Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand (2014)

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.

38 Impacts 2050 is a menu-driven spreadsheet model that state and regional transportation decision makers can use to play out the many ways that changing socio-demographic factors in a region may impact travel demand over time. The tool is designed to be a strategic model: Strategic models are an emerging trend in long-range planning, where there is an awareness that one cannot actually forecast the future, but that many scenario possibilities need to be studied so that a policy or investment strategy that minimizes risk, or moves towards some desired goal(s), can be followed. Like other types of strategic models, Impacts 2050 has been designed to produce qualitatively accurate representations of how different vari- able relationships will evolve over time, rather than numerically precise forecasts for one particular sector. Being qualitatively accurate means the relationships are in the right direction and make intuitive sense. 6.1 Overview of the SD Model Impacts 2050 is powered by an SD model that simulates a regional population over time starting from a base of the 2000 Census and span- ning a period of 50 years. The model depicts five sectors: (1) socio- demographics, (2) travel behavior, (e) land use, (4) employment, and (5) transportation supply. Model results include the travel demand effects of the changing population, which were modified by feedback from the employment, land use, and transportation supply sectors. For example, population increase could increase congestion, which if not alleviated could lead to some people relocating within the region over the longer term, and eventually to a change in the location of employment and/or mode choice. Like other SD modeling tools, Impacts 2050 enables exploratory modeling of changes in these sectors due to socio-demographic changes, the interplay among them, and external factors that are intertwined with socio-demographics, such as attitudes and technology. The latter was accom- plished through predefining our four scenarios in the tool. These represent “what if” conditions that moderate the outcome of the business-as-usual scenario, Momentum, and the tool is set up to enable a user to modify the scenario inputs in order to test many different hypotheses about the future (i.e., different scenarios)—not just the four scenarios developed by the research team. A distinguishing feature of our SD model is an emphasis on dynamics that can result from the relationships pertaining to travel demand that will likely change over time, requiring changes in these relationships over time. Figure 6-1 depicts that there can be substantial time delays in C H A P T E R 6 Scenario Planning Tool: Impacts 2050 Chapter 6 Takeaways • Impacts 2050 is a strategic modeling tool. • Four scenarios are predefined in the tool, but many other scenarios can be tested. • The underlying SD model contains 700 variables in five sectors: socio- demographics, travel behavior, land use, employment, and transport supply. • Model simulation starts at 2000 and runs through 2050 in half-yearly increments. “Everything must be made as simple as possible. But not simpler.” Albert Einstein, theoretical physicist

Scenario Planning Tool: Impacts 2050 39 the system, such as those related to supplying new transportation infrastructure or new housing and commercial infrastructure. Even decisions to change residence or business locations can take some time to occur, so one cannot adjust immediately to changes in prices, congestion, job availability, etc. The SD methodology is specifically designed to reflect these types of dynamic phenomena. Figure 6-1 shows the main feedback relationships between the sectors. Underneath this conceptual representation are many endogenous and exogenous variables— the model and scenario variables, respectively. • Model variables affect and are affected by the rest of the system. There are two types of these variables: (1) those that define the current state or base conditions, and (2) those that define the transitions or rates of change in its state. • Scenario variables are outside variables that affect but are not affected by the behavior of the system. In our SD model, these are the variables that distinguish the four scenarios. Figure 6-2 presents the model variables that comprise Impacts 2050. The most detailed sector is the socio-demographic sector, which will evolve the population over time. The other three sectors—transport supply, land use, and employment—are modeled in a more aggregate manner. These sectors are not the primary focus of the model, but it is important that they be represented, as they have a crucial influence on the evolution of the population and travel within a region. Detailed documentation for each of the sectors can be found in Appendix B. The overall design of our SD model was informed by other strategic land use/transportation models—specifically the DELTA, GreenSTEP, UrbanSIM, and Dynamic Urban Model. • DELTA (Development, Transition, Location, Employment, and Area-quality), developed by David Simmonds Consultancy, is designed to operate in iteration with a local transport Figure 6-1. Overall relationship between the sd model sectors.

40 The Effects of Socio-Demographics on Future Travel Demand model. DELTA operates in one-year steps, with interaction with the local transport model typically occurring every two or five years, depending on the run times of the transport model. • The GreenSTEP model (Greenhouse gas Statewide Emissions Planning), developed by Brian Gregor, forecasts greenhouse gas (GHG) emissions from the transport sector for the state of Oregon DOT. • UrbanSIM is a simulation model for integrated planning and analysis of urban development. It was developed by Paul Waddell at the University of Washington (now at UC Berkeley), and is available as a public domain software package intended for MPO use. • The Dynamic Urban Model (DUM) was developed by John Swanson of Steer Davies Gleave (SDG) to simulate the interaction between transport, land use, population, and economic activity in an urban area. The heart of the SD model is the socio-demographic sector. To predict changes in this sector over time, Impacts 2050 first profiles the base-year population across a range of attributes that are associated with travel behavior. It then evolves this population over time, simulating transi- tions from one category in each of these attributes to another category. Generally, the SD model segments a region’s population by age, household structure, income, race/ethnicity, accultura- tion, residence location, area type, and work status. The model then “evolves” this population Figure 6-2. Impacts 2050 model structure.

Scenario Planning Tool: Impacts 2050 41 over time, simulating the population’s transitions from one category in each of these segments to another category over time (e.g., aging the population into different categories: 0–15, 16–29, 30–44, 45–59, 60–74, 75 and older). The evolution of the population over time affects travel behavior, which is indicated in terms of car ownership, trip rates and distance, and mode choice. The presence of other sectors enables the incorporation of feedback loops that represent the dynamics of the transportation system. The model contains more than 700 variables represent- ing five sectors (socio-demographics, travel behavior, land use, employment, and transport sup- ply) that are linked by mathematical formulas. The model simulation starts with base values for the year 2000 and runs through 2050. The impacts on travel behavior are calculated in terms of car ownership, trip rates, and choice of transportation mode. Changes in expected transitions can be tested as scenario variables. This process is illustrated in Figure 6-3. 6.2 Scenarios in Impacts 2050 The four scenarios are integrated into Impacts 2050 through their underlying assumptions. The scenario assumptions are the attitudes, policies, or other phenomena that were used to develop and differentiate the scenarios. A problem is that the language of the scenarios—which is very broad and qualitative—does not translate directly into the language of the model. To help with this process, we introduced a separate conceptual layer called “scenario variables,” which is illustrated in Figure 6.4. The scenario variables represent classes of assumptions, such as “attitudes toward having children” or “environmental conditions,” that affect clusters of model variables. Each scenario will affect the scenario variables differently; for example, under Gentle Footprint, people will choose on average to have fewer children. The scenario variables can then be translated directly into model variables, such as fertility rates. Scenario variables can affect more than one model variable, and particular model variables may be affected by a number of different scenario vari- ables. Note that the scenario variables are not a functional part of the model; their role is to help calibrate the model for different scenarios, design new scenarios, and communicate the connec- tion between the scenarios and the model. Impacts 2050 contains default programs that run each of the scenarios. The assumptions are clearly indicated for each scenario. When the model is run for given scenarios, the results highlight their distinct futures. There are key differences in population structure, workforce Data on age, household structure, income, ethnicity, acculturation, residence location type, Workforce participation Population in Year B Baseline travel behavior • Car ownership • Trip rates • Choice of transportation mode Changes in socio- demographics Momentum Scenario Changes in travel behavior • Car ownership • Trip rates • Choice of transportation mode Other Scenarios: Changes in assumptions Changes in travel behavior Population in Year A (Base Year) Figure 6-3. Evolving the population over time in travel: Impacts 2050.

42 The Effects of Socio-Demographics on Future Travel Demand participation, immigration, and travel behavior across the scenarios. For example, auto VMT per capita in 2050 is much lower under the Global Chaos and Gentle Footprint scenarios than in the other two scenarios. This is due to differing assumptions. In Gentle Footprint, the lower VMT per capita is by choice, whereas in Global Chaos it is due to poor economic conditions and lack of opportunities. With the same reasoning, car-sharing and walk/bike modes are much higher under the Global Chaos and Gentle Footprint scenarios than in the Momentum and Technology Triumphs scenarios. An innovative feature of Impacts 2050 is that users can modify the default structures to create alternatives to these four scenarios. The assumptions can be changed, and new scenarios can be run, by modifying the parameters indicated for model variables. 6.3 Regional Settings for Model Testing A goal of this project was to develop a tool that state and local transportation agencies can use to understand how socio-demographics will impact travel behavior in the long-term future, and to examine how the illustrative scenarios developed in this project could play out in their regional jurisdictions. So, it was important to identify a set of metropolitan areas that could serve as test sites during tool construction. The research team did not focus on metro areas that were specifically illustrative of the four scenarios, since the objective was to apply and test the scenarios in the different regional settings. The following relevant characteristics were important for differentiating the regional settings: • Population growth trend: population change over last 50 years, net growth rate. • Nature of growth: domestic versus international migration. • Spatial distribution: land area, density. • Economic base: socioeconomic status, income disparity, unemployment rate. • Diversity: household structures, age, racial/ethnic composition. Scenarios Momentum Technology Triumphs Global Chaos Gentle Footprint Attitudes Toward Having Children Healthcare Technology/ Availability Environmental Conditions Scenario Assumptions Exogenous Effect on Fertility Rate Exogenous Effect on Mortality Rate Exogenous Effect on Gasoline Price Model Variables Figure 6-4. Using the technology triumph scenario as an example, conceptual representation of interaction among scenarios, assumptions, and model variables.

Scenario Planning Tool: Impacts 2050 43 • Transportation system orientation: highway versus transit supply, congestion levels, com- mute mode share. • Location: major region of the country, as this tends to reflect era of development as well as various socio-demographic traits, such as age, household type, educational attainment, wealth, and housing. In addition, the team thought it was important to distinguish among the five U.S. Census regions in defining the base sampling frame: Northeast, Southeast, Midwest, Southwest, and Pacific West. Each of these areas has a distinctly different character that would be reflected in the attributes listed above (see Table 6-1). With these criteria in mind, the following metro areas were selected: • Southeast: Atlanta, GA • Northeast: Boston, MA • Midwest: Detroit, MI • Southwest: Houston, TX • Pacific West: Puget Sound (or Seattle, WA, as short-hand) Impacts 2050 has embedded data for these five metropolitan regions. The simulation model used to compute the scenario indicators uses a custom database for each region. Four sets of data must be specified to define the year 2000 base conditions for the simulated region: demographic, land use, employment, and transportation supply. Data have already been entered into Impacts 2050 for the five regions. For any other region, these data must be input (as is fully explained in the User Guide). For this reason, the team has ensured that the input data will be publicly available from the U.S. Bureau of the Census and the Federal Transit Administration. Characteristics Boston Atlanta Detroit Houston Seattle Population growth since 1960 77% 443% 16% 382% 219% Net growth rate 5.03% 9.69% 3.5% 13.5% 8.16% Population growth in center city –7% –13% –57% 125% 11% Population in center city 13% 8% 17% 35% 18% Metropolitan statistical area land area 3,506 8,377 3,914 8,928 5,894 Population density 1,298 654 1,099 666 583 Average household size 2.50 2.68 2.53 2.83 2.49 Households with children 31% 38% 33% 41% 32% Home ownership 62% 66% 71% 63% 62% College degree 42% 34% 27% 28% 37% White 75% 51% 68% 40% 68% Hispanic 9% 10% 4% 35% 9% Median age 38.5 34.9 39.1 33.2 36.8 Unemployment 7% 9% 13% 7% 7% Median income $69.9k $57.5k $52.4k $55.2k $65.4k Population in poverty 19% 20% 24% 21% 17% Transit commuting 12% 3% 2% 2% 8% Auto commuting 69% 77% 84% 79% 70% Roadway congestion index 1.09 1.27 1.14 1.15 1.08 Daily VMT per capita 23.4 27.9 25.6 33.3 22.1 Table 6-1. Summary of the characteristics of selected metro regions.

44 The Effects of Socio-Demographics on Future Travel Demand 6.4 Running Impacts 2050 for the Test Regions A set of quantitative indicators is built into Impacts 2050 (See Table 5-1 for the scenario framework, including the key indicators used in the scenario development.) The indicators are the results of the model simulations related to the scenarios. They were developed to provide an overview of “what might happen in the future” (i.e., travel impacts) and “why it might be hap- pening” (i.e., socio-demographic trends). The Impacts 2050 output is presented in both table and graphic form. Figure 6-5 illustrates graphic output directly from Impacts 2050. Tables 6-2 through 6-5 compare the results of 2050 projections for the five regions with 2010 Census estimates for a subset of these indicators. When compared with 2010, 2050 looks differ- ent from today for most regions under the various scenarios. These results support our scenario approach, which was to develop coherent pictures of the future that move off in different directions. The Momentum scenario is our business-as-usual case. It simulates today’s population in a region through time to 2050, considering relationships among model variables and feedback Figure 6-5. Outputs from Impacts 2050, population by age group (Seattle momentum scenario).

Scenario Planning Tool: Impacts 2050 45 Indicators 2010 Statistics Atlanta 2050 Projection Momentum Tech Triumphs Global Chaos Gentle Footprint Auto VMT per capita 11,115 10,251 11,461 5,451 4,167 Percent noncar owning 2.5% 3.0% 2.5% 5.2% 4.1% Percent car-sharing 22% 22% 17% 34% 29% Average car occupancy 1.6 1.6 1.5 1.8 1.8 Transit mode share 2% 2% 2% 2% 3% Walk/bike mode share 11% 11% 10% 19% 22% Work trips per capita 0.5 0.5 0.5 0.4 0.3 Nonwork trips per capita 2.9 3.0 2.9 1.7 1.7 Population 5,262,023 8,225,550 7,205,888 5,694,525 7,910,911 Percent under 16 22% 23% 20% 17% 15% Percent over age 60 14% 19% 23% 19% 27% Percent over age 75 4% 6% 9% 4% 9% Percent Hispanic 8% 12% 11% 11% 13% Percent low income 32% 33% 28% 51% 36% Percent high income 19% 27% 32% 17% 26% Percent foreign-born 16% 13% 11% 11% 24% Percent in workforce 47% 39% 46% 43% 48% Table 6-2. 2010 statistics and 2050 projections in atlanta by scenario. Indicators 2010 Statistics Boston 2050 Projection Momentum Tech Triumphs Global Chaos Gentle Footprint Auto VMT per capita 9,874 8,709 9,741 4,445 3,293 Percent noncar owning 4.6% 5.0% 4.3% 8.2% 7.2% Percent car-sharing 24% 23% 19% 36% 30% Average car occupancy 1.6 1.7 1.6 1.9 1.8 Transit mode share 3% 3% 3% 4% 6% Walk/bike mode share 17% 18% 17% 28% 31% Work trips per capita 0.5 0.4 0.4 0.3 0.3 Nonwork trips per capita 2.9 3.1 3.0 1.8 1.8 Population 4,662,662 6,149,585 5,356,991 4,242,692 6,227,814 Percent under 16 18% 22% 19% 16% 14% Percent over age 60 18% 21% 25% 21% 30% Percent over age 75 6% 7% 10% 5% 11% Percent Hispanic 8% 10% 10% 10% 12% Percent low income 28% 31% 27% 48% 35% Percent high income 23% 28% 34% 18% 27% Percent foreign-born 19% 17% 15% 15% 31% Percent in workforce 51% 39% 46% 43% 48% Table 6-3. 2010 statistics and 2050 projections in boston by scenario.

46 The Effects of Socio-Demographics on Future Travel Demand Indicators 2010 Statistics Detroit 2050 Projection Momentum Tech Triumphs Global Chaos Gentle Footprint Auto VMT per capita 10,126 9,580 10,586 4,886 3,775 Percent noncar owning 3.0% 3.3% 2.8% 5.4% 4.5% Percent car-sharing 31% 30% 25% 45% 38% Average car occupancy 1.6 1.6 1.6 1.9 1.8 Transit mode share 1% 1% 1% 2% 3% Walk/bike mode share 11% 12% 11% 20% 23% Work trips per capita 0.4 0.4 0.4 0.3 0.3 Nonwork trips per capita 3.6 3.7 3.6 2.2 2.2 Population 4,372,010 5,245,748 4,663,877 3,764,011 5,251,724 Percent under 16 20% 22% 19% 17% 15% Percent over age 60 18% 20% 25% 20% 28% Percent over age 75 6% 7% 10% 5% 10% Percent Hispanic 3% 5% 4% 4% 5% Percent low income 26% 31% 27% 49% 35% Percent high income 20% 28% 34% 18% 27% Percent foreign-born 11% 12% 10% 10% 24% Percent in workforce 44% 39% 46% 42% 47% Table 6-4. 2010 statistics and 2050 projections in detroit by scenario. Indicators 2010 Statistics Houston 2050 Projection Momentum Tech Triumphs Global Chaos Gentle Footprint Auto VMT per capita 9,560 9,072 10,171 4,859 3,720 Percent noncar owning 3.7% 4.2% 3.6% 6.5% 5.5% Percent car-sharing 30% 30% 25% 43% 37% Average car occupancy 1.7 1.7 1.6 1.9 1.8 Transit mode share 2% 2% 2% 2% 3% Walk/bike mode share 11% 12% 11% 20% 22% Work trips per capita 0.5 0.5 0.5 0.4 0.3 Nonwork trips per capita 2.7 2.8 2.7 1.6 1.6 Population 5,944,540 9,291,817 8,092,777 6,380,152 8,951,738 Percent under 16 23% 23% 21% 18% 16% Percent over age 60 13% 19% 23% 19% 26% Percent over age 75 4% 6% 9% 4% 9% Percent Hispanic 34% 42% 40% 40% 45% Percent low income 38% 35% 31% 53% 39% Percent high income 15% 25% 31% 16% 23% Percent foreign-born 24% 17% 15% 15% 28% Percent in workforce 46% 40% 47% 43% 48% Table 6-5. 2010 statistics and 2050 projections in houston by scenario.

Scenario Planning Tool: Impacts 2050 47 loops for the different sectors. Looking at the universal changes across regions for this scenario, Impacts 2050 results seem to indicate auto VMT per capita will drop, walk/bike mode share will increase, and people over age 60 will comprise a much larger share of the regional population. But unique changes are indicated in the five regions. Under the Momentum scenario, Atlanta in 2050 will have about 3 million more residents (see Table 6-2). Compared with today, a greater portion of these residents will be over the age of 60 and Hispanic, and will have a higher income. There will be fewer workers. In this socio- demographic context, Atlanta will be experiencing a slight change in auto VMT, but transit and walk/bike shares will be about the same as today. The percentage of people without a vehicle will be the same as today—about 3 percent. Travel behavior in Atlanta looks the most different from today under the Global Chaos and Gentle Footprint scenarios. Under these two scenarios, auto VMT decreases substantially (by more than 50 percent). Nearly one-third of residents will carpool or car-share. The walk/bike mode share will increase from 11 percent to between 19 and 22 percent. The socio-demographic profile of Atlanta changes the most compared with today under the Gentle Footprint scenario. Nearly one of three residents will be over the age of 60, and there will be a much smaller percentage of children. The share of Hispanics will increase from 8 to 13 percent, and the percentage of foreign-born residents will increase from 16 to 24 percent. But there will be about the same number of workers in the region as today. In 2050 under the Momentum scenario, Boston will have about 1.5 million more residents (see Table 6-3). Demographically, Boston will look a lot like today, except for the fact that a greater portion of residents will be under the age of 16 and there will be substantially fewer workers. In this socio-demographic context, Boston will be experiencing lower auto VMT, but in most other statistics, travel behaviors will not change much. As in Atlanta, the Global Chaos and Gentle Footprint scenarios appear to alter current travel behavior patterns most when compared with those of today. Under these two scenarios, VMT decreases substantially and car-sharing and nonmotorized travel increase by about 10 percent. The socio-demographic profile of Boston alters the most under the Gentle Footprint scenario in much the same way as Atlanta—people live longer, so there are more people over age 60 and age 75. Green businesses and increased farming activity increase the immigrant population. Detroit, under the Momentum scenario, will have about 1 million more residents in 2050 (see Table 6-4). But the demographic makeup of Detroit will be much like that of today. The significant change will be that income inequality will grow with both more low-income and more high-income residents. In terms of travel behavior, Detroit will retain its current auto orientation: high auto VMT per capita, high carpooling or car-sharing, and low transit and walk/bike shares. In the future under the Momentum scenario, auto VMT per capita will decrease, but by a small percentage. Carpooling or car-sharing will not decrease much, and the share of nonmotorized modes will not increase. In terms of the impact of the other scenarios on travel behavior, Global Chaos and Gentle Footprint will cut auto VMT per capita significantly, while under Tech Triumphs it will increase slightly. Car-sharing will decrease slightly under Tech Triumphs, but will increase sub- stantially under Global Chaos and Gentle Footprint. Walking and biking also will increase under the latter two scenarios, but will stay about the same under Tech Triumphs. The percentage of children will decrease under all three alternative scenarios, but the decrease will be greatest under Gentle Footprint. The percentage of people over age 60 will increase sub- stantially under both Gentle Footprint and Tech Triumphs. Income inequality will be lessened under the Tech Triumphs scenario and exacerbated under Global Chaos and also to some degree under Gentle Footprint.

48 The Effects of Socio-Demographics on Future Travel Demand Houston, under the Momentum scenario, will experience a population increase of slightly more than 3 million people in 2050—about the same projected increase as for Atlanta (see Table 6-5). Compared with today, the population distribution will be older (nearly one in five people will be over age 60). The portion of people in Houston who are older than 60 will increase substan- tially under Tech Triumphs and will increase even more under Gentle Footprint. There is a large variation in the percentage of the population that is in the workforce under the different scenarios. Workers will decrease significantly under the Momentum scenario and less under the Global Chaos scenario, and will increase under the Tech Triumphs and Gentle Footprint scenarios. Today, large percentages of Houston’s population are Hispanic (34 percent) and immigrant (24 percent). Under the Momentum scenario, the percentage of Hispanics will increase to 42 percent, but the percentage of immigrants will decrease to 17 percent. Under the Tech Triumphs and Global Chaos scenarios, the growth in Hispanics will be lower than under the Momentum scenario and will be even higher under the Gentle Footprint scenario. In terms of the immigrant population, it will increase to 28 percent under the Gentle Footprint scenario; however, under the other two scenarios, the percentage of foreign-born residents will decrease to 15 percent. Under the Momentum scenario, travel behavior will not change much in Houston by 2050, with only a slight projected decrease in auto VMT per capita. Under Tech Triumphs, auto VMT per capita will slightly increase and will significantly decrease under the Global Chaos and Gentle Footprint scenarios. Car-sharing and walk/bike mode shares will increase significantly under the latter two scenarios. Under Tech Triumphs, car-sharing will actually decrease relative to today and to the Momentum scenario. Seattle, under the Momentum scenario, will experience a population increase of slightly less than 2 million persons in 2050 (see Table 6-6). Compared to today, the population distribution Indicators 2010 Statistics Seattle 2050 Projection Momentum Tech Triumphs Global Chaos Gentle Footprint Auto VMT per capita 9,916 8,728 9,822 4,528 3,351 Percent noncar owning 5.6% 6.4% 5.4% 10.9% 8.9% Percent car-sharing 18% 17% 14% 28% 23% Average car occupancy 1.5 1.6 1.5 1.8 1.7 Transit mode share 4% 4% 4% 5% 7% Walk/bike mode share 18% 20% 18% 29% 32% Work trips / capita 0.6 0.4 0.5 0.4 0.3 Nonwork trips /capita 2.6 2.8 2.7 1.6 1.6 Population 3,522,980 5,365,107 4,632,781 3,656,502 5,299,978 Percent under 16 18% 22% 19% 17% 14% Percent over age 60 16% 21% 25% 20% 29% Percent over age 75 5% 7% 10% 5% 10% Percent Hispanic 7% 9% 8% 8% 10% Percent low income 28% 32% 27% 49% 36% Percent high income 17% 27% 33% 17% 26% Percent foreign-born 18% 16% 14% 14% 28% Percent in workforce 49% 39% 46% 43% 48% Table 6-6. 2010 statistics and 2050 projections in Seattle by scenario.

Scenario Planning Tool: Impacts 2050 49 will be both older and younger. The percentage of workers will decrease by almost 10 percentage points. As in the other regions, there will be a decrease in auto VMT per capita. In terms of the other scenarios, under Tech Triumphs auto VMT per capita is about the same as today as will be most other travel behavior patterns. The percentage of older per- sons increases substantially, that of children not so much. Income disparities among the regional population will decline. There will be fewer immigrants than today or compared to Momentum. The percent of the population in the workforce will be about the same as today. Global Chaos and Gentle Footprint have the same impacts as in other regions— dramatically decreasing auto VMT per capita. In Seattle, however, these two scenarios also have an influence on the percent of noncar owning households, which will increase signifi- cantly. Under Global Chaos, the poor economy leads to a significant drop in workers and a significant increase in the percent of low-income household. Under Gentle Footprint, the percent of the population in the workforce remains about the same as today but income inequality increases substantially. Table 6-7 presents the comparison of the 2050 outcomes with 2010. We summarize this result as percentage or percentage point changes accordingly. The magnitude and direction of changes are indicated by arrows pointing up (increasing), down (decreasing), and sideways (no or small change). A data table presenting the statistical back-up for Table 6-7 is found in Appendix C. An analyst can examine this information and attempt to understand the underlying influences that likely affect the movement in one direction versus another. Hypotheses can be tested by running with new assumptions in mind. For example, auto VMT per capita decreases over time in 13 out of the 20 scenarios tested. Tech Triumphs deviates from this main trend in all regions. A potential reason for this deviation is increased overall economic growth as well as employment growth. Under the Momentum, Global Chaos, and Gentle Footprint scenarios, the auto VMT per capita decreases probably because of both the aging of the population (i.e., percentage of seniors goes up over time) and lower workforce participation (i.e., percentage of people who are employed goes down). These demo- graphic changes happen in the Tech Triumphs scenario as well; however, this scenario indicates an increase in income, which leads to high car ownership and less car-sharing. The decrease in auto VMT per capita is strong in the Global Chaos and Gentle Footprint scenarios due to distinct reasons. In the Global Chaos scenario, fewer trips are made due to poor economic conditions. In Gentle Footprint, we see an increase in environmental consciousness that is associated with both a decrease in reliance on the auto and an increase in the use of alternative modes of travel. Results are not always the same across the regions. For example, within the Global Chaos sce- nario, transit mode share increases by 24 percent in Atlanta, 8 percent in Boston, 21 percent in Detroit, 17 percent in Houston, and 19 percent in Seattle. One can test our assumptions about why this might be happening. For example, the relatively high transit share in Boston at the start of the modeling period leads to this outcome in the direction of the trajectory. The next chapter discusses the value of Impacts 2050 to state DOTs and MPOs. While many different inputs went into writing that chapter, one data set was derived from onsite demonstra- tions and beta tests of the poll that the research team conducted with three of the five regions for which data are included in the tool. Detailed information about these evaluative activities is provided in Appendix C. Here we present seven broad findings: • Generally, Impacts 2050 was favorably received in each of the demonstrations, and most par- ticipants were receptive to the tool concept. • The perceived utility of Impacts 2050 is initially tied to how far the transportation agen- cies have progressed with regard to their long-range transportation planning process; after

Note: M= Momentum, TT=Tech Triumphs, GF=Gentle Footprint, GC=Global Chaos Legend: M TT GC GF M TT GC GF M TT GC GF M TT GC GF M TT GC GF Auto VMT per capita Percent non-car owning Percent car-sharing Average car occupancy Transit mode share Walk/bike mode share Work trips per capita Non-work trips per capita Population (millions) Percent under 16 Percent over age 60 Percent over age 75 Percent Hispanic Percent low income Percent high income Percent foreign born Percent in workforce Seattle Indicator Atlanta Boston Detroit Houston greater than 25% increase 10-25% increase -10% to 10% change 10-25% decrease greater than 25% decrease Table 6-7. Selected indicator trajectories (2010–2050) by scenario by metropolitan region.

Scenario Planning Tool: Impacts 2050 51 consideration and discussion, participants discovered other uses for Impacts 2050 beyond its contribution to the development of their long-range plans. • There was agreement that Impacts 2050’s scenario analysis function will be useful to transpor- tation agencies. Participants welcome better ways to conduct scenario planning, and thereby reach agreement on changes from the status quo. • Two major and important advantages of Impacts 2050, compared with the models currently being used for long-range planning, are (1) it runs scenarios and produces output much faster than other models, and (2) its inclusion of socio-demographic linkages with transportation and land use fills a transportation planning gap. • A drawback to Impacts 2050, which could be a potential factor that deters receptivity to it, is that many transportation agencies have already invested in a wide array of modeling and forecasting tools; they are wary of adding another new tool that someone will have to manage and maintain (when some staff have not yet mastered those already being used). • Most modelers are used to working with spatial data, so the limited spatial definition of Impacts 2050 (urban, suburban, regional) could be seen as a drawback to its applicability. • Two keys to Impacts 2050’s adoption and use are in the quality and level of detail provided through the User Guide and the usefulness of the User Guide. In sum, the feedback received in the demonstrations indicated a need for, and interest in, Impacts 2050 and a quality User Guide.