Checklist 1b summarizes the direct and indirect effects of projects for consideration from Step 1.3. • Checklist 1c provides a list of questions to address contextual considerations, from Step 1.4. C h e C k l i s t 1 Checklists for Framing the Project

Checklist 1a. Tolling actions and the impact-causing aspects of tolling actions. Impact Causing Aspect(s) of Toll Implementation or Rate Change Action Introduce Transaction Cost Increase Transaction Cost Create Uncertain Transaction Cost Form or Payment and/or Credit and/or Fixed Cost Requirements for User Accounts Change in Access to Highway Network Create or Increase Highway, Bridge or Tunnel Footprint Decrease Distance between Community and Highway, Bridge or Tunnel Your project?- Comments T o l l I m p l e m e n t a ti o n o r R a t e C h a n g e A c ti o n 1. New Toll Road or Bridge (mix of toll and general purpose lanes) √ Possibly (choice of toll or general lanes can be trip-specific) Possibly √ √ √ 2. New Toll Road or Bridge (all toll lanes) √ Possibly Possibly √ √ √ 3. Partial Conversion of Existing Highway, Bridge or Tunnel Travel Lanes for Tolling (mix of toll and general purpose lanes) √ Possibly (choice of toll or general lanes can be trip-specific) Possibly Possibly 4. Full Conversion of Existing Highway, Bridge or Tunnel Travel Lanes for Tolling (all toll lanes) √ Possibly Possibly 5. Partial Conversion with Widening of a Highway, Bridge or Tunnel (mix of toll and general purpose lanes) √ Possibly (choice of toll or general lanes can be trip-specific) Possibly Possibly √ Possibly (depends on whether inner or outer widening) 6. Full Conversion with Widening of a Highway, Bridge or Tunnel (all toll lanes) √ Possibly Possibly √ Possibly (depends on whether inner or outer widening)

7a. Increase Tolls on an Existing Toll Facility √ 7b. Change in method of payment on existing facility Possibly Possibly Possibly 7c. Change Toll Collection Technology on an Existing Toll Facility Possibly Possibly Possibly 7d. Introduce Variable or Dynamic Tolls on an Existing Toll Facility Possibly √ 8. Change Operator of an Existing Toll Facility Possibly (dependent on authorized ability to increase tolls) Possibly (dependent on authorized ability to increase tolls) Possibly

Checklist 1b. Potential direct and indirect effects. Potential Direct Effects Associated with an Action and Its Aspects Your project? (Likely, maybe, not likely) Your project? Comments Change in road use patterns (diversions to alternate routes or modes) for some users Change in mobility for some users Change in accessibility for some users Change in travel reliability for some users Change in trip-making behavior and trip purposes (trip frequency, trip timing) for some users Change in disposable income Change in “disposable time” for some users Array of direct effects on the natural and human environment from a transportation capital project that could occur regardless of tolling aspects Changes in health (air quality, noise, vibration) from shorter source-receptor distances, and other facility proximity effects on the natural and human environment that could occur regardless of tolling aspects

Increased travel on alternate routes or modes lead to degradation of level of service on the alternate routes or modes Changes in health (air quality, noise, vibration) for residents nearby alternate routes that have degradation in level of service Changes in quality of life from: Degraded environmental quality and pedestrian safety along alternate routes Contracted possibilities for employment and services because of degraded accessibility, mobility, and travel reliability Activities and purchases foregone because of reduced disposable income or less available time outside of travel time Array of indirect effects stemming from direct facility footprint effects that could occur regardless of tolling aspects Array of indirect effects stemming from facility proximity effects that could occur regardless of tolling aspects Potential Indirect Effects Associated with an Action and Its Aspects Your project? (Likely, maybe, not likely) Your project? (Comments)

Checklist 1c. Context considerations at different project stages and phases. 3 Stages 6 Phases Example EJ Issues for Tolling Your Project: Comments, Additional Questions P o l i c y & P l a n n i n g 1. Visioning and Policy Have needs and funding gaps identified in the long- and short-range plans evoked consideration of tolling alternatives? This phase provides broad direction to guide the transportation development process and includes the development of legislative policy at the national and state levels and a vision for the future transportation system at the local, regional, and state levels. Do projects that can partially or completely “self-fund” through tolling advance more quickly than projects that rely on traditional funding? 2. Long-Range Planning and Programming Has (or will) public involvement during plan development provide early notice of potential EJ concerns and potential alternatives? During this phase, a long-range (20 to 30 years) plan for future transportation investments and a short-range (3 to 6 years) program of projects approved for funding are developed at the regional and state levels. P r o j e c t D e s i g n 3. Environmental Studies and Preliminary Design Have assumptions about toll rates used to support the traffic analysis and, hence, the design, been shared with the EJ community for input?

Environmental studies are mandated by federal and state laws and are used to find the most appropriate design concept for a proposed facility, out of a pool of alternative concepts. The selected alternative is developed into a preliminary design. Has the affected EJ community been involved in the environmental review process, including input on alternatives and perceptions of impacts and potential avoidance or other mitigation alternatives? If a private sector partner is brought on for a design-build or design-build- operate-maintain contract and project, it is often in this stage. Do the operating and contractual agreements include flexibility and financial provisions for design changes and mitigation, if needed, including long-term agreements and backup provisions in case of operator insolvency? 4. Final Design and Right-of-Way Has the affected EJ community been kept informed about design changes and proposed resolution of obstacles and constraints? Have their suggestions on avoidance or other alternatives been considered? In this phase, refinements are made to the preliminary design based on various constraints and practical considerations. Necessary right-of- way acquisitions are identified. I m p l e m e n t a ti o n 5. Construction Are potential bidders aware of and do they understand the EJ-related design commitments articulated in the construction contract? This phase consists of the physical construction of the transportation facility. This is typically considered the last step in the traditional project development process but does not mark the end of the project’s life cycle. Do any contract change orders alter or compromise design commitments, or do they create new EJ issues? (continued on next page)

3 Stages 6 Phases Example EJ Issues for Tolling Your Project: Comments, Additional Questions 6. Operations and Maintenance Do the pricing mechanisms and changed traffic patterns of road pricing projects result in disproportionately high and adverse effects on low-income and/or minority populations? Operations and maintenance are critical elements of the life cycle of a transportation facility. Operations activities include toll pricing, discounts, and exemptions; toll account and transponder usage policies; gate timing; and pavement markings. Maintenance includes actions such as camera and pavement repair. Are there barriers to accessing or securing transponder access and replenishing accounts for populations without credit cards or unbanked populations? Has the planned toll rate change been the subject of a regional planning-level study that considers the impacts to low-income and minority populations? Have pre-implementation surveys and/or other data collection established a baseline for before and after conditions? Have monitoring protocols been established? Have metrics been established? Have thresholds and/or time periods been established at which point the operator will go back to the community for further discussion and feedback? Checklist 1c. (Continued).

373 Purpose. The checklist provides a suggested template to track the progress and status of the analysis and the disposition of each resource topic area in an EJ toll project analysis. How to Use. Early in the project, but no later than Step 4, create a one-page checklist for each pertinent resource topic area. Resource topic areas are described in Steps 4, 5, and 6, with increasing detail as the analysis evolves. Resource Tables 4.1, 5.1 and 6.1 in the Resource Tables section of the Toolbox provide example frameworks for working with the resource topics. Maintain the checklists with other project files in the documentation folder. Use the checklists as a reminder when completing each step of the Guidebook to organize data, analysis, and public outreach and feedback into appropriate resource topic categories for easier retrieval, reminders, and action. As resource topic areas are determined to have no impact, based on analysis and community input, they can temporarily be set aside or dismissed, documenting the status in Steps 5, 6, and 7 as appropriate. Steps 6 and 7 will likely have the most intense documentation efforts, but rigorous outreach and analysis efforts as well as documentation from earlier steps will assuredly make Steps 6 and 7 proceed more smoothly than they would have without the established community ties and input. The checklist is linked to Step 7 but is referred to through- out the Guidebook. For each question, check yes or no and follow the instructions to either proceed to the next question, stop, or go back to an earlier question for additional detail of resolution. With a “YES” answer, proceed to an end decision point (with some “NO” answers—indicating no impact or a likely FHWA decision), or return to an earlier step to complete data validation or a similar task. If answers do not seem to apply, use your judgement to proceed. Step numbers 4, 5, 6, and 7 refer to the steps in the Guidebook. Steps 4a, 4b, 4c, 5a, 5b, etc., are only used within the checklist for navigating the checklist; there are no corresponding substeps in the Guidebook. c h e c k l i s t 2 Suggested Instructions for the Documentation Checklist

Step # Assessment Status/Question Yes If Yes No If No Comments Major Topic Area (e.g., Accessibility, Mobility, and Safety) Resource Topic Area (e.g., Transportation Accessibility) 4a Are there EJ affected populations? To 4b To 7a, stop 4b Are there potential adverse impacts? To 4c To 7a, stop 4c Are data collection, organization, and assessment sufficient? To 4d To 4a 4d Public and EJ outreach conducted on community issues and attitudes towards project: has coordination, access to information, and participation been documented? To 5a To 4a 5a Has each resource topic area been evaluated? Has a degree of effect been determined? To 5b To 4c 5b Has public and EJ outreach to affirm the degree of effect been conducted and documented? To 6a To 4d 6a Have adverse impacts between EJ and non-EJ populations been compared? (Document) To 6b To 5a 6b Are adverse impacts predominantly borne by EJ populations? (Document) To 6c To 6c 6c Are EJ adverse impacts more severe/greater in magnitude than impacts on the non-EJ population? (Document) To 6d To 7a, stop 6d Have alternatives been considered to avoid, minimize, and/or mitigate issues? (Document) To 6e To 6a or 7d 6e Have offsetting benefits been considered? (Document) To 6f To 6e 6f If yes to 6d and 6e: Are impacts still disproportionately higher and more adverse for EJ populations? To 6g To 7a, stop Checklist 2. EJ impact documentation checklist.

6g Have other mitigation measures been considered that would avoid or minimize effects? To 6h To 7d 6h Have affected populations been involved in the decision-making process? (Document) To 6i To 7d 6i If yes to 6f, 6g, 6h, and EJ population is minority: Does project fill substantial public need and meet overall public interest? To 6j To 7d 6j If yes to 6i: Do alternatives with less adverse impacts also have adverse impacts, or are too costly? (Document) To 7b To 7d 7a If there are no adverse impacts, impacts minor with no controversy, or impacts would be the same for EJ and non-EJ populations: Document in letter to file with required wording. Stop 7b Document findings, review process, EJ involvement, and practicability analysis with FHWA region. FHWA may approve. To 7c 7c Establish environmental commitments for the project. Implement and monitor. Stop 7d FHWA will likely not approve project. Stop Additional comments on resource topic area: For choices not covered in steps above, refer to the FHWA flowchart and in the Federal Highway Administration Environmental Justice Reference Guide (2015) or use judgement to proceed.

s e c t i o n 5 Scenarios

C o n t e n t s 381 Introduction 382 Scenario A: Untolled Bridge to Tolled Bridge, P3 395 Scenario B: HOV Lanes to HOT Lanes 406 Scenario C: Rate Change

381 This section illustrates how to apply the tools and case examples included in the Toolbox through the suggested eight-step process framework for assessing the envi- ronmental justice (EJ) implications of toll implementation or rate changes. The pro- cess framework and application of tools is scalable depending on what analysis and public engagement indicate as the potential for disproportionately high and adverse effects of the toll implementation or rate change on EJ populations. Three basic scenarios illustrate important considerations and questions relevant to each step in the process framework. In these scenarios, the hypothetical practi- tioner considers the distinctive impact-causing aspect(s) of toll implementation or rate change actions. The practitioner uses the various tools and case examples in the Toolbox for technical assistance in identifying and addressing potential EJ impacts within a tolling context. The three scenarios are: • Scenario A is an untolled bridge transitioning to a tolled bridge. The project sponsor is considering using a Public-Private Partnership (P3) to design, build, manage, and operate the bridge. • Scenario B is a high-occupancy vehicle (HOV) facility converting to a high- occupancy tolling (HOT) facility, operated by the state Department of Transpor- tation (DOT). • Scenario C is a rate change on a toll road, operated by a toll authority. A parallel highway with signal-controlled intersections might experience increased traffic diverting from the toll road. Each scenario illustrates the practitioner’s use of relevant tools at each step. In the supporting tables, tools are referred to by the short titles presented in the Summary in Table S-1. Introduction

382 Untolled Bridge to Tolled Bridge, P3 Step 1. Frame the Project Scenario A considers a bridge (not part of the Interstate system) built within the past 50 years as an untolled facility using federal assistance. To help finance the rehabilitation needed to improve the structural integrity of the structure and improve the functional operation of the facility, the state’s DOT is sponsoring implementation of a variable toll on the bridge (variable by time of day, congestion level, and type of vehicle). Although there would be no change in the number of travel lanes on the facility, traffic flow across the bridge and at its approaches is expected to improve with implementation of the variable toll. A contracted public-private partnership (P3) entity would design, construct, and manage the facility and be responsible for operations and maintenance, future refurbishment, repaying shareholders, and supporting negotiated transit services through the funds generated. Under the scenario, the state DOT would apply for loan guarantees from FHWA as part of financing the P3 with loans repaid by toll revenue. Tolls would be collected by transponders and license plate recognition with all-electronic tolling (AET) lanes. The bridge is the primary crossing from part of the region to the city central business district (CBD). Facility users who do not wish to pay the toll would have the option of using a non-tolled bridge located 3 miles away, which would be a more roundabout route to the CBD for most users of the bridge to be tolled. The practitioner refers to Checklist 1 in the Checklists section and finds useful definitions for toll system attributes and various context-setting questions. The checklist gives a preview of the types of questions that would likely need to be addressed throughout the study to support an EJ analysis. By reviewing the types of toll implementation actions in Step 1 of the Guidebook, the practi- tioner identifies that the proposed project is ready for toll implementation by fully converting all existing facility travel lanes to toll lanes. The practitioner then identifies which of the eight potential impact-causing activities of a tolling project would apply, as illustrated in the text box, Impact-Causing Activity of the Tolling Project. The practitioner consults Step 1 to identify the typical direct and indirect effects of these impact-causing activities associated with a full conversion of the existing bridge facility to all tolls with variable pricing. From Step 1, it can be learned that the potential direct effects could include the following: change in road use patterns (diversions to alternative routes or modes); change in trip-making behavior and trip purposes (trip frequency, trip timing); change in household disposable income and household financial burden; and change in “disposable time.” Reviewing prior regional plan studies prepared by the metropolitan planning organization (MPO), the practitioner observes that the principal highway that carries traffic to the subject bridge passes through communities that may have higher concentrations of low-income and S c e n a r i o a

Tools Why Practitioner Uses the Tool Checklist 1 To identify the type of toll implementation action from among the range of types. To identify the impact-causing aspects typically associated with the action from among the range of aspects. To consider potential adverse and beneficial direct and indirect effects that can be associated with an action and its aspects. To make an initial characterization of the context of connections of minority and low-income users of the tolled facility and various opportunities relative to user trip-making, income, and value-of-time. To preliminarily screen for EJ issues associated with the potentially adverse and beneficial effects of the toll implementation action for further exploration and evaluation in subsequent steps. To understand the toll system attributes under consideration and key definitions. To situate the project in a stage of transportation decision-making. To understand what studies on purpose and need, tolling concepts, and alternatives have already been conducted. To scan for existing transportation policies, plans, studies, and analytical models relevant to toll pricing and facility implementation. To consider what outreach has been previously undertaken with potentially affected EJ populations and representative organizations. l titi l li i tif f ll i l ti ti f f . i tif i - i i ll i i ti f f . i ti l i l i i i i i ti i . i iti l i ti f f ti f i i l -i f ll f ili i iti l ti i - i , i , l - f-ti . li i il f J i i i ti ll i l f ll i l ti ti f f l ti l ti i . ll tt i i ti iti . i j i f ti i i - i . i , lli , l ti l . f i ti ti li i , l , i , l ti l l l ll i i f ili i l ti . i i l i ti ll J l ti ti i ti . Step 1. Frame the project. Impact-Causing Activity of the Tolling Project Potential Impact-Causing Activity Applicable? Introduction of a transaction cost–paying a toll–on the facility user. Yes, for those facility users who want to continue to use the facility. Increase of a transaction cost to the facility user. Not applicable because the project would involve initial toll implementation. Creation of uncertainty in the transaction cost for the facility user (near term and/or longer term). There is some uncertainty in the toll payments for specific trips across the facility because of the variable toll. Form of payment by the user (i.e., cash or cashless). The tolls would be cashless. Newly implemented or revised toll payer account terms such as minimum balance, monthly fees, up-front transponder deposit or purchase costs, account replenishment options (cash, credit, debit), and discount plan (resident, commuter, low-income resident/commuter, HOV, “green” car, HOT, transponder owners). Those with accounts with the operating entity would be provided with a transponder for the toll transaction and a monthly bill. Those with transponders would need to pay a $25 deposit for the transponder and a monthly $1 service fee to maintain an account. Those without a transponder would have trips identified through license plate recognition, and the bill for the trip would be mailed to the vehicle’s owner at a higher cost. Change in access to the highway network for the user. No change in access to the highway network is expected. Creation of, change in, or increase in a toll facility footprint. A minimal increase in the footprint for the gantries and equipment for AET is expected. However, this addition would be within the operating right-of-way of the existing facility. Closer proximity between a community and a toll facility. No change in the location of the facility relative to nearby communities would occur.

384 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox minority populations before the bridge crossing to the CBD. U.S. Census data shows that the CBD is the governmental, commercial, and services center for the region and, by far, the largest employment center of the region. In discussing the project with the state DOT travel demand modeler assigned to the project, the practitioner learns that 10 to 15 percent of the traffic that now uses the bridge would use the alter- native route and bridge to access the CBD. Select link analysis of diverted trips from those traffic analysis zones that predominantly generate trips over the bridge and into the CBD indicates that select communities with higher concentrations of low-income and minority populations would have a higher propensity of diversion than other areas. Further, those who would benefit from the tolling in terms of improved travel times would be from areas that are not high-minority or low-income communities. The practitioner documents the results of Step 1 in a technical memorandum noting the potential for disproportionately high and adverse effects on minority and low-income popula- tions from the proposed tolling of the bridge. Further study is needed to assess how changes in travel behavior attributable to toll implementation might vary by income of traveler. This level of analysis is needed not only because of what the preliminary identification of issues reveals from the mapping and travel forecasts, but also because as the first toll implementation project in the region, there is no precedent in the region on which to base an assessment. It is possible that more in-depth technical studies related to travel patterns and behavior would be needed to identify a need for mitigation and monitoring in response to EJ considerations. It is likely the P3 partner would need to provide appropriate data, including surveys and other public outreach, to support such analysis. Step 2. Identify the Applicable Requirements Governing Decisions Having used Reference Table 2.1 in the Reference Tables section, the practitioner considers what policy and environmental review requirements may be triggered. • With regard to fulfilling the requirements of the National Environmental Policy Act (NEPA) analysis, the practitioner finds that the state DOT considers the project to be a categorical exclusion because the action involves deployment of electronic equipment [23 CFR 771.117(c) (21)]; takes place within the existing operational right-of-way [23 CFR 771.117(c)(22)]; and involves bridge rehabilitation [23 CFR 771.117(c)(28)]. However, in light of the potential EJ issue identified in Step 1, the practitioner advises the state DOT project manager that addi- tional studies should be conducted to assess the adverse and beneficial effects of the project on minority and low-income populations, including how the project’s toll pricing policies and design might minimize adverse effects or benefit EJ populations. • Full and fair participation of low-income and minority populations is one of the fundamen- tal principles of EJ, and meaningful public involvement processes should be employed at all stages of decision-making, including the environmental review under NEPA during project design. • The bridge project qualifies for the MAP-21 mainstreaming of tolling projects. DOT require- ments under 23 U.S.C. 129(a) include a recommended memorandum of understanding with the FHWA Division Office as well as annual audits and certifications on spending of funds received. FHWA provides information and templates for agreements online at: http://www. fhwa.dot.gov/ipd/revenue/road_pricing/tolling_pricing/section_129.aspx. The practitioner reviews select tools in Step 2.

Untolled Bridge to Tolled Bridge, P3 385 The practitioner documents the results of Step 2 in a technical memorandum describing the discussions with the state DOT manager on why a categorical exclusion is unlikely for this proj- ect because of the need for additional study. The technical memorandum cites the regulations and references the potential adverse impacts borne by low-income and minority populations identified in Step 1 and the need to further consider the potential for offsetting benefits for low- income and minority populations in the studies to be undertaken. Step 3. Recognize the Relevant Decision-Makers and Stakeholders The practitioner, who is already familiar with the planning stages, reads Step 3 in the Guide- book and refers to Reference Tables 3.1, 3.2, and 3.3 in the Reference Tables section to consider in detail and determine the range of stakeholders that would need to be engaged. • FHWA—all stages • State DOT—all stages • P3 entity—implementation stage • Minority and low-income populations—all stages beginning with planning stage The practitioner also seeks out the expertise of others in the agency who have led public involvement efforts to get their advice and contact lists. An experienced public involvement spe- cialist in the agency, who has substantial contacts in the minority communities near the bridge and surrounding areas, is available to work part-time throughout the project. The practitioner reviews select tools in Step 3. The practitioner documents the results of Step 3 in a technical memorandum describ- ing the stakeholders identified, including contact information for the DOT project manager, the P3 manager, and other community leaders and affected organizations in the study area, including the school board, the local director of public health, and the local director of social services, among others, using guidance provided in the “Public Involvement Plan” tool. Tools Why Practitioner Uses the Tools Reference Table 2.1: Applicable Requirements Governing Tolling Projects To identify the major legislative acts, regulatory requirements, authorities, and guidance governing tolling projects, including EJ, Title VI and nondiscrimination, environmental review, and transportation. To recognize guiding authorities for framing an EJ analysis and identifying the criteria for making an assessment of disproportionately high and adverse effects. Public Involvement Plan Community Characteristics Inventory To recognize that identifying and addressing low-income and minority populations requires attention to key guidance described in the FHWA and DOT orders and supporting reference guidance. To understand that “full and fair participation” by all potentially affected communities is a fundamental principle of EJ in transportation. To appreciate that EO 13166, “Improving Access to Services for Persons with Limited English Proficiency” (LEP) and Title VI of the Civil Rights Act are closely related. Preparing an LEP plan is an important step in addressing barriers to participation for foreign-born populations. Step 2. Identify the applicable requirements governing decisions.

386 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Step 4. Scope Approach to Measure and Address Impacts The practitioner begins to scope the project in earnest, within and outside the agency after reviewing the key questions of Step 4: • Who could be affected by the toll project or rate change? • Is there a potential EJ concern? • What are the needs, concerns, attitudes, and perceptions of EJ populations? During this step, the practitioner refines the project boundaries that inform the approach to technical studies, outreach, and impact assessment. The refined project boundaries include a direct impact area, defined as the area along the principal highway feeding the bridge before it crosses into the CBD from the point between the bridge and the beltway that would be the primary point for trips diverting to the alternative bridge route and an extended impact area, defined as the primary catchment area for bridge users based on a license plate survey of existing bridge users and on the travel demand model. Working with the state DOT’s public involvement specialist, the practitioner secures space and sets up a project information kiosk at community street fairs and other gatherings hosted by civic and community organizations in the affected communities. In addition, the P3 private operator has a budget for public outreach and agrees to work on potential outreach strategies, including surveys and focus groups. The practitioner is encouraged to work with leadership from several respected local community organizations to help in recruiting focus groups near affected communities. The practitioner works with Census-related datasets and the state DOT’s travel modeler to explore regional travel patterns and how they differ by race and income. The P3 private operator has some non-confidential information to share in this regard from its traffic and revenue estimates and value-of-time by household income estimates. • The locations of EJ populations are mapped for the two study areas. Census block groups with higher concentrations of low-income and minority populations are identified based on a “meaningfully greater” threshold method using the two counties in the extended areas as reference communities. The locational patterns are mapped using GIS tools; race, income, and other social characteristics such as linguistic isolation are also inventoried. • The study area boundaries and initial mapping of locations of the EJ populations based on Census data are shared with leadership from several community organizations in some early scoping meetings to determine if other low-income and minority populations may have been overlooked. Tools Why Practitioner Uses the Tool Reference Tables 3.1, 3.2, and 3.3: Decision-makers and Stakeholders: Actions, Decisions, and Concerns – Policy and Planning, Project Design including NEPA, and Implementation To appreciate the responsibilities of agencies and stakeholders as they relate to tolling decisions and the assessment of EJ impacts. To ensure scoping and outreach processes are robust and inclusive. To make sure the “right” organizations are at the table at various stages. Public Involvement Plan To identify potential community leaders and organizations to contact based on local research. Step 3. Recognize the relevant decision-makers and stakeholders.

Untolled Bridge to Tolled Bridge, P3 387 • Major concerns and questions are raised in outreach discussions and through focus groups: – Mobility, Access, and Safety: The AET system could be a barrier for those without a bank account or credit card if cash replenishment options are not available or readily accessible. Would there be fees for transponders? Would minimum balances be required on transpon- ders? Could those fees or minimum balances be waived? Could low-income individuals be charged a lower rate? – Social and Economic: How much impact could the toll have on the budget of a low-income household? Could the toll limit access to work, schools, medical facilities, and/or recreation? – Physical and Environmental: The closest non-toll bridge is 3 miles away. Would low-income people have to reduce trip-making or make long detours that take more time based on the expected peak and non-peak toll levels? Would buses travel toll-free and improve services, making that a potential option? Could HOVs travel free or at reduced rates? The practitioner reviews the tools that would be used in Step 4. The practitioner refers to Checklist 2 in the Checklists section to document the analytical steps and process for each of the resource areas: mobility, access, and safety; social and economic; physical and environmental; and cultural and historic resources. The practitioner continues to thoroughly document the public outreach efforts and the outcomes of those efforts, in terms of Tools Why Practitioner Uses the Tools Checklist 2 To ensure that findings are based on careful, step-by-step reasoning. Community Characteristics Inventory To examine criteria for defining project and regional travelshed study areas for impact assessment. To identify data sources and variables for preparing community profiles of race, income, and other social characteristics to identify concentrations of EJ populations. This may include Census and other data sources drawn from online tools and local contacts (e.g., social services, housing authorities, and schools) for more insight into locations of minority or low-income populations. To consider methods and limitations of applying “meaningfully greater,” Council on Environmental Quality (CEQ) guidance and “no-threshold” criteria as means for identifying locations of low-income and minority populations. To recognize limitations of various secondary datasets such as the Census, and the importance of scoping interviews, field work, surveys, focus groups, and community engagement to better understand the locations, concerns, and travel behavior of protected populations. Using PUMS Using NHTS Travel Demand Models To understand regional and local travel patterns that facilitate understanding of similarities and differences in travel behavior by race, income, and other social characteristics. To introduce the travel demand model colleagues to the types of detailed information more supportive of analysis of tolling impacts by income and, potentially, race. User Costs, HH Burden To explore local data and develop estimates of household burden for the proposed toll based on income and, potentially, race. This analysis would be used to support community discussions and address concerns about potential cost impacts. Step 4. Scope approach to measure and address impacts. (continued on next page)

388 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Tools Why Practitioner Uses the Tools Reference Table 4.1: Examples of Resource Topic Considerations Added by Tolling Public Involvement Plan Focus Groups Conducting Surveys Case Examples: Surveys with Local Liaison, KY-IN Surveys at Grocery Stores KY-IN Avoid Impacts FL Bridge To ramp up the public involvement of EJ stakeholders identified in Step 3. The questions in Reference Table 4.1, Tolling Considerations, would help spark fruitful conversations. To recognize potential barriers to participation and prepare a Public Involvement Plan that, in this case, includes outreach with small group meetings at neighborhood centers and information booths at community events. To work with trusted liaisons and community groups to convene focus groups on both sides of the river with EJ populations to delve deeper into preliminary findings from neighborhood outreach. Step 4. (Continued). numbers; low-income and minority representation; and ideas, concerns and feedback on study areas and other issues shared by participants. The practitioner asks the modelers to document their methods and summarize their findings in non-technical terms; the practitioner documents the research findings and summarizes the implications of the initial model efforts on travel pat- terns, behavior, and likely financial or time burdens for low-income households. Step 5. Conduct Impact Analysis and Measurement The practitioner uses Reference Table 5.1 in the Reference Tables section to make sure that the key topics affected by toll implementation or rate change are examined. Subject matter experts in air quality, noise, cultural cohesion, cultural and historical resources, and natural resources are asked to assess whether there could be adverse impacts or benefits to low-income or minor- ity populations for their specific resource areas. The impact assessment considers the adverse and beneficial effects, including direct, indirect, and cumulative impacts as well as short-term (construction) and long-term (operational) impacts. The practitioner gets familiar with the “Travel Demand Models” and “Select Link Analysis” tools and the type of information that can be obtained from the models, including level of detail on income levels and travel patterns. The practitioner conveys this information to the MPO travel and air quality modelers and explains the types of impacts the community is concerned about and requests that they analyze these impacts. • The practitioner confirms the modeler’s expectation from Step 1 that 10 to 15 percent of the traffic that now uses the bridge would use the alternative route and bridge to access the CBD. Select link analysis of diverted trips from those traffic analysis zones that predominantly gen- erate trips over the bridge and into the CBD indicates that a higher propensity of diversions would occur from places with higher concentrations of low-income and minority populations (i.e., communities of concern) than from other areas. Further, those who would benefit from

Untolled Bridge to Tolled Bridge, P3 389 the tolling in terms of improved travel times would be from areas that are not classified as communities of concern. • Based on regional models, disproportionately high and adverse effects on minority and low- income populations are possible from the proposed tolling of the bridge vis-à-vis dimin- ished mobility and accessibility to opportunities in the CBD. In addition, it appears that minority and low-income populations would not necessarily share in the benefits of the project. The practitioner references several tools related to surveys and the “Surveys at Grocery Stores KY-IN” and “Pre- and Post-Surveys Atlanta” case example to design and implement a sur- vey that investigates travel behavior and attitudes about tolling that has been tailored to better understand the impact on low-income populations. The practitioner secures the services of a market research firm and asks that questions be designed to permit a pre- and post-implemen- tation monitoring of the toll bridges effects. The practitioner uses a combination of intercept methods, license plate, and computer-assisted telephone interview (CATI) survey techniques and an inclusive and robust sampling frame to facilitate a comparative assessment of income segments’ (including low-income populations’) attitudes toward tolling, transponder usage, and travel behavior in the peak and off-peak periods. The practitioner uses focus group discussions to assist in the initial survey design and to explore issues related to transponder usage and tolling account policies. The practitioner continues to extend the outreach efforts and, with the assistance of the public involvement specialist, fosters a stronger relationship with several community organizations who assist in providing venues and publicizing a series of community meetings as well as assist in recruiting low-income and minority populations for select focus groups. Several tools are referenced in Step 5. Tools Why Practitioner Uses the Tools Mobility, Access, and Safety Travel Demand Models Select Link Analysis Using PUMS Using NHTS To evaluate impacts on: Travel accessibility - The closest non-toll bridge is 3 miles away. Travel models and outreach indicate that low-income and minority populations use the bridge for commuting, access to hospitals and social services, and education. - The toll could increase household costs and reduce trip-making for some trip purposes. - A large minority of low-income households do not have credit cards or bank accounts. Purchasing and maintaining a minimum balance for a transponder could be a significant financial burden. - Without convenient access to cash replenishment options, the transponder could become a barrier to use. Transportation choice - Bus service crossing the bridge is infrequent and may not serve all destinations, especially for late night and early morning shift work. Transportation safety impacts - No anticipated impacts. Step 5. Conduct impact analysis and measurement. (continued on next page)

390 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Continuing documentation, the practitioner works through the Step 5 portion of Checklist 2 in the Checklists section for each of the resource areas: mobility, access, and safety; social and economic; physical and environmental; and cultural and historic resources. The practitioner con- tinues to thoroughly document the public outreach efforts and the outcomes of those efforts, including developing, conducting, and recording the findings of the survey, as well as contin- ued face-to-face outreach. With additional information from the survey, the modelers are able to better calibrate and validate their models. The practitioner asks the modelers to document their methods and summarize their findings in non-technical terms; the practitioner documents research findings and summarizes the implications of the advanced model efforts on travel pat- terns, behavior, and likely financial or time burdens for low-income households. The practitioner also aggregates and tracks the findings of the other resource topic areas from the subject matter experts on air quality, noise, cultural cohesion, and cultural and historical resources. Tools Why Practitioner Uses the Tools Cultural and Historic Resources Reference Table 4.1 Public Involvement Plan To evaluate impacts on: Cultural and Historic Resources - No anticipated impact because the bridge is less than 50 years old. - Public outreach does not identify concerns. Social and Economic VOT/Willing to Pay(using survey results) User Costs, HH Burden (using refined data including survey results) Public Involvement Plan Focus Groups Conducting Surveys To evaluate impacts on: Community cohesion - The toll would have a limited impact based on community input by disrupting some social activities across the river. Economic impacts, including household burden - Those traveling without a transponder would be billed a higher fee, and possibly an administrative fee higher than the toll itself. - The toll is expected to be burdensome if the rate exceeds $2 each way (hypothetical) or $20/week, which would be 10% of income for a part- time employee at 30 hours per week at minimum wage. Noise - No anticipated impact other than temporary construction noise. Visual quality - No change/no anticipated impact. Land use, prices, and property values - No anticipated impact. Physical and Environmental Travel Demand Models To evaluate impacts on: Air quality - Improved traffic flow on bridge, less congestion on approaches, and reduced travel. The alternative route has sufficient capacity to absorb diverted trips without a degradation in traffic flow level-of-service or in air quality. Hazardous materials - No anticipated change in patterns of carrier use or changes in risk. Water quality and drainage - Improvements in roadside water filtration systems to capture salt, oil, and other run-off could result in improved water quality. Surveys at Grocery Stores KY-IN Pre- and Post-Surveys Atlanta Step 5. (Continued).

Untolled Bridge to Tolled Bridge, P3 391 Step 6. Identify and Assess Mitigation Strategies The surveys, analysis, and continued outreach from Step 5 identify mobility, access, and safety impacts that are adverse and may disproportionately burden low-income populations. The practitioner reviews the impact analyses from Step 5 in light of the definitions of dispro- portionately high and adverse effects. The practitioner reads the methods and criteria for con- sideration in the Guidebook and applies the “Disproportionate Effects” tool. These preliminary findings are consistent with feedback from many low-income and minority residents and bridge users through outreach and surveys. The practitioner convenes discussions with the responsible FHWA representative to confirm their collective determination. The impact assessment of adverse effects and benefits conducted in Step 5 are compared in various ways such as project/no project (Build vs. No-Build) and project impact on EJ popula- tions versus impact on non-EJ populations. The analyses build from Step 5 analysis using analyti- cal and outreach tools and applying the methods from the “Disproportionate Effects” tool. The percentage difference tests reveal adverse impacts borne by low-income populations that would be appreciably more severe and greater in magnitude with regard to social and economic impacts. In assessing mitigation strategies, the practitioner notes that the “Avoid Impacts FL Bridge” case example that evaluated the potential for having a free local road access to the bridge alter- native is not feasible. However, the “Full Cycle I-10–I-110 LA County” and the “Mitigation NCTCOG” case examples suggest the potential to devise a low-income assistance plan extend- ing credits to public transportation users or frequent bridge crossing users. The practitioner arranges a study to evaluate the feasibility in terms of revenues and costs with the understanding that mitigation for some users, in some form, could be a condition for FHWA project approval in light of the disproportionately high and adverse user cost impacts on low-income populations. Several tools are relevant for this step in terms of supporting impact assessment, mitigation, and the application of criteria relevant to formulating an assessment of disproportionately high and adverse effects as shown in Step 6. Tools Why Practitioner Uses the Tools Disproportionate Effects Travel Demand Models Select Link Analysis Conducting Survey Pre- and Post- Surveys Atlanta Public Involvement Plan Focus Groups Determine disproportionately high and adverse effects on: Transportation accessibility - The toll increases costs and reduces trip-making for some trip purposes among EJ populations. - The purchase of a transponder, minimum monthly balance, and replenishment options limited to credit cards and bank accounts present adverse impacts that are appreciably more severe for low-income and minority populations than for higher-income populations. - Avoiding the toll by traveling several additional miles each way costs time and money, and again, the impact is appreciably more severe and greater in magnitude for low-income populations. Transportation choice - Bus service crossing the bridge is infrequent with a limited service area, and the adverse effects would be predominantly borne by low-income populations. Transportation safety impacts - No disproportionately high and adverse impacts. Mobility, Access, and Safety Step 6. Identify and assess mitigation strategies. (continued on next page)

392 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Tools Why Practitioner Uses the Tools Conducting Surveys Full Cycle I-10– I-110 LA County Mitigation NCTCOG Pre- and Post- Surveys Atlanta User Activity Monitoring or cash payment (through arrangement with local grocery stores and convenience stores). The bridge authority would partially subsidize additional bus service, and buses and qualified van pools can use the bridge for no fee. (The practitioner conducts research and confirms that such strategies are legal in the state.) Regular bus riders would earn credits that entitle them to a discounted bus ride or a credit discount for use of the toll facility. The bridge authority and transit agency would investigate arrangements with employers to provide late night and early morning shift van shuttle services. The bridge authority would fund an objective "after" survey on travel patterns and usage, comparable to the "before" survey, to monitor effectiveness of mitigation and usage by different populations. The toll private operator would need to share user account information illustrating spatial patterns of travel to support monitoring. Social and Economic Disproportionate Effects VOT/Willing to Pay User Costs, HH Burden Public Involvement Plan Focus Groups Conducting Surveys Determine disproportionately high and adverse effects on: Community cohesion - No disproportionately high and adverse impacts. Economic development and other economic impacts, including household burden - Community outreach, surveys, models, and analysis of the toll’s impact on low- wage workers indicate that the adverse effect would be appreciably more severe and greater in magnitude for low-income populations. Noise - No disproportionately high and adverse impacts except in construction. Visual quality - No disproportionately high and adverse impacts. Land use, prices and property values - No disproportionately high and adverse impacts. Physical and Environmental Disproportionate Effects Determine disproportionately high and adverse effects on: Air quality - No disproportionately high and adverse impacts. Hazardous materials - No disproportionately high and adverse impacts. Water quality and drainage - No disproportionately high and adverse impacts. Cultural and Historical Resources Disproportionate Effects No disproportionately high and adverse impacts. Mitigation and Commitments Recycling Toll Revenue Replenishment Options Mitigation alternatives are discussed and agreed upon with the community. The agency drafts commitments that would need to be implemented, in part, through agreements with the toll private operator and other entities: For income-qualified low-income households, distribution of specially coded transponders with no up-front fee and no monthly maintenance fee. User accounts could be replenished by the applicant with a credit card, debit card, Step 6. (Continued). The practitioner uses the Step 6 portion of Checklist 2 from the Checklists section for each research area, the “Disproportionate Effects” tool, and the backup documentation for each resource topic analysis to document the findings of disproportionately high and adverse effects (or in some cases the dismissal of the resource topic), including the mitigation agreements with the community and the agreements for long-term monitoring.

Untolled Bridge to Tolled Bridge, P3 393 Step 7. Document Results for Decision-Makers and the Public The practitioner uses Checklist 2 in the Checklists section to ensure thorough documentation of various outreach and analytical steps (Steps 4, 5, and 6) that provide a sound, technical basis for the EJ impact findings for each of the resource topics. Using the “Public Involvement Plan” tool, the practitioner is more aware of potential barriers to public involvement from tradition- ally underserved populations and more proactive in making efforts to overcome these barriers. The practitioner thoroughly documents public involvement activities throughout the process, including her team’s efforts to foster meaningful involvement opportunities with low-income and minority populations. Mitigation commitments are clearly identified that have been made in support of the EJ impact findings. The documentation distills the analysis from previous steps by resource topic to substantiate the rational basis for impact assessment, the disproportionately high and adverse effect findings, EJ findings, and mitigation commitments, as shown in Step 7. Tools Why Practitioner Uses the Tools Checklist 2 Multiple tools from previous steps Public Involvement Plan Mobility, Access, and Safety Transportation accessibility - Document the disproportionate financial burden and/or time burden that would be appreciably more severe and greater in magnitude for low-income populations and resulting mitigation commitments, including no minimum balance and reduced fee transponders with cash replenishment options and form of credit for frequent users of public transportation and drivers who qualify for low-income assistance plan. Transportation choice - Document agreements to support improved bus service and investigate encouraging employers to provide late night and early morning shift van shuttle services. Transportation safety impacts - Document no anticipated impacts. Social and Economic Community cohesion - Document that the proposed transportation accessibility mitigation would address this concern to the satisfaction of the affected community. Economic development and other economic impacts, including household burden - Document the disproportionately high and adverse impact, and that the mitigation proposed for transportation accessibility would also address household burden. Noise - Document no anticipated impact. Visual quality - Document no anticipated impact. Land use, prices, and property values - Document no anticipated impact. Physical and Environmental Air quality - Document no disproportionate impact. Hazardous materials - Document no anticipated impact. Water quality and drainage - Document a potential beneficial impact. Cultural and Historical Resources - Document no anticipated impact. Step 7. Document results for decision-makers and the public.

394 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Step 8. Conduct Post-Implementation Monitoring The practitioner advises the agency that it is appropriate to monitor whether the travel diver- sion effects predicted in the travel demand model prove to be true after operations begin and to assess whether the cash replenishment option and low-income assistance/toll credit mitigation commitments are effective in practice. A post-implementation survey and user activity monitor- ing of the toll account records are also recommended. Monitoring commitments are clearly defined, including milestones for definitive measure- ments; ongoing monitoring mechanisms; what would be measured; how it would be measured; what constitutes a threshold for taking action; and, if action is needed, who would initiate the action and with what public and private operator consultation, if appropriate. Several tools describe strategies for post-implementation monitoring that could apply under this scenario (Step 8). Tools Why Practitioner Uses the Tools Public Involvement Plan Recycling Toll Revenue User Activity Monitoring Full Cycle I-10–I-110 LA County Pre- and Post-Surveys Atlanta Determine monitoring strategies: Mobility, Access, and Safety Transportation accessibility - Monitor income-qualifying eligibility levels and low-income population use of the toll facility. - Monitor enrollments and use of reduced fee transponders and readily accessible cash replenishment options. Establish sunset/renewal provisions for existing users to prevent or limit fraud. - Publicize programs to ensure that new, potentially qualified people have information about registering. Transportation choice - Monitor bus service and ongoing agreement. - Monitor bus ridership on subsidized services. - Monitor potential implementation of vanpool or shuttle services working with community members, toll authority, local transit agency, local commuter service center, and employers, if necessary. Social and Economic Community cohesion - Work with community members to monitor any community cohesion challenges that may arise as a result of the toll. Economic development and other economic impacts, including household burden - Conduct travel behavior and attitude survey of crossing users within 12 and 24 months of full-implementation to explore usage levels and impacts of low-income assistance/credit programs by income and race. - Periodically monitor household burden and work with community and other stakeholders to modify transportation accessibility and transportation choice programs as necessary to alleviate the household burden on low-income users. Step 8. Conduct post-implementation monitoring.

395 Step 1. Frame the Project Scenario B considers a 6-lane roadway (Interstate) with one HOV lane in each direction. Pur- suant to 23 U.S.C. Section 166, the state DOT would seek FHWA authorization to convert the presently underutilized HOV lanes to HOT lanes. The HOT lanes would have variable pricing (by time of day, congestion level, and type of vehicle) on the HOT lanes priced to keep the lanes at close to free-flow speeds, maintaining free HOV service at HOV 3+ levels. This represents a change from the current free HOV service at HOV 2+. Tolls on the HOT lanes would be collected via transponders with AET. HOV transponders would be equipped with on-off switches to go from free to paid operation depending on the number of riders. A toll authority arm of the state DOT would manage and operate the entire facility. The conversion is intended to (1) increase utilization of HOV/HOT lanes, (2) decrease overall highway congestion, and (3) raise funds to support the cost of roadway operations and maintenance. In consulting the types of toll implementation actions in Step 1 in the Guidebook, the prac- titioner classifies the proposed project as a toll implementation by partially converting exist- ing facility travel lanes to have a mix of toll and general purpose lanes. The practitioner then identifies which of the eight potential impact-causing activities of a tolling project apply, as illustrated in the text box, Impact-Causing Activity of the Tolling Project. The practitioner then consults Step 1 in the Guidebook to identify the typical direct and indi- rect effects of these impact-causing activities associated with a partial conversion of the existing highway to dedicate one travel lane, the existing HOV lane, to a HOT lane. From Step 1, it can be learned that the potential direct effects could include the following: change in road use patterns (diversions to alternative routes or modes); change in trip-making behavior and trip purposes (trip frequency, trip timing); change in household disposable income and household financial burden; and change in “disposable time.” In discussing the project with the state DOT travel demand modeler assigned to the project, the practitioner learns that the travel model shows that the net effect of the project would be an overall reduction in congestion on the highway because more vehicles would use the existing underutilized HOV lane when it is converted to a HOT lane. The travel modeler also notes to the practitioner that little, if any, diversion of traffic would occur from the highway with HOT lane implementation because of the overall improvement in traffic flow relative to existing conditions and that more vehicles may be attracted to the highway during certain travel times because of the improved traffic flow (the term for this attracted traffic is latent demand). The practitioner documents the results of Step 1 in a technical memorandum noting that, while there is some potential for adverse impact on minority and low-income communities that must be explored further, the project purpose and expectation of overall improved traffic flow S c e n a r i o B HOV Lanes to HOT Lanes

396 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Impact-Causing Activity of the Tolling Project Potential Impact-Causing Activity Applicable? Introduction of a transaction cost–paying a toll– on the facility user. Yes, for those facility users who wish to use the HOT lane and are not HOV-3. The change from HOV 2+ to HOV 3+ would create an additional burden for HOV 2 drivers to add another rider to the carpool in order to avoid the toll. Increase of a transaction cost to the facility user. Not applicable because the project would involve initial toll implementation. Creation of uncertainty in the transaction cost for the facility user (near-term and/or longer term). There is some uncertainty in the toll payments for specific trips in the HOT lane because of the variable toll. Form of payment by the user (i.e., cash or cashless). The tolls would be cashless. Newly implemented or revised toll payer account terms such as minimum balance, monthly fees, up-front transponder deposit or purchase costs, account replenishment options (cash, credit, debit), and discount plan (resident, commuter, low-income resident/commuter, HOV, “green” car, HOT, transponder owners). Current HOV 3+ users would be required to obtain a transponder; many systems have a switch on the trans- ponder that can shift from HOV to single-occupancy vehicle (SOV)/HOV 2 mode, depending on the number of riders. No initial fee or regular monthly maintenance fee is proposed for HOV 3+ registered vehicles. Current HOV 2+ users would be required to recruit another rider or pay the full fees. SOV users (full price or discounted) would need to establish an account with the toll authority, and would receive a transponder for the toll transaction and a monthly bill. Full price transponder users would need to pay a $25 deposit for the transponder and a monthly $1 service fee to maintain an account. If low-energy “green” vehicles are allowed free access, this may be perceived as an advantage to those who are well-off. Changing HOV 2+ to HOV 3+ may warrant additional review of other exemptions. Those without a transponder would have trips identi- fied through license plate recognition, and the bill for the trip would be mailed to the vehicle’s owner at a higher rate than the transponder rate, plus fee. Change in access to the highway network for the user. No change in access to the highway network would occur because HOT access to highway exits would be the same as the existing HOV access. Creation of, change in, or increase in a toll facility footprint. A minimal increase in the footprint for the gantries and equipment for AET would be required. However, this addition would be within the operating right-of- way of the existing facility. Closer proximity between a community and a toll facility. No change in the location of the facility relative to nearby communities.

HoV Lanes to HoT Lanes 397 should mean that the potential for disproportionately high and adverse effects on minority and low-income populations is low. In addition, because all users of the facility would benefit from the project, it appears that minority and low-income populations would share in the benefits of the project. The practitioner notes that a low-income housing complex abuts the highway. Because of the already high volume of traffic that uses this highway, the practitioner wants to evaluate the potential for latent demand (more traffic attracted to the highway with improved traffic flow) and, if there is a basis for latent demand, what effect higher traffic volumes may have on air quality and health in the low-income housing complex. The practitioner refers to and completes Checklist 1 in the Checklists section and finds useful definitions for toll system attributes and various context-setting questions to get him better situ- ated. The practitioner discovers that the checklist provides a preview of the types of questions that would likely need to be addressed throughout the study to support an EJ analysis. Step 2. Identify the Applicable Requirements Governing Decisions Having used Reference Table 2.1 in the Reference Tables section, the practitioner begins to consider what policy and environmental review requirements may be triggered. • Public engagement of minority and low-income populations. Minority and low-income popula- tions would need to be engaged beginning with the planning stage and continuing throughout the right-of-way acquisition, construction and operations, and maintenance stages. State DOT planners would refer primarily to the public involvement plan tool beginning in Step 3. • Environmental processing. With regard to fulfilling the requirements of the NEPA analysis, the practitioner finds that the state DOT considers the project to be a categorical exclusion because the action involves deployment of electronic equipment [23 CFR 771.117(c)(21)] and takes place within the existing operational right-of-way [23 CFR 771.117(c)(22)]. Although the potential for EJ impacts is relatively low, based on the assessment in Step 1, the practitio- ner advises the state DOT project manager about the potential for adverse air quality effects on the low-income housing complex if additional traffic is attracted to the highway. • Air quality conformity. The region is in non-attainment for National Ambient Air Quality standards for fine particulate matter (PM2.5), and an air quality conformity analysis would be needed as part of demonstrating conformity of the MPO’s long-range transportation plan with the state’s air quality improvement plan (the State Implementation Plan). A PM2.5 analysis could also be required because of potential air quality and health impacts on the low-income housing complex, and a quantitative mobile-source air toxics analysis evaluation would be anticipated on the basis of relatively high traffic volumes. The practitioner refers to the state DOT’s air quality specialists for analysis (beginning in Step 4). • MPO endorsement. Under Section 166 of Title 23, existing HOV lanes may be converted to tolled operation provided that the local MPO endorses the use and amount of tolls on the converted lanes. The practitioner checks to see whether or not this endorsement has been obtained and that the amount of tolls endorsed by the MPO, in this case, the assump- tions regarding varying tolls, is consistent with the travel and air quality model input for assumed tolls. • Annual reporting. The practitioner notes that ongoing annual reporting documenting con- ditions on the converted lanes is required under 23 U.S.C. 166, and if the HOV facility becomes degraded the sponsor must bring the facility into compliance either by increasing HOV occupancy requirements, increasing tolls, increasing capacity, or eliminating access to paying motorists. The practitioner advises the state DOT project manager that, even though there may not be EJ effects, because the project is planned and based on assumptions about

398 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox HOT lane and general lane performance, it would be prudent to make provisions for includ- ing an EJ assessment in a future annual report if actual conditions differ from expected conditions. In planning for this HOV to HOT lane conversion, some concerns may be expressed that the primary function of the HOV lanes—to provide an incentive to form carpools and to “reward” such an incentive by guaranteeing a mostly delay-free trip—would be degraded. However, to avoid this, an HOV facility that allows tolled vehicles or any class of qualifying energy-efficient vehicles must annually certify that the subject lanes are “not degraded.” This annual certifica- tion states that, for a 180-day continuous reporting period, the lane(s) operate at greater than 45 miles per hour for 90 percent of the time. The practitioner documents the results of Step 2 in a technical memorandum describing the discussions with the state DOT manager on the need for analysis for adverse air quality effects on the low-income housing complex if additional traffic is attracted to the highway. The technical memorandum cites the regulations and references the likelihood of benefits for all users and the need to evaluate the potential for induced demand on the facility. Step 3. Recognize the Relevant Decision-Makers and Stakeholders The practitioner reads Step 3 of the Guidebook and refers to Reference Tables 3.1, 3.2, and 3.3 in the References Tables section to consider in detail and determine the range of stakeholders that would need to be engaged. Major stakeholders include the following: • FHWA—all stages • State DOT—all stages • Toll Authority—all stages • MPO—planning and project design stages • Minority and low-income populations—all stages beginning with the planning stage In this Step, the practitioner also refers to the “Public Involvement Plan” tool to develop a public involvement plan that would seek out additional community leaders for information and identify venues to meet with low-income and minority populations. The practitioner reviews select tools in Step 3. The practitioner documents the results of Step 3 in a technical memorandum describing the stakeholders identified, including contact information for the FHWA regional representative, the DOT project manager, the MPO, the local director of public health, the local director of social services, and community leaders, using guidance provided in the “Public Involvement Plan” tool. Step 2. Identify the applicable requirements governing decisions. Tools Why Practitioner Uses the Tools Reference Table 2.1: Applicable Requirements Governing Tolling Projects To identify the major legislative acts, regulatory requirements, authorities, and guidance governing tolling projects, including EJ, Title VI and nondiscrimination, environmental review, and transportation. To recognize guiding authorities for framing an EJ analysis and identifying the criteria for making an assessment of disproportionately high and adverse effects.

HoV Lanes to HoT Lanes 399 Step 4. Scope Approach to Measure and Address Impacts The practitioner gathers information to answer the key questions: • Who would be affected by this project? • Is there a potential EJ concern? • What are the needs, concerns, attitudes, and perceptions of EJ population groups about the proposed project and its impacts? The practitioner works with a planner skilled in demographics and GIS mapping to confirm the locations of low-income and minority populations within the project study area. In this Step, the practitioner develops some working assumptions about project boundaries, but discovers in a field visit that the defined direct impact area contains not only the identified housing complex, but the corridor has several small clusters of minority and low-income populations that are not identified using the “meaningfully greater” threshold approach. A “Select Link Analysis” determines that 80 percent of all trips in and through the corridor are to or from 150 traffic analysis zones in the regional model. These zones are defined as the extended impact area, with minor modifications to add zones for contiguous analysis. Many EJ communities are located in clusters along the corridor and in the adjoining counties. The results of the initial mapping of locations of EJ populations are shared in some early scoping meetings with community and civic leaders, which include discussions of the purpose and need for the contemplated project and its anticipated benefits and possible impacts. Major concerns and questions are raised in discussion and research: • Mobility, access, and safety: Would low-income people need to change their travel times, join or expand a carpool or vanpool, take transit, or prepare for longer travel times to avoid tolls? Would other travelers now decide to use the interstate when they had avoided it because of congestion? • Social and economic: What would be the travel times on the general purpose (free) lanes at rush hours? Would the toll limit access to work, schools, medical facilities, and/or recreation? • Physical and environmental: What would be the effect on air quality in the housing complex adjacent to the facility? The practitioner refers to Checklist 2 in the Checklists section to document the analytical steps and process for each of the resource areas: mobility, access, and safety; social and economic; physical and environmental; and cultural and historic resources. The practitioner continues to thoroughly document the public outreach efforts and the outcomes of those efforts, in terms of numbers; low-income and minority representation; and ideas, concerns, and feedback on study areas and other issued shared by participants. The practitioner asks the modelers to doc- ument their methods and summarize their findings in non-technical terms. The practitioner Step 3. Recognize the relevant decision-makers and stakeholders. Tools Why Practitioner Uses the Tool Reference Tables 3.1, 3.2, and 3.3: Decision-Makers and Stakeholders: Actions, Decisions, and Concerns (Policy and Planning, Project Design including NEPA, and Implementation) To appreciate the responsibilities of agencies and stakeholders as they relate to tolling decisions and the assessment of EJ impacts. To ensure scoping and outreach processes are robust and inclusive. To make sure the “right” organizations are at the table at various stages. Public Involvement Plan To identify potential community leaders and organizations to contact based on local research.

400 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox documents the research findings and summarizes the implications of their initial model efforts on travel patterns and behavior, in particular induced demand. Step 5. Conduct Impact Analysis and Measurement The practitioner refers to Reference Table 5.1 in the Reference Tables section and mobilizes subject matter experts to conduct impact assessment of the toll implementation. The practitio- ner closely considers for each resource topic whether there could be adverse impacts or benefits to low-income or minority populations. Continuing documentation, the practitioner works through the Step 5 portion of Checklist 2 in the Checklists section for each of the resource areas: mobility, access, and safety; social and economic; and physical and environmental; and cultural and historic resources. The practitio- ner continues to thoroughly document the public outreach efforts and the outcomes of those efforts. Traffic models forecast little induced demand, with air quality impacts from more traf- fic offset by higher operating speeds and fewer emissions. The practitioner asks the modelers to document their methods and summarize their findings in non-technical terms. The prac- titioner documents the research findings as well as the implications of the advanced model- ing efforts on travel patterns, behavior, and likely financial or time burdens for low-income Step 4. Scope approach to measure and address impacts. Tools Why Practitioner Uses the Tools Checklist 2 To ensure that findings are based on careful, step-by-step reasoning. Community Characteristics Inventory To examine criteria for defining project and regional travelshed study areas for impact assessment. To identify data sources and variables for preparing community profiles of race, income, and other social characteristics to identify concentrations of EJ populations. To consider methods and limitations of applying “meaningfully greater,” CEQ guidance and “no-threshold” criteria as means for identifying locations of low-income and minority populations. To recognize limitations of various secondary datasets such as the Census, and the importance of scoping interviews, field work, surveys, focus groups, and community engagement to better understand the locations, concerns, and travel behavior of protected populations. To identify potential issues of concern (e.g., low-income housing adjacent to roadway). Travel Demand Models Select Link Analysis VOT/Willing to Pay To understand regional and local travel patterns and likely future patterns. Public Involvement Plan Conducting Surveys Case Examples: Surveys at Grocery Stores KY-IN To think through strategy for conducting outreach through meetings at churches, neighborhood centers, a local Hispanic- oriented mall, and booths at community events. To develop questions for a toll attitude and behavior survey. To refer to examples to ensure a broad distribution and good response rate from EJ populations.

HoV Lanes to HoT Lanes 401 Step 5. Conduct impact analysis and measurement. Tools Why Practitioner Uses the Tools Mobility, Access, and Safety Travel Demand Models Select Link Analysis Full Cycle I-10–I-110 LA County Pre- and Post- Surveys Atlanta To evaluate impacts on: Travel accessibility - Limited impact. Travel models and outreach indicate that travel speeds in the general purpose lanes would improve over existing conditions because some vehicles would pay a toll to take advantage of superior travel speeds in the HOT lanes. Transportation choice - Limited impact. In addition to general purpose lanes, EJ populations would have the option of using transit and HOV 3+ on toll lanes at no charge (although current HOV 2 drivers and riders will have to recruit another rider); many commuter vans are currently in use in the region, and bus service is frequent along the corridor. Transportation safety impacts - No anticipated impacts. Social and Economic VOT/Willing to Pay User Costs, HH Burden Public Involvement Plan Conducting Surveys Full Cycle I-10–I-110 LA County Pre- and Post- Surveys Atlanta To evaluate impacts on: Community cohesion - No impact is anticipated. Economic development and other economic impacts, including household burden - Low-income households with regular commute patterns are expected to form, expand, or join carpool or vanpool options or use transit in order to use the HOV/HOT lanes for free or use the general purpose lanes ifshared rides are not feasible. - Low-income households without established HOV arrangements would not be anticipated to use toll lanes during peak hours except in emergencies. Use of general purpose lanes on a regular basis would not be burdensome if the anticipated congestion reduction in the general purpose lanes materializes. - Those who do not have a transponder who feel compelled to use the HOT lane for whatever reason may be charged a higher fee, possibly including an administrative fee higher than the toll itself. Noise - Noise modeling indicates an imperceptible change in sound levels with higher travel speeds and higher peak hour traffic volumes. Visual quality - Minor with AET detection equipment gantries. Land use, prices, and property values - Minor changes because accessibility changes would be minor. Physical and Environmental Travel Demand Models To evaluate impacts on: Air quality - No significant change in traffic anticipated; based on travel model, minor diversions could be offset by improved traffic flow on all lanes. Hazardous materials - No anticipated change in patterns of carrier use or changes in risk. Water quality and drainage - Improvements in roadside water filtration systems as part of the project to capture oil and other run-off could result in improved water quality. Cultural and Historical Resources Reference Table 4.1 Public Involvement Plan To evaluate impacts on: Cultural and Historic Resources - No anticipated impact because the signage, toll recognition fixtures, and ramp improvements would be in locations that have been extensively disturbed by previous construction.

402 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox households. The practitioner also aggregates and tracks the findings of the other resource topic areas from the subject matter experts on noise, cultural cohesion, and cultural and historical resources. Most resource topics are dismissed as having no adverse effects, with the findings recorded in technical memoranda to file. Step 6. Identify and Assess Mitigation Strategies The adverse effects and benefits identified from the analytical and outreach activities under- taken in Steps 4 and 5 are closely evaluated. The practitioner studies the “Disproportionate Effects” tool to fully appreciate the criteria that would be employed to assess whether the project has disproportionately high and adverse effects on low-income and minority populations. No significant adverse impacts are identified because of overall improvements in travel times and ability to choose free or managed lanes. Step 6. Identify and assess mitigation strategies. Tools Why Practitioner Uses the Tools Mobility, Access, and Safety Disproportionate Effects Travel Demand Models Select Link Analysis Full Cycle I-10–I-110 LA County Pre- and Post-Surveys Atlanta Public Involvement Plan Focus Groups Determine disproportionately high and adverse effects on: Transportation accessibility - No disproportionately high and adverse impacts. Transportation choice - No disproportionately high and adverse impacts. Transportation safety - No disproportionately high and adverse impacts. Social and Economic Disproportionate Effects VOT/Willing to Pay User Costs, HH Burden Public Involvement Plan Conducting Surveys Full Cycle I-10–I-110 LA County Pre- and Post-Surveys Atlanta Determine disproportionately high and adverse effects on: Community cohesion - No disproportionately high and adverse impacts. Economic development and other economic impacts, including user costs and household burden. - Low-income SOV households would be less likely to use toll lanes during peak hours except in emergencies. Could be burdensome if toll lanes are needed on a regular basis, but non-tolled alternatives, including transit options, car pools, and van pools on the toll lanes, and better operations on the regular lanes (at least in the near term) are available to avoid costs. - Low-income households without transponders who elect to use the toll lanes would pay a higher fee as well as an administrative cost. Noise - No disproportionately high and adverse impacts. - Construction could be limited to daylight hours to avoid night-time noise impact. Longer term noise comparable to current, with potential for sound barriers based on future noise studies. Visual quality - No disproportionately high and adverse impacts. Land use, prices, and property values - No disproportionately high and adverse impacts.

HoV Lanes to HoT Lanes 403 The practitioner uses the Step 6 portion of Checklist 2 in the Checklists section for each research area, the “Disproportionate Effects” tool, and the backup documentation for each resource topic analysis to make clear the level of analyses conducted and document the rational basis for find- ings on each resource topic in Step 4, Step 5, or Step 6. Step 7. Document Results for Decision-Makers and the Public The practitioner uses Checklist 2 in the Checklists section to ensure thorough documentation of various outreach processes and analytical steps (Steps 4, 5, and 6) taken to provide a sound, technical basis for the EJ impact findings for each of the resource topics. Using the “Public Involvement Plan” tool, the practitioner is more aware of potential barriers to public involve- ment for traditionally underserved populations and thoroughly documents public involvement activities throughout the process. Mitigation commitments are clearly identified in support of the EJ impact findings. The documentation distills the analysis from previous steps by resource topic to substantiate the rational basis for preparing the disproportionately high and adverse impact findings and mitigation commitments. Step 8. Conduct Post-Implementation Monitoring The practitioner advises the agency that it would be appropriate to monitor whether the limited travel diversion effects predicted in the travel demand model prove to be true after operations begin. A post-implementation survey and user activity monitoring of the toll account records are also recommended. Monitoring commitments are clearly defined, including milestones for definitive measure- ments; ongoing monitoring mechanisms; what would be measured; how it would be measured; Step 6. (Continued). Mitigation and Commitments: Recycling Toll Revenue Replenishment Options Full Cycle I-10–I-110 LA County Mitigation NCTCOG The toll authority could include a trigger for EJ evaluation in the annual report if the expectation of benefits from congestion relief for all roadway users does not materialize. Tools Why Practitioner Uses the Tools Physical and Environmental Disproportionate Effects Determine disproportionately high and adverse effects on: Air quality - No disproportionately high and adverse impacts. Hazardous materials - No disproportionately high and adverse impacts. Water quality and drainage - No disproportionately high and adverse impacts. Beneficial impacts enjoyed by all. Cultural and Historic Resources Disproportionate Effects No disproportionately high and adverse impacts.

404 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox what constitutes a threshold for taking action; and if action is needed, who would initiate the action and with what public operator consultation, if appropriate. As noted in Step 2, in planning for this HOV to HOT lane conversion, some concerns may be expressed that the primary function of the HOV lanes—to provide an incentive to form carpools and to “reward” such an incentive by guaranteeing a mostly delay-free trip—would be degraded. However, to avoid this, an HOV facility that allows tolled vehicles or any class of qualifying energy-efficient vehicles must annually certify that the subject lanes are “not degraded.” This annual certification states that, for a 180-day continuous reporting period, the lanes operate at greater than 45 miles per hour for 90 percent of the time. Step 7. Document results for decision-makers and the public. Tools Why Practitioner Uses the Tools Checklist 2 Multiple tools from previous steps Public Involvement Plan Mobility, Access, and Safety Transportation accessibility Document no disproportionately high and adverse impacts and commitment for monitoring for potential future impact. The agency would commit to monitor the situation over time. - Document travel times on the toll lanes over time to ensure that historic HOV service is not degraded, consistent with 23 U.S.C. 166 definitions and requirements. - Document travel times on general lanes, based on expressed community preferences. For example, if travel time differences between the HOT and regular lanes exceed an agreed-upon threshold (e.g., 20 minutes on average in peak hours once a week or more) an agreement could be made to take the issue back to the community for input. Transportation choice - Document no impact to minimal impact because buses and vans continue to use the HOV lanes for free, and HOV 3+ vehicles continue to use HOV lanes for free. Document that low-income and minority HOV users were not disproportionately represented among previous HOV 2 users. Transportation safety - Document no anticipated impacts. Social and Economic Community cohesion - Document no impact. Economic development and other economic impacts - Document no impact under current and anticipated traffic conditions on regular use lanes. Noise - Document no disproportionately high and adverse impacts because short- term adverse noise impacts would be mitigated by limiting construction to daylight hours. Visual quality - Not a disproportionate impact. Land use, prices, and property values - Document no disproportionate impact foreseen. Physical and Environmental Air quality - Document no disproportionate impact. Hazardous materials - Document no disproportionate impact. Water quality and drainage - Document a potential beneficial impact. Cultural and Historic Resources Document no disproportionate impacts.

HoV Lanes to HoT Lanes 405 This annual reporting documenting conditions on the converted lanes is required under 23 U.S.C. 166, and if the HOV facility becomes degraded the sponsor must bring the facility into compliance either by increasing HOV occupancy requirements, increasing tolls, increasing capacity, or eliminating access to paying motorists. If the HOV facility becomes degraded, and such actions are taken, it is likely that the regular lanes are already worse and would become even worse. Even though there may not be current EJ effects, because the project is planned and based on assumptions about HOT and general lane performance, it would be prudent to make provisions for including an EJ assessment in a future annual report if actual conditions differ from expected conditions. Several tools describe strategies to ensure post-implementation monitoring. Step 8. Conduct post-implementation monitoring. Tools Why Practitioner Uses the Tools Public Involvement Plan Conducting Surveys Focus Groups User Activity Monitoring Full Cycle I-10–I-110 LA County Pre- and Post-Surveys Atlanta Determine monitoring strategies: Mobility, Access, and Safety Transportation accessibility - Monitor the travel time situation over time. - If travel times on the HOT lanes become degraded, as defined under 23 U.S.C. 166, take actions to restore the HOT lane operations as appropriate. - If travel time differences between the HOT and regular lanes exceed an agreed-upon threshold, recommend taking the issue back to the community for their input. Social and Economic Noise - Confirm that construction agreements and contracts limit construction near concentrations of housing to daylight hours to avoid night-time noise impact. Monitor noise in the future and maintain contact with residents to see if sound barriers would be warranted at some point.

406 Step 1. Frame the Project Scenario C considers a proposed rate change on a tolled facility, an interstate toll road. The interstate toll road is operated by a state toll authority. An alternate untolled highway parallels the toll road. The alternate route operates through towns and urbanized areas with signalized intersections. Many low-income and minority persons currently use the interstate toll road on a regular basis, and many more live in close proximity to an alternate route. The previous toll rate increase instituted by the toll authority generated substantial public controversy about the affordability for certain interstate toll road users, and there were widespread reports that some trucking operators had diverted to local roads to avoid the paying the increased toll. These diver- sions contributed to heavy truck traffic along routes predominantly occupied by minority and low-income populations. The practitioner completes Checklist 1 in the Checklists section and determines that many of the standard considerations and actions do not apply, but that a substantial number do apply. The action is a “pure” rate increase and does not involve a change in toll collection technology, method of payment, user account terms, or access to the toll road, and does not involve any con- struction. In consulting Step 1 of the Guidebook, the practitioner concludes that potential direct effects of the rate increase could include: a change in road use patterns (diversions to alternate routes or modes) for some users; a change in trip-making behavior and trip purposes (trip fre- quency, trip timing) for some users; a change in household disposable income and change in household financial burden; and a change in “disposable time” for some users. In addition, a number of indirect effects related to the direct effects could occur as a result of the rate change on minority and low-income users of the toll road and on the minority and low-income com- munities through which the alternate route passes. Step 2. Identify the Applicable Requirements Governing Decisions The practitioner checks with toll authority management and counsel and learns that a rate change is considered a routine or continuing administrative action and, thus, is not subject to federal environmental review requirements. Specifically, the rate change involves no discretion- ary action by FHWA and, therefore, is not subject to review under NEPA. The toll authority also interprets the state’s environmental review procedures as not requiring environmental review of administrative actions such as rate changes. However, the toll authority recognizes that the rate increase may affect travel patterns and could divert traffic, and have a particularly adverse effect on low-income users of the toll road. Further, the region is in non-attainment for the NAAQS for fine particulate matter (PM2.5). Because diesel trucks are a major contributor to S c e n a r i o c Rate Change

rate change 407 fine-particulates, and because heavy truck diversions to the alternate route were reported with the previous toll increase, the practitioner concludes that a PM2.5 hot spot analysis is warranted. These effects, as well as the increased cost of travel, can generate public controversy as evidenced by what followed the previous toll increase. Because of the controversy surrounding the previous toll increase and in consideration of the state’s social and environmental culture, the toll authority decides to study the effects of the proposed rate change to meet the intent of the state’s environmental review statute and, in so doing, inform the public of the potential effects of the proposed rate change and address poten- tially significant adverse effects. Also, in conducting the study, the toll authority decides to take a proactive approach to engaging with affected users and communities. Step 3. Recognize the Relevant Decision-Makers and Stakeholders The practitioner reads Step 3 of the Guidebook and refers to Reference Tables 3.2 and 3.3 in the Reference Tables section to identify the range of decision-makers and stakeholders that may need to be engaged in the public engagement with minority and low-income populations. In referring to Checklist 2 in the Checklists section, the practitioner is reminded of the impor- tance of having thorough documentation to substantiate impact assessment findings for each relevant resource topic. The topic areas of greatest concern in this situation appear to be trans- portation access, mobility, and safety; community cohesion; user costs and household burden; and air quality. This perception of the major concerns would be confirmed through discussions with the relevant stakeholders and decision-makers and in subsequent outreach to formulate a responsive scope of work for analytical activities (see Step 4). The practitioner refers to the “Public Involvement Plan” tool to inform the development of a public involvement plan that would lay out strategies for ensuring that low-income and minor- ity populations have an opportunity to learn about the proposed toll change and comment on its potential effects in terms of their own travel behavior and how it may affect their community. The strategy would include early scoping meetings and dialogue with community leadership in select communities that have been heavily burdened in the past by truck traffic. The practitioner is also reminded of the importance of documenting the outcomes of these involvement efforts, including records of participation in community meetings (e.g., sign-in lists and estimates of attendance, contact lists) and findings from surveys, focus groups, and other methods for facilitating community feedback. Outreach outcomes include survey findings, con- cerns identified by residents, agreements reached on study areas, study methods, mitigation efforts, monitoring protocols, and related project information. Step 4. Scope Approach to Measure and Address Impacts During Step 4, the practitioner defines working assumptions about study boundaries for the primary and extended impact areas. Mapping would identify clusters of minority and low-income populations residing in the corridor near toll road exits and along the parallel alternate route. In this scenario, initial outreach suggests that various community groups have concerns about the cumulative financial burden effects of the toll rate increase when combined with previous toll increases particularly for low-income populations. Concerns are also expressed about the potential of traffic diverting from the toll road to the alternative local route, causing increased traffic congestion, air pollution and additional pedestrian and vehicle crashes in the vicinity

408 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox of minority neighborhoods that are traversed. Such diversion would also have impacts on the roadway users who divert to the alternate route to avoid tolls, including longer travel distances and longer travel times. Initial research and outreach suggest several concerns including: • Mobility, access, and safety: Would low-income travelers currently using the toll road decide to divert to the non-tolled alternate route, yielding longer travel times to avoid tolls, or could they eliminate some trips? Are these low-income travelers willing to spend their limited dis- cretionary income to pay for higher tolls? Would substantial numbers of people of all income levels, including truckers, divert from the toll road to the alternate route to avoid tolls? What Tools Why Practitioner Uses the Tools Community Characteristics Inventory To identify geographic boundaries for study, map demographic areas with low-income and minority populations. - Area is comprised of an assumed five miles on either side of the toll road and the alternate route, plus the area between the two routes. Travel Demand Models Using NHTS To consider available data from the MPO and the tolling authority that can be used to model existing and prospective travel patterns on the toll road and the alternate route. To gather detailed demographic characteristics of people within the region and their travel patterns. To examine existing toll transponder data to consider the specific toll plaza origin and destination patterns by time of day. “Desire lines” are used to show the weighted frequency of travel between different traffic analysis zones along the corridor. Public Involvement Plan To establish a public involvement plan: - Contact lists already compiled by the toll agency are found too insufficient for reaching potentially affected minority and low-income communities along the alternate route. - Website is developed to describe the proposed rate change, publicize upcoming meetings, and invite public comment. - Convene open house meetings scheduled for day and night-time sessions in the vicinity of major toll plazas and near the alternate routes at accessible public sites. Staff who are trained to assist persons with disabilities and the elderly, provide translation services, and supply childcare are on-hand. The meetings seek to elicit concerns and comments about the toll change and inform the public of technical studies underway and reporting schedule. - These outreach activities also publicize upcoming focus groups and surveys and other upcoming public events. Conducting Surveys Pre- and Post- Surveys Atlanta Focus Groups Surveys with Local Liaison KY-IN To design a survey and sampling plan to explore travel patterns and usage of the toll system and the effects of the proposed toll rate change and demographic information. - Working with several community groups that work with low-income and LEP populations, participants are recruited to attend focus groups. The focus groups explore the clarity of the draft survey questions and explore in-depth particular concerns related to travel choices in response to tolls and interest in various mitigation options. - The survey is distributed to registered transponder users via emails and to cash users via a postcard containing pre-paid mail-back cards and/or online survey links. Using PUMS User Activity Monitoring Step 4. Scope approach to measure and address impacts.

rate change 409 impacts could diverted traffic have on current traffic on the alternate route? Would the increased traffic impact pedestrian, bike and vehicle safety on the alternate route? • Social and economic: Would the increased toll limit access to work, schools, medical facilities, and/or recreation? • Physical and environmental: Would the diverted traffic and increased traffic congestion on the alternate route adversely impact air quality along the alternate route, particularly with regard to Particulate Matter (PM2.5)? The practitioner reviews the “Conducting Surveys” tool and determines that it is appropri- ate to develop a short survey to better understand the probable travel response to a rate change and to explore how it may differ by various demographic groups (including race and income). Reviewing the “Focus Groups” tool, the practitioner determines that it would be productive to explore certain questions and themes with LEP populations and low-income populations to inform the development of the survey, including willingness to pay for tolls, awareness of existing toll rates and policies, and preferences toward various mitigation options, including potential toll credit options. Step 5. Conduct Impact Analysis and Measurement The key questions the practitioner would address in Step 5 are as follows: • Would the proposed rate change have an impact on low-income and minority populations? • Would impacts extend beyond the categories of mobility, access, and safety; social and eco- nomic; physical and environmental; and cultural and historic resources? • Would other resource topics be dismissed as having no foreseeable impact? The practitioner reviews the “Travel Demand Models” and “Select Link Analysis” tools and dis- cusses with the MPO and toll authority travel demand modelers whether existing regional travel demand models are suitable for addressing the key questions, in particular, for differentiating Tools Why Practitioner Uses the Tools Mobility, Access, and Safety Public Involvement Plan Conducting Surveys Pre- and Post-Surveys Atlanta Focus Groups Travel Demand Models Select Link Analysis VOT/Willing to Pay Select Link Analysis Evaluate impact on: Transportation accessibility - Analysis, surveys and outreach indicate that some truck trips and some travel for low-income households could be diverted from the toll road to the alternate route due to higher cost. - The alternate route could be considerably slower than the toll route, both because of traffic lights, speed limits and geometry, and because of additional traffic congestion due to diversion from the toll road. Transportation choice - Only infrequently scheduled public bus service or van pool service is available along the toll route. - Transit service along the alternate route could be slowed by the traffic diversion. Some individuals carpool but there is currently no toll discount provided for such behavior. - Diverted traffic to the alternate route could negatively impact pedestrian, bicycle and vehicle safety. Step 5. Conduct impact analysis and measurement. (continued on next page)

410 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox Tools Why Practitioner Uses the Tools Social and Economic VOT/Willing to Pay User Costs, HH Burden Public Involvement Plan Evaluate Impact on: Community cohesion - Community outreach did not identify adverse impacts to community cohesion. Economic development and other economic impacts, including household burden - Low-income households could be adversely impacted. The toll increase could impose a significant household burden on low- income/minimum wage workers who use the toll road on a daily basis for work, school or services; therefore, it is assumed that a segment of travelers would divert to the alternate route, which could cost significant time. For those working multiple jobs or with daycare or other constraints, such time might be a burden. Noise - Diverted traffic might cause increased noise, but the impact is assumed to be unnoticeable. Visual quality - No adverse impacts anticipated. Land use, prices and property values - No change anticipated. Physical and Environmental Travel Demand Models (Air quality and noise are not specifically addressed in the Travel Demand Models tool, but underlying travel patterns and analysis support such efforts.) Evaluate Impact on: Air quality - MPO Travel Demand and Air Quality modelers investigate potential changes in regional air quality due to the diversion to the alternate route and resulting greater congestion. The practitioner requests that modelers perform MSAT “hot spot” analyses along the alternate route, including in areas with concentrations of low-income and minority populations. Hazardous materials - Conversations with major hazardous waste carriers determine that the majority of carriers value time and safety over the tolls, and therefore would be unlikely to divert to the alternate route for through trips. Water quality and drainage - No adverse impacts. Cultural and Historic Resources No adverse impacts. Step 5. (Continued). travel by income. The region’s MPO is converting from a traditional 4-step model to an ABM and there may be an opportunity to apply the tool for this project. The “Conducting Surveys” and “VOT/Willing to Pay” tools would be particularly impor- tant when considering the trade-offs that individuals at various income levels must make when deciding whether to pay a higher toll or make a trip on an alternate route or via a different mode that may require much more travel time. Discerning trip patterns and VOT for different income levels may not be a customary activity of the MPO travel demand modelers, and therefore, it is important for the practitioner to coordinate and streamline requests to the regional modelers and ensure consistent model assumptions and results. Air quality and noise are not specifically

rate change 411 addressed in the tool, but underlying travel patterns and analysis would produce travel informa- tion that would be input to the air quality and noise models. Step 6. Identify and Assess Mitigation Strategies The key questions the practitioner would address in Step 6 are as follows: • Could the proposed rate change contribute to a financial burden effect that is disproportion- ately high and adverse for low-income populations? • Alternatively, could the diversion to the alternate route result in disproportionately high and adverse effects on low-income users in terms of travel time? • And, could the diversion to the alternate route create a disproportionate and adverse effect on low-income and minority populations residing in communities adjacent to the alternate route? Tools Why Practitioner Uses the Tools Disproportionate Effects User Costs, HH Burden Determine disproportionately high and adverse effects on: Mobility, Access, and Safety Analysis found: A disproportionately high and adverse financial burden impact would likely be borne by low-income populations using the corridor, taking into consideration the cumulative impacts of toll increases. Daily tolls for low-income workers would consume an appreciably greater share of their household income than for higher- income individuals. For households working multiple jobs or penalized by late arrivals, the slow travel times on the alternate route would not be a reliable alternative. A disproportionately high and adverse “time burden” impact would likely be borne by lower income populations that divert to the slower non-tolled alternative route in comparison to higher-income population segments that would not divert. A “hot spot” analysis demonstrates that low-income and minority populations living in close proximity to the alternate route would be more likely to be exposed to disproportionately higher levels of particulate matter resulting, in part, from diverted traffic. Social and Economic Community cohesion - No disproportionately high and adverse effect. Economic development and other economic impacts, including user costs and household burden - Mitigation alternatives would be studied to address the potential EJ impact. Physical and Environmental The toll authority could agree to test deploying highway advisory signs on comparable travel times (on the toll route versus the alternate route) to reduce diversion to the alternate route, especially for through travelers. Cultural and Historical Resources No disproportionately high and adverse effects. Step 6. Identify and assess mitigation strategies. (continued on next page)

412 assessing the environmental Justice effects of Toll implementation or rate changes: Guidebook and Toolbox If any or all of these questions could be answered “yes,” for any of the resource topic areas examined, are there mitigation measures or offsetting benefits that could be considered to address the disproportionately high and adverse impacts? The practitioner uses the “Focus Groups” tool to help structure in-depth conversations about concerns, trade-offs, and desirable mitigations. Through outreach efforts, the practi- tioner learns that low-income households have substantial concerns about the added cost for travel, especially commuting, if using the toll road, but are also concerned about travel to access other opportunities and services (e.g., education, medical care, etc.). The alternate route would be much slower, and is perceived to be significantly slower with extensive traffic conflicts (e.g., left hand turns, plus through traffic). The practitioner seeks to further vet and develop the ideas discussed in focus groups through design of the survey and through other discussions with the agency and other stakeholders to assess feasibility. Six potential mitigation measures are generated, in part, from focus group discussions for toll authority review as most promising and potentially helpful. The practitioner uses the “Disproportionate Effects” tool to analyze the data collected in Step 5. The analysis reveals adverse effects that appear to be appreciably more severe and greater in magnitude for low-income households. In response, several mitigation strategies are further considered by the tolling authority and commitments are made to conduct further studies and pilot one or more of the most feasible options. The toll authority determines that it would be appropriate to work with the region’s public transportation agency to explore the feasibility of increasing the frequency of transit service using the toll road as well as work with the state DOT to identify other potential ridesharing options. The toll authority also agrees to consider a pilot study of implementing a lower-cost transponder tied to HOV use for committed existing car pools. The agency also agrees to further study the potential revenue and cost impacts of instituting a low-income assistance plan specifically targeted to income-qualified low-income households to discount tolls for frequent users. Tools Why Practitioner Uses the Tools Mitigation Strategies and Commitments Recycling Toll Revenue Replenishment Options Full Cycle I-10–I-110 LA County Mitigation NCTCOG Several mitigation strategies could be considered and commitments made by the tolling authority to study and pilot one or more of the feasible options: The toll authority could exempt HOVs (3+ individuals) from tolls. The mitigation would incentivize desired behavior, reduce peak loads, reduce diversion to the alternate route, and reduce funding to the tolling authority. The toll authority could issue special transponders for HOV users; trips could be free as long as there are three or more in the car. The transponder could have a switch and could revert to a full or reduced fare if not HOV. The toll authority could agree to study the feasibility of issuing transponders to income-qualified low-income households to discount the toll rate based on travel. The toll authority could agree to work with the transit agency to investigate demand for public transit and ridesharing services (car pools, van pools) along the toll road. The city and county could prohibit through trucks from using the alternate route, posting signs and monitoring violations, to lessen the traffic impact. Step 6. (Continued).

rate change 413 Step 7. Document Results for Decision-Makers and the Public The practitioner uses Checklist 2 in the Checklists section to ensure thorough documen- tation of various outreach processes and analytical steps (Steps 4, 5, and 6) taken to provide a sound, technical basis for the EJ impact findings for each of the resource topics. Using the “Public Involvement Plan” tool, the practitioner would have thoroughly documented public involvement activities throughout the process and documented how those processes would have informed the impact assessment and the mitigation commitments. The documentation would distill the analysis from previous steps by resource topic to sub- stantiate the rational basis for impact assessment, the disproportionately high and adverse effect findings and mitigation commitments. Step 8. Conduct Post-Implementation Monitoring The tolling authority could agree to partially fund one or more of the mitigation options in partnership with the transit agency and other partners and report back within 18 months as to its progress in pilot testing one or more of the mitigation options. The practitioner advises the tolling authority that it is appropriate to monitor whether the travel diversion effects predicted in the travel demand model and expressed in surveys prove to be true following the toll increase. A post-implementation survey and user activity monitoring of the toll account records would also be recommended. Monitoring commitments would be clearly defined, including milestones for definitive mea- surements; ongoing monitoring mechanisms; what would be measured; how it would be mea- sured; what constitutes a threshold for taking action; and if action is needed, who would initiate the action.

Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACI–NA Airports Council International–North America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FAST Fixing America’s Surface Transportation Act (2015) FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TDC Transit Development Corporation TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S.DOT United States Department of Transportation

TRA N SPO RTATIO N RESEA RCH BO A RD 500 Fifth Street, N W W ashington, D C 20001 A D D RESS SERV ICE REQ U ESTED N O N -PR O FIT O R G . U .S. PO STA G E PA ID C O LU M B IA , M D PER M IT N O . 88 ISBN 978-0-309-44673-0 9 7 8 0 3 0 9 4 4 6 7 3 0 9 0 0 0 0 A ssessing the Environm ental Justice Effects of Toll Im plem entation or Rate Changes: G uidebook and Toolbox N CH RP Research Report 860 TRB

Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Research Report 860: Assessing the Environmental Justice Effects of Toll Implementation or Rate Changes: Guidebook and Toolbox provides a set of tools to enable analysis and measurement of the impacts of toll pricing, toll payment, toll collection technology, and other aspects of toll implementation and rate changes on low-income and minority populations.

The guidebook shows the practitioner when and how to apply the tools in the toolbox through an eight-step process framework corresponding to the typical transportation project planning and development process. The guidebook and toolbox together provide an assessment framework and supporting tools to measure the impacts of tolling on such factors as mobility, access, and household expenditures, as well as tools to engage low-income and minority populations.

This report is accompanied by NCHRP Web-Only Document 237: Environmental Justice Analyses When Considering Toll Implementation or Rate Changes—Final Report. This report presents information gathered in the development of the guidebook and the toolbox. This web-only document summarizes the technical research and presents the technical memorandum that documents the literature, existing case studies, resource documents, and other reports compiled.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!