National Academies Press: OpenBook
« Previous: Front Matter
Page 1
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 1
Page 2
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 2
Page 3
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 3
Page 4
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 4
Page 5
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 5
Page 6
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 6
Page 7
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 7
Page 8
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 8
Page 9
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 9
Page 10
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 10
Page 11
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 11
Page 12
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 12
Page 13
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 13
Page 14
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 14
Page 15
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 15
Page 16
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 16
Page 17
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 17
Page 18
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 18
Page 19
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 19
Page 20
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 20
Page 21
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 21
Page 22
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 22
Page 23
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 23
Page 24
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 24
Page 25
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 25
Page 26
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 26
Page 27
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 27
Page 28
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 28
Page 29
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 29
Page 30
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 30
Page 31
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 31
Page 32
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 32
Page 33
Suggested Citation:"Overview and Summary." National Academies of Sciences, Engineering, and Medicine. 2012. Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities. Washington, DC: The National Academies Press. doi: 10.17226/22791.
×
Page 33

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.

16 – Pedestrian and Bicycle Facilities OVERVIEW AND SUMMARY Pedestrian and bicycle facilities form essential elements of the overall transportation system, whether utilized for walking or bicycling as the primary form of travel, or as the means of access- ing other transportation modes. The first pedestrian facilities, of course, date back thousands of years. Bicycle advocates have been demanding paved facilities since the 1880s. For much of the 20th century in the United States, however, particularly during the great expansion of metropoli- tan areas into the suburbs after World War II, pedestrian and bicycle facilities received significantly less attention than was desirable. U.S. Federal funding for non-motorized transportation (NMT) improvements was increased sub- stantially in the 1990s, and interest in pedestrian and bicycle facilities grew dramatically. In the 21st Century, public health concerns have joined with transportation and environmental objectives as major forces supportive of “active transportation” enhancements (Centers for Disease Control and Prevention, 2010). The U.S. Department of Transportation has declared that walking and bicy- cling should be considered “as equals with other transportation modes” and adopted “complete streets” principles. Complete streets policy calls for “well-connected walking and bicycling net- works” (LaHood, 2010). The importance of understanding the roles and potential of walking and bicycling in the satisfaction of both travel demand and the desire for recreation and exercise has expanded accordingly.1 This chapter examines pedestrian and bicyclist behavior and travel demand outcomes in a rela- tively broad sense. It covers traveler response to NMT facilities both in isolation and as part of the total urban fabric, along with the effects of associated programs and promotion. It looks not only at transportation outcomes, but also recreational and public health outcomes, which are primarily covered as part of the “Related Information and Impacts” discussion. This “Overview and Summary” section presents: • “Objectives of Pedestrian and Bicycle Improvements,” which highlights goals and purposes of these applications. 16-1 1 Walking and bicycling, and the facilities they utilize, are together referred to as non-motorized transporta- tion (NMT). “Active transportation” is an alternative term often employed in urban planning and public health circles. Both include “any self-propelled, human-powered mode of transportation.” “Active trans- portation” is often meant to include public transportation, with its heavy reliance on walk and bicycle access. “Complete streets” are “roadways designed and operated to enable safe, attractive, and comfortable access for all users, including, but not limited to, pedestrians, bicyclists, motorists and transit riders of all ages and abilities.” NMT trips made in the service of transportation needs, such as for commuting or going to retail establishments, are classed as “utilitarian” trips. That leaves pedestrian and bicycle trips made for recreation and exercise as the other primary NMT trip category (Federal Highway Administration, 2007, Centers for Disease Control and Prevention, 2010). Overlap of motivations—and thus purposes—for walking and bicy- cling is known to exist, but is poorly researched.

• “Types of Pedestrian and Bicycle Improvements/Programs,” which categorizes and de- scribes the characteristics of the various treatments and approaches, for purposes of organization. • “Analytical Considerations,” which discusses the limitations of available information and the conclusions which may be drawn from it. • “Traveler Response Summary,” which highlights key findings presented in the “Response by Type of NMT Strategy” section. Following the “Overview and Summary” are sections on: • “Response by Type of NMT Strategy,” providing traveler response coverage of a variety of pedestrian and bicycle facilities and programs. • “Underlying Traveler Response Factors,” examining—from the perspective of travel behavior— influences affecting response to NMT facilities and programs. • “Related Information and Impacts,” addressing related issues such as NMT activity levels, safety, user health benefits, and economic considerations. • “Case Studies,” including one compilation of varied mini-studies, five selected examples of response to bicycle and pedestrian facility availability and implementation, and one example of NMT marketing. An “Adult and Child Public Health Relationships Summary,” similar to the “Traveler Response Summary” but health-focused, is provided within the “Related Information and Impacts” section at the end of the subsection on “Public Health Issues and Relationships.” Not covered by this chapter are the specific impacts of direct and indirect safety, operational, and design support for pedestrian and bicycle travel and facility development such as bicycle park- ing ordinances, safety education and enforcement programs, or construction design guidelines. Chapter 16 focuses on the travel behavior and public health implications of pedestrian/bicycle area-wide systems, NMT-link facilities such as sidewalks, bicycle lanes, and on-transit accommo- dation of bicycles, and node-specific facilities such as street-crossing treatments, bicycle parking, and showers. Discussion of the implications of pedestrian and bicycle “friendly” neighborhoods, policies, programs, and promotion is also incorporated. Related topics are addressed in Chapter 15, “Land Use and Site Design,” Chapter 17, “Transit Oriented Development,” and Chapter 19, “Employer and Institutional TDM Strategies.” Chapter 1, “Introduction,” serves to provide guidance for effective use of this and all TCRP Report 95 chapters. See especially the information and suggestions offered in the “Use of the Handbook” sec- tion of Chapter 1. Objectives of Pedestrian and Bicycle Improvements Bicycling, and to a lesser extent, walking, were—in the post-World-War-II half century—viewed mainly as recreational activities. It has been increasingly recognized, however, that walking for short 16-2

trips and bicycling for medium-short trips represent efficient, non-polluting, inexpensive modes of travel (Goldsmith, 1992). Moreover, even in communities designed in the motor age, a large proportion of trips by auto and almost every trip by bus, rail, air, or boat begins or ends with non- motorized travel. Thus, from a transportation and community perspective, objectives of pedestrian and bicycle facility improvements have evolved to include numerous aspects of providing viable and safe active transportation options for all ages, abilities, and socioeconomic groups. NMT objectives include (LaHood, 2010, Centers for Disease Control and Prevention, 2010, U.S. Environmental Protection Agency, 2010, Litman, 2011a): • Support for trips too short to be effectively served by motorized transportation. • Reduction of vehicular trips and parking demand through – diversion of short- and intermediate-distance auto trips to non-motorized travel. – reduction in chauffeuring of unlicensed youth and elders. – enhancement of public transportation through access improvement. – diversion of automotive transit access trips to non-motorized access modes. • Achievement of associated local and global environmental and security benefits through – pollutant and carbon emissions reduction. – conservation of oil and other energy resources. • Provision of economic benefits through transportation and health care cost savings. • Enhancement of mobility and safety, with attendant improvements in equity, for – unlicensed youth and elderly persons. – physically or mentally challenged individuals who cannot drive. – low income persons who cannot readily afford an automobile. – other persons without access to an automobile, temporarily or long term. – all members of society regardless of auto ownership, income status, or age. • Enhancement of quality of life through – making available a broader array of viable and attractive transportation choices. – improving conditions for pedestrians and cyclists of all types and circumstances. – providing expanded, enjoyable recreation and exercise opportunities. – expanding opportunities for chance social and community interaction. – supporting more livable, vibrant, healthy, and sustainable communities. Starting in the 1970s and brought to a head in the Surgeon General’s 1996 report, Physical Activity and Public Health, inactivity has been identified as a public health crisis now roughly of the same magnitude as smoking (Committee on Physical Activity, Health, Transportation, and Land Use, 2005, Pratt et al., 2000). The inactivity of a majority of U.S. adults is estimated to lead to 200,000 or more premature U.S. deaths annually (Heath et al., 2006). About 1/3 of U.S. adults were obese in 2007–2008. Both adult and youth obesity percentages increased markedly in the 1980s and 1990s, more than doubling or even tripling—depending on age/gender category—in 25 to 30 years (Flegal et al., 2010, Committee on Prevention of Obesity in Children and Youth, 2005). Moderate intensity activity has been shown to be of substantial benefit, with brisk walking the most univer- sally practical form. Relevant forms of active transportation include bicycling, in-line skating, skateboarding, and use of public transportation with its attendant walking for access and egress (Department of Health and Human Services, 2008, Besser and Dannenberg, 2005). The public 16-3

health objectives of pedestrian and bicycle facility improvements include the following (Centers for Disease Control and Prevention, 2010): • Expand opportunities for – safe and health-enhancing transportation choices. – convenient and affordable exercise. • Achieve increases in exercise attainable from – walking and bicycling for utilitarian travel purposes. – walking and bicycling for pleasure. • Achieve decreases in – excess body weight. – disease for which inactivity is a risk factor. Clearly the goals and objectives for pedestrian and bicycle facilities are very diverse. The corre- sponding diversity of associated benefits leads to a situation where benefit analysis based on one objective alone, such as energy conservation, will lead to a severe understatement of advantage to the public welfare. This circumstance is further expanded on in the “Economic and Equity Impacts” discussion within this chapter’s “Related Information and Impacts” section. Types of Pedestrian and Bicycle Improvements/Programs Area-wide, link-specific, and node-specific types of pedestrian and bicycle treatments are all addressed in this chapter. Area-wide approaches include providing comprehensive systems of pedestrian and bicycle facilities, pedestrian- and bicycle-friendly neighborhoods, policies and ordinances under girding NMT provisions, and active transportation promotion and information. Link-specific treatments include sidewalks, bike lanes, routes, paths, and connections to transit and activity centers. Node-specific treatments include intersection improvements and point facil- ities like showers and bicycle parking. In practice pedestrian and bicycle facilities can range widely in complexity and can involve many different simultaneous treatments. It is helpful to bear this in mind as the types of facilities and actions, and the responses to these approaches, are discussed. Sidewalks and Along-Street Walking. Paved sidewalks are constructed alongside motorized vehicle travel ways with the intent of providing a safe, attractive environment for walking, sepa- rated from motor vehicles. ADA provisions such as avoidance of sidewalk obstructions and abrupt changes in cross-slope facilitate their use by the mobility disadvantaged and the general public. While sidewalks are found to the side of almost all streets in high-density urban sectors, they are not always consistently found in lower-density city and suburban areas. Where low-density resi- dential area sidewalks are lacking, walking along low-volume, low-speed residential streets may serve as a generally inferior but workable substitute, as does use of paved shoulders in suburban and rural situations. Although it is not typically desirable for adult bicyclists to use sidewalks, there are exceptions, and such utilization does in any case occur. Street Crossings. A range of traffic engineering approaches including crosswalk and related pavement markings, signs, warning beacons, and traffic signals, as well as crossing-related traf- fic calming, can help make crossing streets at grade less of a barrier for pedestrians and bicyclists. Many such improvements carry trade-offs between conveniences to motorists versus pedestrians 16-4

and even vis-à-vis pedestrian safety. ADA requirements call for curb ramps. Marked crossings are most commonly located at intersections, but mid-block locations may be appropriate in some circumstances.2 Where a reasonably safe and appropriate crossing solution cannot be provided at grade, the layout is amenable, and the typically high costs can be justified, pedestrian/bicycle grade separations may be employed. Pedestrian Zones, Malls, and Skywalks. Pedestrian zones, malls, and skywalks, typically found in urban commercial cores, more extensively separate walkers from motorists and provide added walking space. Pedestrian zones are areas in which vehicle traffic is restricted and pedestrian travel is encouraged, generally resulting in a small-area system of pedestrian streets. The form of “Pedestrian Mall” now classified as “Traditional Pedestrian Streets” is for pedestrian use only except for off-hour use by delivery and service vehicles. Extensive landscaping and street furni- ture is typical. “Shared Malls” are similar but provide a narrow traffic-calmed passage for vehi- cles, normally a single lane in one or both directions, with or without parking. “Transit Malls” are likewise pedestrian oriented but share the right-of-way with exclusive transit vehicle lanes. Shelter for waiting passengers, and related amenities, are commonly provided. Skywalk systems connect between and through buildings above-grade to enable pedestrians to walk without traffic conflicts between business district activities. They typically utilize climate-controlled second-level pedes- trian bridges, most often mid-block. Underground tunnel networks perform the same function below-grade. Pedestrian zone, mall, and skywalk installations are often intended as strategies for stabilizing or enhancing the viability of central business district (CBD) retail and office space (Robertson 1994). Bicycle Lanes and Routes. Conventional on-road bicycle lanes are designated by signing and pavement markings, including lane striping that sets aside a portion of the roadway pavement width for preferential or exclusive use by bicyclists. An alternative provision is to have wider-than- normal shared-roadway right-hand lanes to give additional passing room for bicycles and vehi- cles, but to not actually stripe the lanes (AASHTO, 1999). Variations on the common right-side bike lane include left-side bike lanes and contra-flow bike lanes on one-way streets. Newer approaches include buffered bike lanes, with a marked buffer strip between bicycles and motor vehicles, and cycle tracks, where physical separation is employed. Physical separators may be created with raised medians, bollards, on-street parking, or by constructing a raised cycle track to introduce a grade differential. Cycle tracks may be one-way or two-way (NACTO, 2011). Another on-road approach gaining in acceptance is bicycle boulevards, a shared-roadway bicycle facility on low- volume, low-speed streets enhanced for cycling with preferential traffic calming, intersection crossing assists, pathfinder signing, and other treatments (Alta Planning + Design, 2009a). Other streets conducive to bicycling may simply be designated as bicycle routes. All such shared-roadway alternatives are designated with signs and may also receive shared-lane pavement markings, known as “sharrows,” with the included chevrons indicating recommended bicycle positioning (NACTO, 2011). Pathfinder signing may obviously be used with any type or combination of bicy- cle (or pedestrian) facilities. 16-5 2 It is critical to note that under model U.S. vehicle codes pedestrians are in a legal crosswalk, even if it is unmarked, so long as (when unmarked) they are crossing at an intersection on the extension of one of the inter- secting street’s sidewalks or shoulders (Federal Highway Administration, 2005).

Shared Use, Off-Road Paths and Trails. Off-road paths or trails have a distinctive place in the hierarchy of non-motorized facilities, being totally separated from street traffic except at road- ways. They are frequently located on old roadbeds in abandoned or “banked” railroad rights-of- way no longer used for their original purpose (“rail trails”), or similarly on canal towpaths. They may also be placed in linear and other parks, on river levees, and adjacent to vehicular roadways. Although commonly called bike paths, these shared use facilities are normally used jointly with pedestrians, joggers, and—when design and surface conditions allow—in-line skaters and other wheeled non-motorized conveyances. In this case, “shared use” means use by multiple NMT modes, but not “shared roadway” use in conjunction with motor vehicles.3 Nevertheless, low-speed motorized wheelchairs and scooters for the physically disabled are generally allowed by law (AASHTO, 1999). Pedestrian/Bicycle Systems and Interconnections. Pedestrian and bicycle systems and system expansions that provide system continuity through sheer size and good design are included in this category. Also included are pedestrian and bicycle facility segments intended to eliminate “miss- ing links” and provide important NMT network connections. Examples include bridges for pedes- trians and bicycles that cross major barriers such as freeways, railroads, or rivers. Other examples include short segments of sidewalks, pavement, or paths that join up unconnected sections or allow detour-free passage through traffic diversions or closed (or never opened) street segments such as may be encountered with traffic calming. Pedestrian/Bicycle Linkages with Transit. Access to transit and facilitation of transit trips are important roles for pedestrian and bicycle facilities. Treatments include physical connections and bicycle storage at transit stops and stations. Transit oriented development (TOD), which ideally places the most dense development in closest proximity to transit service and provides pedestrian- and bicycle-friendly design throughout the community, is a model integration of pedestrian and bicycle treatments with public transit. Also within the scope of this topic is bicycle access/egress integration with transit service in the form of bike-on-bus and bike-on-rail programs allowing tran- sit riders to bring their bicycles with them. Point-of-Destination Facilities. Point-of-destination facilities encompass those necessities and amenities required at work and other non-home destinations to enable walking and bicycling to be workable and convenient transportation modes. They thus serve to eliminate barriers to NMT use. Examples are bicycle parking, secure from theft and preferably weather-protected, and shower and locker facilities for cleaning up and changing clothes at work. Included are workplace or activity center features, such as walkability and availability of convenience services, that reduce need to have one’s car along. Bikesharing, providing short-term rentals of utilitarian bicycles, is also examined as a point-of-destination facility. Pedestrian/Bicycle Friendly Neighborhoods. A variety of neighborhood land use and site design characteristics have been identified as having an impact on the amount and frequency of walking 16-6 3 Shared use, off-road paths are unfortunately referred to by any number of potentially confusing names, includ- ing but not limited to multi-use path or trail, bike path or trail, bikeway, sidepath, hiker-biker trail, greenway, pathway, bike/ped path, and walkway, not to mention design-specific terms such as rail-trail, towpath, and boardwalk (Patten et al., 1994, AASHTO, 1999). Note that although “path” is the preferred technical term for urban applications, regional and facility-specific uses of “trail” have been adopted here where known. Thus shared-use, off-road paths specifically located in urban areas such as Seattle, Minneapolis-St. Paul, and Washington, DC, are referred to as shared use, off-road “trails” to conform with local usage.

and bicycling. Included are development density, land use mix (diversity), design features, dis- tance to transit, accessibility to goods, services, and other needs at destinations, and the overall neighborhood environment from both adult and child perspectives. NMT Policies and Programs. Supporting policies, programs, and funding at the federal, state and local level are a key ingredient in implementing pedestrian and bicycle improvements and do have their own measurable influence on growth in use of non-motorized transportation modes. Included are encouraging, retrofitting, linking, and expanding pedestrian and bicycle facilities and accommodations of all types and making them work together as an integrated NMT system. Examples include city-level programs such as those in Portland, Oregon, Davis, California, and Boulder, Colorado; national-level programs most commonly associated at present with Northern European countries, but coming to the United States with adoption of Complete Streets require- ments at the federal level; and schoolchild-focused programs such as Safe Routes to School (SRTS) activities. Walking/Bicycling Promotion and Information. Not an “improvement” in the infrastructure sense, but an adjunct for encouraging more walking and bicycling activity for transportation and health, is the promotion of active transportation and the provision of information both on NMT options available and benefits. Included are mass market information and promotions focused on inducing mode shifts to active transportation, group-targeted information and promotion with the same objectives, and similar activities focused on introducing new facilities to the public and encouraging their use. Also included is one-on-one personal promotion tailored to the interests and needs of the individual, known as “individualized marketing,” a voluntary-behavior-change approach to concurrent assistance with and encouragement of walking, cycling, and transit use choices (Brög and Ker, 2008). Finally, physical activity promotions and interventions used by pub- lic health practitioners to encourage walking and bicycling for exercise and its health benefits are covered. Analytical Considerations Well into the 1990s, NMT travel data and travel demand studies were quite limited and mainly descriptive in nature. NMT research focused primarily on safety and capacity investigations, while travel demand and behavioral aspects of walking and cycling received relatively little attention (University of North Carolina, 1994, Schwartz and Porter, 2000). Since then, NMT information and insight has received a remarkable infusion from physical activity research spurred by public health concerns (Clifton and Krizek, 2004). Substantial progress has been made in establishing existence of a significant connection between physical activity and the built environment. In the process, advances have been made in evaluating what aspects of the transportation system, community design environment, and active transportation encouragement policies tend to be associated with increases in walking and bicycling and how much so (Handy, 2004, de Nazelle et al., 2011). The process of making sense of the multitudinous new findings is moving forward. Comprehensive evaluation of NMT facility impacts and active transportation policies potentially involve numerous complex factors, yet remain relatively undeveloped compared to motorized transportation analyses (Victoria Transport Policy Institute, 2007, Committee for Determination of the State of the Practice in Metropolitan Area Travel Forecasting, 2007). The presence or absence of pedestrian and bicycle facilities can affect travel choices on many dimensions: frequency, mode, route, and time of day. On a broader level, the presence of such facilities may influence destina- tion choice, and even housing and employment location choice, thereby impacting trip distribu- tion. Facility availability for exercise and recreation may induce more physical activity, just change 16-7

where it takes place, or both. Such interrelationships are poorly understood, and deficiencies in data collection and analysis remain widespread (Clifton and Krizek, 2004, de Nazelle et al., 2011, Kuzmyak et al., 2011). It has been stated, for example, that walking is “the least understood major mode of transportation in the United States” (Agrawal and Schimek, 2007). Users of this chapter should be aware of a number of specific walking and bicycling data and analysis issues encountered in the practice of NMT research and planning and which limit the degree to which quantitative conclusions can confidently be offered. The remainder of this subsec- tion addresses such issues broadly. Additional specifics are introduced at points throughout the chapter where of special relevance. National and Regional Non-Motorized Transportation (NMT) Data Derivation of walking and bicycling activity and demand response information from national and regional NMT data, such as household travel information from the National Household Transportation Survey (NHTS) or a regional travel survey, presents a largely different set of issues than does use of data from counts and surveys of NMT facility users. These two major data sources are addressed separately here, followed by discussion of NMT trip purpose versus moti- vation identification. Purpose identification is an issue that affects all types of survey questions and applications seeking to classify NMT trips and understand reasons for NMT trip making and mode selection. Modal Definitions for Multi-Modal Trips. There are several aspects of survey-based household travel information on NMT trips that are crucial to keep in mind. One is the matter of modal def- initions and priorities. Conventional travel survey processing assigns a single mode to any trip made up of individual segments, i.e., to “linked” multi-modal trips. This mode selection is done on the basis of a hierarchy that normally gives the least priority to NMT travel modes or else employs some other identification of a single “primary” mode (Victoria Transport Policy Institute, 2007, Schneider, 2011). Thus, bicycling to a commuter rail station (for example), if it is picked up at all, will be subsumed into the commuter rail trip and not be counted in the bicycle mode. Walking to or from a bus or a parking lot a few blocks away has been almost never identified in processed regional or national travel survey data. To get at this type of NMT activity, one must typ- ically utilize specialized surveys such as transit mode-of-arrival surveys, parking facility user interviews, individual building-occupant or visitor/patron surveys, pedestrian and bicyclist inter- views, and travel surveys specifically structured to garner quantitative travel data on all configu- rations of NMT trips. Since 2001, the NHTS has served as an example of a standardized survey where NMT transit access information is at least asked about and retained in publicly available data files. Even in the NHTS, however, walking and cycling trips made to access and egress transit are not entered as individual trip records in the trip data identified by mode (Clifton and Krizek, 2004). As in those regional sur- veys which do obtain raw data on NMT access to transit, extra analyses requiring special interest, effort, and expertise are necessary to isolate and evaluate NMT travel linked to transit use. The NHTS does not obtain information on NMT travel in connection with auto use, such as walking between remote parking and the office (Agrawal and Schimek, 2007). To summarize, the processed data and documentation of most traditional large-scale travel surveys—the usual source of regional data and statements about proportions of trips made by walking and cycling—understate overall NMT activity to a substantive degree. Only trips made exclusively by walking or bicycling are iden- tified as NMT trips in the standardized data compilations and reporting of such surveys, and in many cases information on the NMT component of multi-modal trips is never obtained at all. 16-8

Poor Survey Respondent NMT Trip Recall. A second important consideration in using travel- survey-based information is the extent to which survey respondents may not recall or understand they should report NMT trips, particularly when not prodded to do so. The trips least well recalled are non-work trips and short trips. Walk trips in particular are predominantly both of these, and thus tend to be underreported (Victoria Transport Policy Institute, 2007, Agrawal and Schimek, 2007). Non-work vehicle trips are sometimes adjusted upward in regional studies on the basis of screenline information, but this is a procedure rarely if ever attempted for NMT trips. Presently available count information is typically insufficient for such use. The gradual shift from trip-based to activity-based surveys has helped survey respondent recall, but the problem has not been eradicated. Use of Global-Positioning-Systems (GPS) devices for regional surveys holds future promise for addressing poor trip recall effects. Work remains, however, on determining whether and which differences between survey-reported and GPS-recorded travel inventory results reflect survey- respondent recall problems or GPS-recorded trip misreporting (Bricka et al., 2011).4 Changes in survey methodology necessarily have an adverse effect on the validity of comparisons over time. For example, walk trip shares from the NHTS cannot be directly compared with those from the predecessor National Personal Transportation Surveys (NPTS) because of significant walk trip reporting increases obtained through trip recall prompts instituted in the 2001 NHTS pro- tocol (Hu and Reuscher, 2004) and maintained in the 2009 NHTS survey. Also, whereas the NPTS did not survey the trips of children under age 5, the NHTS does so (Liss et al., 2003). A less crucial but nevertheless troublesome survey respondent recall problem (affecting all travel modes) is the tendency to report trip start and end times to the nearest 5 or even 15 minutes. This tendency can warp trip-survey-based calculations such as estimation of average walking speed. The 2001 NHTS reports walk times and distances (Agrawal and Schimek, 2007) that suggest an average walk speed of only 2.2 miles per hour (mph). The reason for this dubious value may lie in imperfect trip start and end time reporting. The comparable value from the 2009 NHTS is 2.8 mph (Kuzmyak et al., 2011), more reasonable, but suggesting instability in the calculation inputs.5 Limited Non-Work and Child Trip Data. A third consideration is that some surveys address only travel to and from work, most notably the widely used decennial U.S. Census travel data and its replacement, the yearly American Community Survey (ACS). As noted in a Victoria Transport Policy Institute publication, referring to U.S. weekday travel, “Only 7% of walking trips and 8% of 16-9 4 Survey-reported versus GPS-recorded differences seen in a 265-sample analysis of 2009 Indianapolis regional- survey GPS-trial results do not fully conform with conventional wisdom. Trip totals were lower for survey- reported than GPS-recorded records for lower-income families, as would be expected, but the same discrepancy was also found for busy professionals and volunteers. Among groups with work-trip reporting results most closely matching between methodologies were non-Caucasians, middle-income workers, and persons commuting via transit and NMT. More non-work trips were obtained from survey responses than GPS readings for the elderly and retirees, including persons reporting walk and bicycle trips. In Indianapolis, dependence was placed on GPS readings for determining the beginning and ending of trips and workplace orientation (Bricka et al., 2011), unlike the 2007 Portland, Oregon, studies of bicycle trip route choice described later, which asked each participant to key trip beginning, ending, and purpose information into the GPS devices (Dill and Gliebe, 2008). 5 The 2009 NHTS travel diary sample size of 150,000 households is, to its benefit, well over twice the 2001 NHTS sample size of 64,000 including localized sample add-ons (Kuzmyak et al., 2011).

cycling trips are to work, a far smaller portion than for motorized travel, so surveys that focus on commute trips are particularly likely to under[-emphasize] non-motorized travel” (Litman, 1999). School travel is necessarily omitted in surveys that obtain only work-trip information, and even when sought, there are concerns that travel by children is underreported (Victoria Transport Policy Institute, 2007). Even when data have been available for non-work or recreational walking and bicycling, effects of trip purpose acting in combination with other factors have received only spotty attention. Cross- classifications involving purpose have only rarely been developed in pedestrian and bicycle trip data mining, a limitation highlighted in the “Factor Combinations Involving Trip Purpose” discussion of the “Underlying Traveler Response Factors” section. A related issue is that some regional travel survey processing procedures have called for discard- ing (or setting aside) some or all trips with the same origin and destination, typically the home. This protocol may limit the utility of such surveys for analysis of walking or cycling that takes place purely for recreation or exercise, or at lease cause such trips not to be included in standard travel compilations. An example of the latter, involving non-motorized recreational trips starting at and returning to the home without an intermediate stop, is provided by the “Travel Behavior Inventory”—the regional survey for the Minneapolis-St. Paul area. In the 2000 survey, such trips were recorded in the household data along with distance covered, and have been used for univer- sity research, but were not entered in the trip-data files (Filipi, 2011). Depending on the particular survey design, ability to analyze chauffeuring of children and other non-drivers who might con- ceivably walk or bike—were adequate facilities to exist—may also be affected. Bureau of Transportation Statistics (BTS) reporting on NMT data limitations, in addition to touch- ing on many of the other survey issues raised here, observed that travel by preschoolers is rarely obtained. Moreover, trips by children of approximately grade school and junior high age are often obtained from adult proxies, reducing the likelihood of picking up all travel (Schwartz and Porter, 2000). The NHTS has at least attempted to record the trips of children of all ages starting with the 2001 survey (Liss et al., 2003). Nevertheless, many research studies reflect the limitation of having investigated only the travel or exercise of working age adults. The NHTS series of surveys, covering 2001 and 2009 so far, provides a rich resource at the national level of U.S. non-work travel by all modes including walking and bicycling (Agrawal and Schimek, 2007). It also includes trips by children of all ages (Liss et al., 2003). In working with the NHTS it is essential to take into account that it is a 7-day-a-week survey representing an average day, not an average weekday. Of sample days, 29 percent are weekend days (McGuckin and Srinivasan, 2005). Trip purpose distributions from the full spectrum of NHTS trip records reflect the inclusion of weekend travel, whether for motorized or non-motorized travel modes. Lack of Consistency in Trip Counting Protocols. Also requiring close attention is a lack of com- mon protocols in travel-survey-based information, especially when walking and bicycling for recreation and exercise is involved. For example, there appears to be no consensus protocol for defining a one-way trip equivalent for so-called “loop trips” that begin and end at the same point (if such trips are counted at all). Loop trips from the home, and sometimes from work, are typi- cally encountered when the trip purpose is recreation/exercise. Despite the terminology, they include simple out-and-back trips using the same routings in both directions but involving no des- tination activity at the farthest-out point. The NHTS splits such loop trips into two one-way trips by asking the survey respondent to identify—as a “destination”—the farthest point reached (Agrawal and Schimek, 2007). Other 16-10

surveys, such as the National Highway Transportation Safety Administration (NHTSA) and BTS 2002 summer survey and the 2007 GPS-based survey of cycling in Portland, Oregon, have treated the entire loop as a single one-way trip for purposes of trip reports (NHTSA and BTS, 2002, Dill and Gliebe, 2008). Thus, survey findings based on an approach similar to the NHTS report twice as many recreational/exercise trips for a given amount of activity, and half the average trip length, as compared to studies using the Portland GIS-study approach. Large differences among trip length reportings for recreational/exercise trips suggest that this is likely a pervasive defi- nitional problem,6 and there may be others like it. A related issue is what constitutes a walk trip that should be recorded at all. “Tours,” series or chains of work-related or non-work trips made starting at and ultimately returning to the same location, present a special problem. The nature of activity required to define the end of one trip and the start of another has not been well defined, a deficiency that remained unsolved for the 2001 NHTS. (Does buying a newspaper on the way to lunch count as a trip-ending/beginning?) Even simple one-leg trips present a problem. (Is crossing the street to visit a neighbor a trip?) It has been suggested that the questioning process include a request to treat each change of address location as a trip (Clifton and Krizek, 2004). Most-Recent Trip Versus Trip-Day Travel Data. Travel-diary and activity-diary surveys focus on a set period, most often a “survey day” during the working week, and so produce information on a typical weekday’s trips. More abbreviated surveys may ask about specific types of trips made on the previous day. Some NMT survey information is, however, obtained for the most-recent walk or bicycle trip. For example, the 2002 summer survey performed by NHTSA and BTS utilized a variant of the “most recent trip” inquiry methodology in that it recorded NMT data for the day (within the last 30 days) of most recent walking or bicycling activity (NHTSA and BTS, 2002). Such surveys overemphasize trips by persons who walk or bicycle less and underemphasize trips by persons who do so more frequently yet have only one trip covered in the survey. This approach can skew data ranging from age distributions to trip purpose percentages to trip length averages. There may be reasons for finding out about the walking and bicycling of persons who engage in these activities less frequently. If one is trying to use such a survey to describe trip making by a representative population, however, there are built-in biases to address. The nature of such biases has apparently not been investigated quantitatively, but it has been suggested that longer recre- ational trips—such as hikes—may be overemphasized. Self-Reported Information and Perceptions. Public health professional involvement in walking and exercise research, in many cases representing a first attempt at epidemiological consideration of the transportation and land use environment, has introduced certain additional issues to be aware of. On the one hand, health researchers are bringing badly needed added attention and rigor to statistical designs, along with new analysis techniques and a broad knowledge base concerning 16-11 6 A circa 2005 survey of bicyclists by a large metropolitan planning organization (MPO) apparently fell into this inconsistency trap. Although questions on the survey instrument inquired about one-way bicycle trips, interviewers were instructed to treat out-and-back recreational and exercise trips as a single trip. Trip lengths were compared with 2001 NHTS results, and the conclusion was drawn that MPO-area recreational and exercise bicycle trips were over twice as long as the national average. This was likely an inadvertent apples and oranges comparison, given that (as described above) loop trips surveyed in the NHTS are split in two, whereas in the survey in question they were apparently not, at least not if they were identified as recreational/exercise trips.

such matters as the reliability of self-reported socioeconomic characteristics, actions taken in daily life, and health status. On the other hand, certain self-reported information—while perhaps use- ful in analysis of perceptions relevant to behavioral intervention—comes with questions related to its use as independent variables for understanding travel behavior. A case in point, which comes from both health and transportation research, is asking a survey respondent if some place or activity, such as shopping, is within walking distance. It would seem that such indicators could vary considerably in their quantitative basis: For example, what would be perceived as “beyond walking distance” for an obese person would likely be quite different than for a fit individual. Moreover, various studies suggest that factors in the environment can affect distance perception. Self-reported distance walked, expressed in terms of travel times, has been subjected to quantita- tive investigation vis-à-vis estimated actual travel times. The study in question surveyed individ- ual perceptions of proximity to various types of businesses and facilities in Minneapolis and its inner and outer suburbs. Actual locations were geocoded, allowing perceived travel times to be compared to both airline and transportation-network distances and corresponding travel times estimated using average walking speeds. Both perceived and actual travel times were grouped into 1-to-5, 6-to-10, 11-to-20, 21-to-30, and over-30 minutes categories. Perceived travel times matched the corresponding estimated-time category only 37 to 38 percent of the time (Horning, El-Geneidy, and Krizek, 2008). Results of surveying or analyzing perceived values such as travel distance need to be treated with suitable caution. Moreover, with regard to findings derived from modeling, it also needs to be kept in mind that the strength of other variables modeled concurrently may be affected by inclusion of variables based on perceptions. Facility Counts and Research Surveys Transportation planners of motorized facilities and services rely on widely accepted vehicle and pas- senger count methods for measuring use and validating estimating models. The available procedures include periodic statistically controlled street and highway counts, vehicle classification counts, full- time highway count stations, and transit passenger counts and rider surveys (Shunk, 1992, Kell, 1992, Cambridge Systematics et al., 2011). Walking and bicycling have been afforded no such consistent data collection and processing systems. Each agency has tended to conduct any NMT counts, sur- veys, and analyses in its own way, and there has been little national sharing of data (Jones, 2009). National Perspectives. A 2004 review of NMT data collection in the United States found many communities and agencies following at least part of an organized pedestrian and bicycle data col- lection process, although many did not. Among agencies collecting data, it was a challenge to for- malize results and make them publicly available. There was no uniform, national NMT data format to rely on. Of 29 pedestrian and bicycle data collection case studies developed by the authors, only 2 were supportive of evaluations that could contribute directly to traveler response evaluation (Schneider et al., 2005). Recognizing the need, the National Bicycle and Pedestrian Documentation Project (NBPDP) was initiated in 2002 as a voluntary cooperative effort by Alta Planning + Design and The Institute of Transportation Engineers (ITE). This effort has sought to establish a “consistent national methodol- ogy” and assemble counts and other information into a starter database. The taking of annual counts is encouraged, focusing on one 2nd-week-of-September weekday in the peak periods (7–9 AM, 16-12

4–6 PM) and one weekend in the midday (12–2 PM). It must be understood, however, that the NBPDP (at least as of January, 2009) is unfunded, with no resources to conduct quality assurance/quality control on incoming data (Jones, 2009). Related limitations include automated pedestrian and bicyclist counters that are more difficult to properly deploy and less reliable than the motorized-vehicle equivalent, requiring calibration to manual counts (Lindsey et al., 2006, Schneider, Arnold, and Ragland, 2009), and sometimes uncritical acceptance of counts, surveys, and observations contributed by NMT advocacy groups. Effects of Exogenous Events and Circumstances. The occurrence of exogenous influences and events is a potential problem for any transportation data collection effort. As discussed in the “Natural or Artificial NMT Volume Variability” discussion to follow, special events may strongly affect pedestrian volumes. Such events include athletic contests, concerts, and any happening that draws large crowds to areas with pedestrian accessibility. In some cases events affect bicycle vol- umes as well. Shared use trail volume modeling in Indianapolis included a “state fair in session” variable in the Monon Trail model to address a key special event impact (Lindsey et al., 2006). Exogenous influences and events that obviously or possibly caused NMT-count, survey, compar- ison, demand-model, or long-term-outcome abnormalities are noted where relevant throughout this chapter. Such instances include sharp growth in university enrollment and employment, lack of hard surfaces in an urban path system, urban Interstate bridge collapse at a CBD cordon, earth- quake destruction, weather conditions, and a mass-transit strike. Some of these real-life examples may seem overly obvious, but if not documented in connection with data presentations, they can raise questions or be overlooked in later applications and interpretations of the data affected. An ever-present exogenous circumstance is the area type context, in terms of land use and demo- graphics, for individual facility improvements and other actions. A new sidewalk or bicycle con- nection (or walking/cycling encouragement program) in a dense, urban, mixed-use area may promote a significant mode shift that is in part thanks to having many people living and working in close proximity to the improvement. Adding the same facility or taking the same action “in a low-density suburban area with separated land uses may produce a minimal mode shift” because residences and activity destinations are simply too far apart for most trip makers to contemplate walking or bicycling. New facilities in such environments may increase recreational walking and bicycling, and increase safety, but utilitarian travel may continue using the auto (Schneider, 2010). In the case of heavily recreational shared-use-path volumes, trail use research in Indianapolis sug- gests that neighborhood demographic makeup may be more important than land use and design in influencing trail usage, although both were found significant (Lindsey et al., 2006). In any case, such findings highlight “the difficulty of drawing general conclusions about the pedestrian and bicycle volume impacts of pedestrian and bicycle facilities in different [. . .] contexts” (Schneider, 2010). Important land use context information is reported where known, and effects of land use are exam- ined in the “Pedestrian/Bicycle Friendly Neighborhoods” discussion within the “Response by Type of NMT Strategy” section. Natural or Artificial NMT Volume Variability. Pedestrian and bicycle counts tend to exhibit much higher variability than observed for equivalent vehicle or transit passenger volumes. Walking and cycling are more affected by the day-to-day and season-to-season weather than other travel modes, making conduct of “typical day” counts more problematic. Commercial and sports/entertainment area pedestrian volumes, reflecting as they do what is going on close at hand, can be strongly affected by special or localized events and situations including conventions, opening or failure of popular stores, and local economic conditions at the time. NMT facilities exhibit significant hourly varia- tion, often making designation of standardized peak periods meaningless (Bruce, 2002a and 2004a 16-13

and b, Jones, 2009). This variability notwithstanding, it is not uncommon to limit NMT counts and surveys to fixed one-or-two-hour peak periods, raising concern about the reliability of such data gathering. A count variability example is provided by pedestrian crosswalk counts presented in this chap- ter’s first case study, “Special Mini-Studies in Montgomery County, Maryland” (within the case study, see “More—Volume Variability”). Counts taken on parallel crosswalks illustrate how strongly pedestrian volume characteristics reflect the events of the day and/or the nature of nearby land development. Counts one year later illustrate how much difference there can be between two counts at the same location, for whatever reason. Pedestrian activity is much more localized than transit or automobile passenger flows. Comprehensive surveys circa the 1970s in Chicago’s Loop found State Street pedestrian volumes between Madison and Washington Streets to be 4 to 5 times the volumes three blocks up the street, and over 7 times the pedes- trian volumes 5 blocks over to the west. Similar phenomena were recorded in 6 other major U.S. cities coast to coast (Pushkarev and Zupan, 1971, Wilbur Smith and Associates, 1970, Levinson, 1982). (For more detail, see “Related Information and Impacts”—“Facility Usage and User Characteristics”— “Sidewalks and Other Provisions in Major Central Business Districts”—“Central Business District Pedestrian Volume Characteristics.”) There has been little study of variability in bicycle volumes. Perhaps as critical as variability in the case of bicycle volumes is the small proportion of all travel, and even of NMT travel, that bicycling today represents in the United States. The previously noted BTS study points up the possible inad- equacy of sample sizes for specific examinations of travel by lesser-used modes such as bicycling (Schwartz and Porter, 2000). Errors in measuring and estimating bicycle volumes will be magni- fied in the context of describing changes from low-activity base-case conditions. NMT Facility Survey Design Issues. Intercept surveys of NMT facility users, in addition to encoun- tering the same traffic flow variations noted above, introduce other methodological issues. NMT/active-transportation survey instrument definitions and application designs have less com- monality than typical for transportation surveys. This lack of standardization perhaps occurs because a broader range of professional backgrounds is reflected in their design, a range encompassing not only transportation planners and traffic engineers, but also parks, recreation, and public health pro- fessionals. A case in point involves determination of the purpose of active transportation trips. There appears to have been a tendency to obscure utilitarian purposes of travel by asking survey questions in such a way that “motivation” has superseded trip “purpose.” This particular problem, not limited to individual-facility studies, has sufficiently widespread implications that it is afforded a separate discussion under the “Trip Purpose Versus Motivation” heading. NMT facility observations and intercept surveys taken “on-line” (on-facility), i.e., at a point on the main walkway or path, may bias analyses if one is attempting to derive conclusions about trip- based use and usage characteristics. In such an on-facility intercept survey, trips in categories asso- ciated with longer trip lengths are more likely to be intercepted, and thus overemphasized in summaries relative to categories associated with shorter trip lengths. This phenomenon will adversely affect discernment of usage characteristics such as trip purpose percentages, trip length, NMT travel-mode mix, child versus adult proportions, and even gender proportions. Classification counts and surveys taken on-facility reflect the mix of user traffic at points along the facility, use- ful for operational analyses of facility traffic, whereas interception of persons starting or ending facility use allows analysis without trip length bias of the mix of users visiting the facility. The lat- ter method, as employed in year 2000 Indiana trails surveys, provides actual trip-based user data in the manner of an attraction survey. 16-14

For example, proportions of trips for commuting and recreation, by adults, and—especially on shared-use facilities—by bicyclists, will probably be over-reported in on-facility intercept surveys. Use for running errands, by children, and by pedestrians will be underreported. If trip-length data were obtained, it should be possible to compute bias corrections, but use of such an approach is presently rare. The only direct approach to avoiding this bias entirely is to conduct intercept sur- veys of persons entering or exiting the facility, the methodology used in the year 2000 Indiana University trail use studies (in the “Case Studies” section, see “Six Urban, Suburban, and Semi- Rural Trails—Indiana Trails Study”). The Monon Trail of Indianapolis, included in the Indiana trail entry/exit interviews, was reexamined just under 4 years later using field observations taken in accordance with the on-facility intercept approach. The on-facility observations showed a 165 percent higher proportion of bicycle- mode users, shifting the reported majority user from pedestrian to bicyclist, and a 24 percent higher proportion of male users, shifting the reported majority user from female to male (Indiana University, 2001, Lindsey et al., 2006). The passage of time, seasonality, and interview versus field observation approaches to data gathering may have had some effect. The major differences fit, how- ever, with the expectation that the proportion of longer trips would be boosted by the choice of an on-facility survey approach.7 (A full numerical comparison is provided in Table 16-107 of the “Related Information and Impacts” section, under “Facility Usage and User Characteristics”— “Off-Road Shared Use Paths”—“Path User Mode Distributions” and “Mode-of-Access Distributions.”) Public Opinion and Preference Surveys. Public opinion and preference surveys, as contrasted to carefully structured and modeled stated preference survey experiments, have been largely discred- ited as a basis for direct estimation of motorized transportation facility usage. This type of public sur- vey has nevertheless commonly been used in the active transportation community to determine motivations and estimate the degree of increased bicycling and walking likely with improved NMT facilities. Typically, potential improvements are described and the persons surveyed are asked if the changes would lead them to walk or cycle. Findings from such surveys are often contradictory, and frequently exaggerate the potential of prospective improvements. For example, 1970s surveys of res- idents in Madison, Wisconsin, found 21 percent willing to bicycle to work if there were better facili- ties. After bicycle lanes and paths were provided, it was found that the share of bicycles in traffic had risen from 4 percent to 11 percent: a very substantial increase but about half that projected (Zehnpfenning et al., 1993). Before and After Surveys. “Before-and-after” surveys and the analyses based on them are essen- tially two-point-in-time longitudinal studies, with one set of observations before an action such as facility improvement, and the other afterward. They do not track individuals, however, unless a panel-survey approach is used. The introduction to this “Analytical Considerations” discussion highlights the many dimensions in which presence or absence of pedestrian and bicycle facilities can affect travel choices including trip frequency, mode, route, time of day, destination, and dis- tribution. The limited number of before and after studies available of response to pedestrian and bicycle facility improvements generally examined only one or two of these dimensions at most, or utilized volume as the travel measure (see the various strategy assessments within the “Response by Type of NMT Strategy Section”). This limitation makes difficult such assessments as quantification of latent or induced demand, frequently discussed in the context of motorized facilities, but also a factor of interest to planning 16-15 7 This assumes that male trail users make longer trips on average than female users, primarily because a lower proportion of females typically choose to bicycle.

non-motorized improvements. Introduction of pedestrian and bicycle facilities may inspire indi- viduals to undertake walk or bicycle trips not made previously by any mode, as well as to shift from motorized to non-motorized modes. On the other hand, NMT volume increases observed in connection with individual facility improvements may largely represent shifts from alternative walking or cycling routes by people already using active transportation. Travel choice sensitivi- ties, discussed further in the “Underlying Traveler Response Factors” section (see “Behavioral Paradigms”), suggest that such route shifts are a major contributor to usage of new facilities, espe- cially within dense networks. Before-and-after surveys may have other analytical problems, such as inadequate sample size, various forms of potential survey response bias, and exposure to exogenous influences that may distort outcomes. A full discussion of such challenges is provided in connection with individual- ized marketing. (Within the “Walking/Bicycling Promotion and Information” subsection of the “Response by Type of NMT Strategy” section, see “Individualized Marketing”—“Critiques of Home/Community-Based Individualized Marketing Assessments.”) It is important to stress that the challenges listed there are not unique to individualized marketing. Indeed, the same or simi- lar issues are encountered in before and after analyses of all types of NMT facility improvement and program actions, and are rarely as well addressed as has been done in a number of individu- alized marketing assessments. A before-and-after survey problem potentially troublesome in NMT facility improvement evalua- tions is the exceptionally long time it may take to fully establish usage patterns for a new facility. Data on this subject is scarce, but time-series counts in Melbourne and Seattle suggest that in the first year of all-new facility availability, weekday usage may be only 15 to 40 percent or so of fully stabi- lized usage. Complete establishment of usage patterns may take 7 or 8 years (Davies, 2007, Moritz, 1995 and 2005a and b). Motorized facility trends suggest usage at the end of two years may be an acceptable albeit incomplete indicator of long-term travel response, but NMT facilities appear to require more time for usage to become established. “After” surveys taken too soon will fail to pick up usage by people requiring more time to take advantage, while attempts to defer such surveys until further demand stabilization has taken place may run afoul of the confounding events that come with the passage of time. Available usage stabilization information is provided in the “Time to Establish Facility Use” subsection of the “Related Information and Impacts” section. Trip Purpose Versus Motivation The challenge of determining the purpose of trips made on NMT facilities, already noted, is appli- cable to both national/regional surveys and facility surveys. Much active transportation may fall into a large gray area where the pedestrians and bicyclists involved are actually killing two birds with one stone—deliberately choosing to exercise while also accomplishing utilitarian NMT travel (trips for “transportation”). When asked, under these circumstances, to provide a single-choice answer to a trip purpose survey question, the interviewee is faced with a dilemma regarding what to answer. Especially in surveys of shared-use paths and other facilities attractive in their own right, there may be an inclination to ask purpose of path use or “trail visit” in such a manner that “motivation” over- rides conventional trip purpose (in the transportation planning sense), obscuring utilitarian purposes of travel. Even purely trip-oriented surveying of trip purpose may be producing unclear results. Exercise or recreational motivation appears to be not infrequently reported as a purpose when there was in fact a utilitarian trip involved, even commuting to work. Identified Cases of Motivation Versus Purpose Confusion. Two studies in particular illustrate “exercise” or “recreational” purpose reporting when there was actually a utilitarian trip involved. 16-16

Interviewers on the Iron Horse Regional Trail in the San Francisco East Bay area made specific note of a tendency for interviewees to give “recreation” as a trail trip purpose when, in fact, they had actual utilitarian destinations but were choosing trail use as an exercise opportunity (East Bay Regional Park District, 1998). The Indiana Trails Studies of 2000 asked users about both “main” and “other” purposes of visiting the trail. On the Monon Trail in Indianapolis, among those users who volunteered an answer to the other-purpose question, 12 percent reported “commute.” Another 3 percent reported various utilitarian secondary purposes ranging from dining to business (Indiana University, 2001). Additional context and information is provided in the “Case Studies” section (see “Six Urban, Suburban, and Semi-Rural Trails—Indiana Trails Study”—“More . . .”). Reporting of Both Motivation and Purpose. One example encountered provides insightful infor- mation on both purpose and motivation. The source is a post-graduate student survey on the Goodwill Bridge NMT crossing of the Brisbane River, in Brisbane, Australia. Although 82 percent of bridge walkers and 72 percent of bridge bicyclists were found to be making a commute to work or school in this weekday-peak-periods survey, 56 percent of walkers (including 59 percent of com- muters) and 60 percent of bicyclists (including 58 percent of commuters) reported an exercise moti- vation for using the bridge (Abrahams, 2002). (For more detail, see Tables 16-24 and 16-25 and the accompanying discussion in the “Response by Type of NMT Strategy” section under “Pedestrian and Bicycle Systems and Interconnections”—“River Bridges and Other Linkages”—“Goodwill Bridge, Brisbane, Australia.”) NMT Modeling and Research Procedures In TCRP Report 95: Traveler Response to Transportation System Changes, the focus is on observed trav- eler response relationships. Demand model findings add important additional insight, however, and also serve to fill gaps in the observed response record. Thus the limitations of NMT demand models, along with research procedures and coverage, are relevant. Common Research Model Limitations. A fairly large number of research models have actually been developed in the last 10 to 15 years in attempts to investigate and describe factors affecting the choice of walking and/or bicycling. The results have contributed greatly to the understanding of NMT travel choices. Most of the research by university and other research organizations has, however, not been of a scale allowing use of regional transportation network data for describing the en route travel characteristics and options affecting choice of travel between individual origin and destination pairs, or even to develop and use accessibility measures employing these charac- teristics. At the same time, relatively few regional agencies have had the transportation network information required to fully describe pedestrian and bicycle facility availability and quality along individual links of the transportation network. Most research modeling has focused on developing and using models exclusively employing neighborhood and other “trip-end” area descriptors—including socio-demographic information and sometimes proximity measures—as contrasted to models taking into account NMT network characteristics specific to individual trips (e.g., Kitamura, Mokhtarian, and Laidet, 1994, Saelens et al., 2003, Krizek and Johnson, 2006, Cao, Handy, and Mokhtarian, 2006). A few studies have used GIS information to identify specific features encountered en route, such as need to cross a busy street or percentage of arterials with sidewalks (e.g., Troped et al., 2001, Moudon et al., 2007). Fewer still have more fully employed regional study network data, either to develop accessibili- ties or to explicitly examine conditions via minimum and/or chosen paths (e.g., Kockelman, 1996, Broach, Gliebe, and Dill, 2009a and b and 2011). Reliance on local-area descriptors dulls ability to describe NMT facility availability and features in a context relevant to specific travel desire lines. 16-17

This lack of full spatial demand and supply representation may be causing underemphasis in model results on effects of facilities, and overemphasis on factors that can be described well with “trip end” data—such as socio-economic factors or attitudes, for example. It is mainly research and subsequent applied models developed for or by regional agencies that have made use of transportation network data (e.g., Reiff and Kim, 2003, Kuzmyak, Baber, and Savory, 2006, Lawrence Frank & Co., SACOG, and Bradley, 2008), but these also are few in num- ber. Limitations in network information on NMT facilities have led to use of surrogates, such as having street intersection density stand in for pedestrian and bicycle system interconnectivity. The street intersection density variable has generally worked well for regional models (Reiff and Kim, 2003, Lawrence Frank & Co., SACOG, and Bradley, 2008), but would fall short in describing any NMT network not closely aligned to the street system. The lower speeds and lesser range of walk-mode trips cause pedestrians to be highly exposed to microscale features of the immediate transportation and land use environments. This circumstance places extra demands on the scale of analysis, with Census-tract levels of detail not up to the task (Clifton and Krizek, 2004). The same can be said of traffic analysis zones (TAZs) and perhaps even blocks. Land use, transportation, and public health research in greater Seattle at the land-parcel-level of detail has pro- duced fruitful results, discussed at several points in this Chapter, but even that research has not man- aged the full integration of fine parcel-level detail with regional transportation model spatial representations of travel demand and facility networks (Moudon et al., 2007, Lee and Moudon, 2006a). These limitations of research-scale and applied NMT travel demand models lead to concern that NMT network characteristics, such as NMT system connectivity, may often have been described in a manner inadequate for identifying the nature and full importance of system effects on choice and use of NMT modes of travel. Cross-Sectional Studies and Causality Issues. The vast majority of the new body of research con- sists of cross-sectional analyses, although a few new “before and after” type evaluations have been added by transportation and health researchers, and some true experimental trials have been run on interventions to promote active-transportation-based exercise. In working with cross-sectional studies—found in both physical activity research and travel demand modeling—it is essential to keep in mind that correlation does not prove causality. Which caused what may not even be read- ily evident. Correlation between built environment characteristics and rates of walking and bicy- cling activity does not necessarily prove the environment directly caused the degree of activity (Committee on Physical Activity, Health, Transportation, and Land Use, 2005). One example of a confounding issue is that of “self-selection”: Did an environment favorable to walking and cycling cause more NMT activity, or did the environment simply attract persons predisposed to walking or cycling? (For more on “self-selection,” see “Underlying Traveler Response Factors”—“Choice of Neighborhood/Self-Selection.”) Deficient Research Methodology. There is always the concern that a deficient analytical approach will result in overstatement, as in the case of the 1970s Madison, Wisconsin, example noted previ- ously in connection with “Public Opinion and Preference Surveys.” Indeed, supposedly sophisti- cated forecasting studies of proposed toll roads and urban rail systems have an international record of overestimation in the past, attributed to the institutional climate, and labeled “optimism bias” (Committee for Determination of the State of the Practice in Metropolitan Area Travel Forecasting, 2007). The equivalent can happen in reverse, however, in active transportation research. As part of a sys- tematic review of interventions to promote walking, studies by public health professionals were 16-18

ranked by Scottish Physical Activity Research Collaboration (SPARC) investigators on a seven- point scale reflecting seven procedure-validity criteria. Examination of 25 studies on promotions of walking found that all studies ranking “6” or “7” for validity, seven studies in all, produced sta- tistically significant findings. Not quite half of the 13 studies that ranked “5” had statistically sig- nificant outcomes. None of the five studies with lower rankings reported statistically significant results. Moreover, the median increase in time spent walking isolated in the seven top-ranked studies was 54 minutes, the median increase for the 13 mid-ranked studies was 32 minutes, and the median increase for the five studies ranked lowest for validity was 0 minutes (Ogilvie et al., 2007). Similar relationships have been seen, but within a much smaller sample of studies, in research on effects of sidewalk presence and condition on exercise by children (Davison and Lawson, 2006). Insufficient Second-Order-Effects Research. Questions about the nature and importance of second-order effects remain to be resolved. A number of examples suggest existence of more nuanced or complex relationships than have been fully investigated to date. Illustrating this point is a paired-neighborhood study which found more walking and cycling in the non-auto-oriented neighborhood, an expected and desired outcome from a transportation perspective, but a level of physical activity overall that was so slightly higher as to not be statistically significant. Other types of exercise were being substituted for active transportation, a secondary effect that appeared to diminish public health implications (Committee on Physical Activity, Health, Transportation, and Land Use, 2005). Another study reported in a synthesis update found, among persons doing at least some walking, that total time spent walking was less in areas with more physical activity facil- ities (Saelens and Handy, 2008). Findings such as these have to be regarded as needing further replication before full acceptance, but they serve to underscore that there has been little exploration of possible second-order effects and counterbalancing outcomes. Deficient Research Coverage. One area of interest in particular stands out as having so little avail- able travel demand and response information as to render impracticable any substantive treatment in this chapter. It is the effect of ADA-compliant pedestrian and multi-use facilities (or lack thereof) on the travel and mobility of people with disabilities. The importance of having facilities which people with disabilities can readily use to get from one place to another is fairly self-evident. Obvious benefits include enhanced mobility with attendant social and economic benefits, a better life for the affected persons, and reduction in need for special social services such as costly door- to-door paratransit. The transportation and active-living public-health literature utilized as a basis for this chapter was not, however, found to contain data or research focused on impact of ADA- compliant pedestrian and multi-use facilities—or even presence or lack of sidewalks—on trip- making or exercise by people with disabilities. Thus, despite its importance to a consequential segment of the population, there has been no basis for inclusion of a full discussion. A few percep- tive observations found in the literature, and one special-purpose capital cost recovery estimate, have been provided. A second area of interest not quite so devoid of research, but only beginning to receive attention in empirical studies, is the effect of “tour” composition on the choice of whether to use or not use active transportation. Advanced travel demand modeling investigations indicate that people are influenced in their travel mode choice decisions by all the travel requirements they encounter during any chain of trips made on a tour of activity stops. Travel data keyed solely to individual trips (links) between pairs of activity stops is effectively divorced from information on other elements of a tour. Since most past travel surveys available for analysis were “summarized using trips as the unit of analysis, this can lead to some problematic interpretations of pedestrian and bicycle mode choice.” An example is use of trip length as an indicator of walking or bicycling likelihood. In many cases trips shorter than a mile, and thus seemingly obvious active-transportation candidates, are components of much longer 16-19

tours—perhaps 10 miles or more in total length. An auto may be needed for the full tour even though individual trips within it could theoretically be taken by walking or bicycling (Schneider, 2010). Research available on this issue has allowed only limited examination here, found mainly in the “Underlying Traveler Response Factors” section (see “Trip Factors”—“Walk Trip Distance, Time, and Route Characteristics”—“Walk Trip Speeds and Lengths”). Traveler Response Summary In this summary of traveler response to pedestrian and bicycle facility implementation and pro- grams, both highly positive and neutral results will be found. However, the general indication is that the phrase “if you build it they will come” does apply to pedestrian and bicycle facilities and programs that are well planned, especially facilities that are well oriented to utilitarian travel pat- terns and points of activity. Of special interest is that the more robust positive results tend to occur when a systems approach encompassing enhancement of connectivity is followed. This will be seen especially in the summaries and corresponding subsections for “Pedestrian/Bicycle Systems and Interconnections” and “NMT Policies and Programs.” In addition to traveler response impacts there are public health effects, again, generally either pos- itive or neutral. Those effects are summarized separately, at the end of the “Public Health Issues and Relationships” subsection, located within the “Related Information and Impacts” section. That digest is titled “Adult and Child Public Health Relationships Summary.” Sidewalks and Along-Street Walking. Neighborhood sidewalks perform a “land service” func- tion, just as do local streets. As with local streets, their usage is often light but important to fronting dwellings. Illustrative pedestrian-volume intersection counts from three San Francisco Bay Area counties range from roughly 12,000 pedestrians/day at a location on one side of the San Francisco central business district (CBD), to some 2,200/day at a suburban intersection with low-rise apart- ments near ethnic gathering spots, to 300–350/day at an exurban intersection with a town hall and dwellings, and on down to 20–25/day at partially developed office/commercial intersections in suburban and exurban locations. The average individual sidewalk at these locations would be han- dling roughly 1/4 of these volumes assuming four-way intersections with sidewalks on all sides. Four quantified cases of neighborhood sidewalk improvement effects show pedestrian volume increases ranging from 46 percent to 400 percent, with a median increase—among case study averages—on the order of 60 percent. Not known from these before-and-after investigations is whether the added pedestrian volumes represent additional walking in the form of new walk trips, more frequent walk trips, or lengthened walk trips, or whether and to what extent the added vol- umes come from walk trips diverted from other routes or destinations. In one California five-site Safe Routes to School (SRTS) evaluation, however, additional walking activity by schoolchildren in response to new sidewalk construction was probed and explicitly demonstrated. Directness of sidewalk routing is sought by pedestrians. Local deviations producing as little as 12 percent extra walking distance have been observed to engender short-cutting by most pedestri- ans, while 6 percent local indirectness may be tolerable. At a larger scale, route directness has been shown to be an indicator of higher walking activity. Studies in Austin, Texas, found walking for its own sake (exercise/health, pleasure, and dog- walking) to vary least across neighborhood types, although traditional gridded neighborhoods did have the most. Walking for shopping was five times as prevalent in traditional neighborhoods as in late modern neighborhoods of roughly similar socio-economic makeup, with early modern 16-20

neighborhoods in between. Closeness of stores was a major factor. Quality of commercial area side- walk connections and presence of low traffic volumes and speeds were stronger indicators of walk- ing for shopping than residential street sidewalk completeness. At least for able-bodied adults, narrow and pleasant low-volume streets appear to have offered suitable compensation for lack of continuous sidewalks. In other cities, five out of six adult-focused cross-sectional walking studies found presence of neighborhood sidewalks or sidewalks in general to be positively related to walking activity, although typically not the strongest indicator. In one of these cases the relationship only held for recreational walking and in one case it explicitly held for utilitarian, recreational, and health- related walking. In addition, three U.S. child-oriented active transportation studies found walking to be positively related to sidewalk availability, while two Australian studies found walking by children to be negatively related to heavy or problematic traffic. A comparative analysis in the Seattle area, the Austin studies, and San Francisco Bay Area shop- ping district research, along with less formal reportings, all underscore the importance of customer- friendly commercial area sidewalk facilities. The Seattle area analysis examined 12 shopping districts and their surrounding neighborhoods, six suburban and six urban—controlled for density and land use mix—using 16-hour shopping area pedestrian cordon counts. The six sub- urban examples, with large blocks and averaging 8 miles total of discontinuous, incomplete side- walks, averaged just 12 pedestrians/hour per 1,000 residents. The six urban examples, with small blocks and averaging 38 miles of sidewalks, averaged 38 pedestrians/hour cordon line flows per 1,000 residents. It is well established that traffic calming reduces traffic crashes. Very limited evidence suggests it also encourages more walking and bicycling along the treated streets, with increases of 60 to 70 percent or so in NMT traffic observed. National surveys report that some 11 to 14 percent of bicycle trips are made mostly on sidewalks. A majority but not all of the limited available research suggests this is undesirable from an adult bicycle crash rate perspective. Two studied instances of attracting cyclists off of sidewalks and onto bike lanes suggest different diversion outcomes. In the central San Francisco example, bike lane installation caused sidewalk usage to drop from 52 percent of 71 2-hour PM peak cyclists to 7 percent of 94 cyclists. In a Fort Lauderdale beachfront installation of bike lanes, sidewalk usage stayed within a percentage point of 44 percent. Bicycle lanes may be less effective where there are fewer commuters and large pro- portions of less skilled cyclists and/or where the bike lanes are substandard. Public health physical activity research has directed extensive attention onto effects of sidewalk system availability, often finding significant and positive relationships with walking sufficiency, exercise, and normal body weight. Public health research—as noted in the introduction to this “Traveler Response Summary”—is summarized at the end of the “Public Health Issues and Relationships” subsection of the “Related Information and Impacts” section (see “Adult and Child Public Health Relationships Summary”). Street Crossings. Reports of pedestrian and bicyclist response to street crossing provisions pre- sent a mix of quantitative research and less formal reporting, but the findings overall largely demonstrate that provision of safe and attractive crossings is an essential element of having an attractive overall NMT system. A before-and-after study of 11 sites in four U.S. cities, all signed for 25 mph, found the proportion of pedestrians crossing at the crosswalk locations to have increased by a range of less than 1 percent up to 12 percent (city averages) in response to new crosswalks. There was also a statistically insignificant increase of somewhat less than 1 percent in pedestrians 16-21

overall. A larger study of marked crosswalks without traffic controls, in comparison with unmarked and uncontrolled intersection crossings, found the proportions of the young and elderly crossing four or more lanes who used the crosswalks to be 76 and 81 percent, respectively, as compared to 66 percent for all types of pedestrians. Unfortunately, the four-or-more lane uncontrolled cross- ings category represents precisely the type of facilities found to have higher crash rates within marked as compared to unmarked crosswalks at higher traffic volumes and speeds. Individual cases where walking volume changes were observed provide indication that an urban- design highway intersection with tight curb radii is more attractive to pedestrians than a rural design with free right turns and “pork-chop” corner islands, and that an urban traffic circle with comprehensive pedestrian provisions is more attractive than a complex conventional intersection. A systems approach to traffic signal placement as part of one-way street revisions in a London entertainment district, with attention to pedestrian desire lines and provision of an increased num- ber of signals along with selective sidewalk widening, was followed by a 9 percent increase in over- all pedestrian flow. Based on GPS tracking of bicyclists in Portland, Oregon, it was estimated—in the context of a 3.5 mile trip—that cyclists will deviate by 16.5 percent to avoid each unsignalized major arterial crossing. Comparable deviation values for other conflict situations include 2.5 percent per unsignal- ized minor arterial crossing, 11.5 percent for each unsignalized left turn from a major arterial, and 4.5 percent for each unsignalized left turn from a minor arterial. Modeling of surveyed shared use trail utilization in Arlington, Massachusetts, found perceived need to cross a busy arterial for trail access cut usage of the trail in half, but the effect could not be isolated for measured need. Three of four research efforts examining the effect on walking and bicycling to school of necessity to cross multiple, busy, or major roads found negative impacts. California Safe Routes to School studies found the travel choice effectiveness of intersection improvements to be only moderately less than the impact of paved sidewalk projects. Crossing signalization was the most effective inter- section treatment, with before-and-after child pedestrian counts showing a 24 percent increase in schoolchild usage of 2 intersections that had been newly signalized. Studies in England made under 1960s conditions found usage of pedestrian grade separations to be highly sensitive to pedestrian crossing time relative to at-grade alternatives. Virtually no one used an overcrossing requiring 25 to 50 percent more crossing time than the at-grade route. Given equal travel time via either the grade separation or an at-grade route, underpasses were found to be chosen by 95 percent of pedestrians, while overpasses were chosen by 20 to 70 percent. Pedestrian Zones, Malls, and Skywalks. Traditional CBD pedestrian streets (malls) have been greatly affected by long-term business activity trends, especially retail trends. Many in the United States were superimposed in the 1960s and 1970s on downtowns in decline. However meritorious, they were often unable to stem the tide toward suburban shopping. Loss of activity led to a deserted feeling, and many were removed or redesigned to reintroduce street traffic lanes and pro- vide a better balance of pedestrian space with pedestrian flows. In other cases, pedestrian streets have been and are highly successful. In the Downtown Crossing pedestrian zone in Boston, the volume of visitors to the area went up by about 10 percent over a 2-year period. Weekday mode shares for worker and shopper trips into and in the pedestrian zone shifted from 48 to 54 percent walk, 37 to 39 percent transit, and 11 to 6 percent auto. Transit malls have had a higher success rate in the United States than purely pedestrian malls, with four out of five completed transit malls covered in a 1970s study report still extant in the 21st Century. On the Nicollet Mall in Minneapolis, average 11- to 12-hour pedestrian counts per side 16-22

for the six blocks central to retail activity were 12,400 to 12,800 in 1958, well before the 1967 intro- duction of the transit mall, and 13,600 in 1973 after transit mall development. Introduction of par- allel skywalks, starting in the mid-1970s, reduced usage into the low 7,000s as measured in 1976 and 2002. The Nicollet Mall attracted 38 to 46 percent of the immediately parallel pedestrian flow on a September day in 2002. The block-wide corridor centered on the Nicollet Mall is estimated to have been attracting—as of 2002—an 11-hour pedestrian flow averaging 15,600 to 18,700 per side (skywalk traffic included), contrasted to the 12,400 average per side in 1958, an increase of some 25 to 50 percent. Comparable data is not available for other transit malls, but simulation-aided pedestrian estimates circa 1980 led to a conclusion that the Portland, Oregon, transit mall had focused pedestrian activ- ity on the mall area and nearby sections of cross-streets. Total mall pedestrian volumes were esti- mated to be 75 percent bus patrons at the time. In contrast, 16 percent of surveyed Nicollet mall pedestrians were headed to or from a bus stop. Two much newer major mall installations, on opposite sides of the Atlantic, utilize context-sensitive combinations of mall facility types. The Oxford, England, installation of June 1999 involves Cornmarket and two other streets, while the New York City trial—now assured permanency— involves Broadway through Midtown Manhattan. Oxford central area pedestrian flows increased 8-1/2 percent between 1998 and 2000, and Broadway pedestrian flows past Times and Herald Squares gained an average of 8-1/2 percent, roughly double the annual upward trend since 1999. No studies have been encountered that explicitly examine the relationship between presence or extent of skywalks or underground walkway systems and prevalence of walking. Skyway system bridge crossings in the three- by four-block core of the Minneapolis downtown have averaged about 10,000 per day from the 1970s to the present, with recent volumes on the remainder of the now-vast 82-bridge system averaging about 1/3 as much. Choice of Skyway over parallel cross- walks in Minneapolis and St. Paul ranged in 1975 from 46 percent in June to 68 percent in November, averaging 61 percent over 12 months. A parametric estimate based on these Minneapolis and St. Paul sidewalk versus Skyway choice differentials by season suggests that induced walk- ing may represent 9 to 30 percent of total annual observed Skyway traffic, with a maximum like- lihood estimate of 15 percent. At a time when the extent of the St. Paul Skyway system was four blocks north-south and east-west, the median CBD walk journey via sidewalks was found to be approximately 2-2/3 blocks, while the median for walks making use of the Skyway system was some 3-1/3 blocks. Bicycle Lanes and Routes. Bicycle lanes have been found to reduce cyclists concerns about con- flicts with traffic and to attract riders from nearby parallel roads, as well as potentially tapping latent demand. GPS route tracking studies in Portland, Oregon, indicate that the average cyclist making a utilitarian trip will go 31 percent out of their way to use a bike lane instead of having to ride in mixed traffic on a street with moderately heavy volumes. The corresponding value for bicy- cle boulevards was found to be 45 percent out of the way. (The research did not encompass cycle tracks.) The user makeup of bicycle lanes, as compared to other types of facilities, may possibly be tilted toward use by adults commuting to work. In-depth before-and-after evaluations in Davis, California, and Toronto, Ontario, suggest that the introduction of bike lanes on a single street or multiple streets results in increased cycling along those streets, but with a substantial portion of the increase attributable to shifts in route choice. The average weekday increase in counted bicyclists on streets receiving bicycle lanes—across four North American cities with usable before-and-after data—is 48 percent, with a range from 23 per- cent in downtown Toronto to 70 percent in San Francisco. The bicycle count on St. Kilda Road in 16-23

Melbourne, Australia, almost doubled in the first year after bicycle lane installation. After 10 years, however, it had increased by a factor of 12 (to 511 cyclists in the AM peak one hour). Two separate studies that examined commute travel mode shifts in response to bicycle lanes, in Minneapolis- St. Paul and Chicago, found average bicycle mode share increases of 64 percent and 91 percent, respectively, with bicycle lane introduction. The response in Chicago was almost certainly ampli- fied by publicity and bicycle parking enhancements. The response in Minneapolis-St. Paul repre- sented a 1.38 percentage point shift in corridor work-trip bicycle share.8 Availability of both weekday and weekend before-and-after data from Oriental Blvd. in Brooklyn indicates that the relative weekday impact of bicycle lane implementation was some 7 to 8 times the weekend increase in bicycling, skating, and scooter use. California SRTS studies found no sta- tistically significant evidence of an effect on bicycling to school with bicycle lane installation. These and a number of other bicycle lane and route studies with relevant data could be the beginning of a still very tentative thesis that bicycle lanes offer relatively little attraction for increased cycling at times or by groups likely to be characterized by presence of youngsters and high proportions of cyclists with modest skill levels. Four progressively comprehensive national-level studies using aggregate cross-sectional data sup- port existence of a strong correlation between bike lane mileage in a city and work-trip bicycling. A 90-city study, utilizing 2006–2008 journey-to-work travel data, has estimated a highly significant statistical relationship between commuter cycling and both bike lanes and bike paths. Typical of this study’s findings, one of the research models estimates 2.5 percent more bicycle commuters for each 10 percent more bike lanes and 2.6 percent more for each 10 percent more bike paths. In a Seattle area study that addressed cycling for all purposes, a perception of bike lane and/or trail presence was found to have a positive relationship to actual bicycling activity, but not objectively measured presence of a bike lane. Objectively measured closeness of an off-road trail proved sig- nificantly positive, however. Cycle tracks, now extensively deployed in the Netherlands and Denmark, have only a short his- tory in the United States. A before-and-after study in Copenhagen found 18 to 20 percent bicycle (and moped) count increases on streets with cycle track installations, as compared to 5 to 7 percent increases on streets with conventional bike lanes added. A comparative study in Montreal encom- passing 6 street pairings found 2-1/2 times the bicycle volumes on cycle tracks relative to mostly parallel streets with no bicycle facilities. Early results from Portland, Oregon, indicate high levels of preference for cycle track and buffered bike lane installations. In California, Palo Alto’s early development of a bicycle boulevard saw 85 to 97 percent bicycle count increases with 35 and 54 percent declines on nearby multi-lane streets. Bicycle volume increases on streets in Vancouver, British Columbia, converted to “Bikeways” (a.k.a., bicycle boule- vards), found weighted-average 2-to-5-year cycling increases per Bikeway of 76 percent, 272 per- cent, and approximately 333 percent, easily exceeding upward secular trends in cycling of roughly 18 percent per year. Surveyed residents fronting a Portland bicycle boulevard report bicycling rates markedly higher than average, and a variety of bicycle trip purposes. Analysis of travel mode shifts in response to implementing signed bicycle routes or “on-street bikeways” found an average increase in three cities of 20 percent in corridor bicycle commuting, with a range from −1 percent in Salt Lake City to +37 percent in Austin, Texas. 16-24 8 “Percentage point(s)” refers to an absolute difference in percentages, rather than a relative difference.

Shared Use, Off-Road Paths and Trails. The GPS route tracking studies in Portland indicate that the average cyclist making a non-recreational trip will go 55 percent out of their way to use an off- road trail instead of having to ride in mixed traffic on a street with moderately heavy volumes, or 26 percent out of their way even if the alternative is a quiet street. The Portland studies indicate a hierarchy wherein conventional bicycle lanes are preferred over all categories of undifferentiated streets except maybe quiet streets, bicycle boulevards are preferred over bike lanes, and off-road trails are preferred over bicycle boulevards, thus according off-road paths highest preference. Earlier studies produced inconsistent findings on bike lane versus shared use path cyclist prefer- ences. Various surveys have indicated willingness to incur extra travel in order to bicycle on off- road shared use paths, including 67 percent extra in the case of Minneapolis respondents bicycling for all purposes—including recreation and exercise—on weekends as well as weekdays. The most recent national-level study—making use of aggregate cross-sectional data for 90 out of the 100 largest U.S. cities—found bike path prevalence to be significantly related to commuter cycling, and with about the same degree of positive effect as bike-lane prevalence. Each 10 percent more bike path miles per 100,000 residents was associated with 2.6 percent more bicycle com- muters per 10,000 residents. Many shared use paths follow natural features and fail to offer direct commuting routes, and may thus appear less attractive as a facility type unless indirectness is explicitly taken into account. Shared use paths serve all NMT users, including pedestrians, joggers, and riders of various non-motorized wheeled vehicles representing all ages and a broad spectrum of capabilities. On six urban, suburban, and semi-rural trails in Indiana, trail users (as distin- guished from spot-count path traffic) ranged from 11 to 54 percent walkers, 5 to 20 percent run- ners, 23 to 77 percent cyclists, and 1 to 13 percent in-line skaters and others. In 2000, when the six studied Indiana trails were relatively new, they attracted August weekday volumes ranging from 170 (Pennsy Trail in Greenfield) to 1,620 (Monon Trail in Indianapolis), with corresponding weekend-day volumes of 190 to 2,350. On the 30-mile Interurban Trail of Ozaukee County, Wisconsin (north of Milwaukee), fully opened in September 2002, the 14-hour, 7-day trail traffic volume in August 2004 was 4,400 per week, averaged across 2 locations. A summer 2003 survey found 25 percent of respondents countywide to have made use of the trail. The correspond- ing rate for a March 2005 survey was 53 percent. The Seattle-area shared use Burke-Gilman and Sammamish River Trails were opened in the late 1970s, joined in 1993 into a 27-mile trail, and then gradually extended closer to downtown. Average 3-to-4-station weekday volume counts grew (not steadily) from 410 in 1980 to 2,190 in 1995, dropped to 1,690 in 2000, and rose again to just over 2,000 in 2005. Corresponding Saturday counts were 1,940, 3,640, 2,080, and 2,290. Possible reasons for the post-1995 drop include weather, reac- tion to trail crowding and high-speed cyclists, and increased recreational opportunity competition from newer trails. In 1985, 54 percent of survey respondents reported use of a car for trail access on weekdays, while the Saturday proportion was 59 percent. A 50 percent decline for Tuesday respondents from 1985 to 2000 in this proportion, and a 22 percent decline for Saturday respon- dents, meshes with the postulate that some earlier recreational users of the trails may have shifted to newly opened facilities. Weekday trip purposes evolved in a continuous shift from 10 percent work/school commute and 90 percent recreation/exercise in 1985 to 48 percent work/school and 45 percent recreation/exercise in 2000, with continuation of the shift toward commuting in 2005. Two individual studies that examined work commute travel mode shifts in response to imple- menting shared use off-road paths obtained a range of results. Bicycle commute share increases along four new trails in Minneapolis-St. Paul averaged 43 percent, starting with corridor shares that were already 4 to 5 times the norm, and represented a 1.38 percentage points gain. Off-road trail commutershed bike shares in Austin, TX, increased by an average of 0.88 percentage points, 16-25

up 24 percent. Work commute mode share outcomes for the off-road paths in the other cities were not statistically significant. The research examined neither walk trips nor non-work purpose trips, and in the cities without significant outcomes, the paths tended to either not be part of an overall bike facility network or else parallel to pre-existing facilities. A review of five other before-and- after studies of cycling activity by residents living nearby new or modified paths likewise pro- duced mixed results. No significant changes in levels of cycling were identified in two studies, one of which examined a 1-mile path, but two other studies did find more cycling, and in the 5th study there was an identifiable increase within 1 mile of the path but only after a promotional campaign. Among local or regional cross-sectional studies examining walking or bicycling activity levels vis- à-vis off-road path proximity or availability, four out of six established a positive relationship. The primary tributary area of an off-road trail has been estimated at 1/2 mile from the facility for commuter cyclists on the Burke-Gilman Trail, and appeared to be 1/2 to 3/4 mile to each side for all users of the Minuteman Trail in Arlington, Massachusetts. On the other hand, decay functions fitted to percentages by access distance of all-purpose cyclists in Minneapolis did not flatten out until about 3 miles from the trail. The functions for work/school and shopping trips dropped off more sharply than the function for recreational trips. Pedestrian/Bicycle Systems and Interconnections. In Portland, Oregon, bikeway system extent (bike lanes, bicycle boulevards, and off-road trails) was increased from 78 miles in 1991 to 256 miles in 2004, a 228 percent increase. Bike facilities were improved or added on four central area bridges. Extrapolation from bridge counts suggests a 210 percent increase in bike trips over the same time span. In the 1990–2000 decade, the citywide bike mode share for work purpose trips increased from about 1 to 3 percent, a comparable increase. In 2005–2008, bridge bicycle counts increased even more rapidly, despite a slowing in system expansion. Possible explanations include a lag effect, gas price increases, individualized marketing, and feedback effects. The bridge bicycle traffic counts from 1991 through 2008 exhibited an exponential growth rate of 9.6 percent per year, pro- ducing a fivefold increase overall, with 16,700 weekday four-bridge bicycle crossings in 2008. With development starting circa 1985, Brisbane, Australia’s shared use path system extended some 7-1/2 miles from the central business district (CBD) in one corridor by 1995 and in three corridors by 2000. A major new pedestrian-and-bicyclist-only bridge was opened in 2001. Walk to work shares for travel to the CBD and the CBD fringe increased almost threefold from 1986 to 2006, reaching 17.4 percent walk. (Housing expansion in the core may have contributed to the increased walking.) Bicycling shares, from within roughly 7-1/2 miles, increased sixfold, reaching 3.0 per- cent bike. Usage over a 2-week period of the new NMT bridge across the Brisbane River ranged, 5 months after opening, from 4,726 (25 percent cyclists) on a Saturday up to 10,854 (18 percent cyclists) on a Tuesday. The prior travel choices of weekday peak period bridge users, 8 months after opening, included 40 percent previously walking or bicycling via an upstream vehicular bridge less safe for NMT users. Prior modes, with multiple responses allowed, were 59 percent walk or bike, 45 percent bus, train, or ferry, 19 percent car, and 6 percent other. Many multi-modal trips were involved, both before and after, and the change from motorized modes was often for the innermost, cross-river leg of the motorized trip only. Example bike and walk bridge volumes range from 350 users a day on a former rail bridge between Lewiston and Auburn, Maine; to 2,120 pedestrians and 940 cyclists on the NMT-only Stone Arch Bridge in Minneapolis, a former rail bridge across the Mississippi River; 4,000 to 5,000 cyclists and walkers a day on a new NMT-only bridge over Town Lake in Austin, Texas, where the parallel highway bridge with 3-1/2 foot sidewalks was formerly used by only 700 to 1,000 per day; and nearly 2,400 pedestrians and roughly 4,400 cyclists per day on Burrard Bridge in Vancouver, British Columbia, after safety and traffic flow improvements to the highway bridge’s bicycle provisions. 16-26

A market survey focused primarily on adjacent neighborhoods found the Burrard Bridge improve- ments were accompanied by a 6 percent decline in reported cross-bridge walking versus a dou- bling of reported cross-bridge cycling. When the NMT-only Millennium Bridge in York, England, was opened in 2001, use of walking and cycling routes on both banks grew between 1999 and 2002 by 73 percent for walkers and 31 per- cent for cyclists, with some route expansion involved. Utilitarian trips went up by 141 percent, going from 25 to 38 percent of the NMT total. Surveys on the Brisbane NMT bridge in Australia and a new harbor-crossing bridge with path in Charleston, South Carolina, have found a majority of walk/bike bridge commuters to be combining intentional exercise with their commute trip. Closure of a 3-mile gap between Seattle’s Burke-Gilman Trail and the Sammamish River Trail saw 1990–1994 before-and-after weekday bicycle count increases of 84 percent at the closest-in end of the gap and 227 percent at the other end. Corresponding weekday pedestrian count changes were −19 percent and +163 percent, respectively. The pedestrian count outcomes seem to reflect an over- all increase combined with redistribution to the new and to the previously less accessible sections of the combined trails. Connection across a missing link in Brisbane’s Centenary Bikeway was accompanied by a 142 percent weekday (164 percent weekend) 2006–2007 cycle traffic increase at the nearest count station, and corresponding 54 percent and 59 percent increases calculated as weighted averages for the 3 count stations within 4 kilometers to each side of the former gap. Post- 2007 annual increases were substantially higher than the almost negligible increases prior to the late 2006 interconnection. Scattered evidence, mostly circumstantial, gives indication that smaller-scale neighborhood and facility linkages are also of substantial importance. Analysis of Seattle neighborhoods found that when neighborhood pedestrian connectivity and vehicular connectivity were about the same, the walk mode share averaged 14 percent. Where pedestrian connectivity was inferior, the walk share was 10 percent, and where pedestrian connectivity exhibited greater directness than vehicular con- nectivity, the walk share averaged 18 percent. Pedestrian/Bicycle Linkages with Transit. Of person trips in the United States, 1.7 percent in 2001 and 2009 involved walking to and/or from bus or rail transit service. This amount represents 16 percent (almost 1 in 6) of all walk trips. Walking is an essential component of all but a small frac- tion of transit trips. Surveys of transit riders have consistently shown transit mode share and walk to transit share to each have a strong inverse relationship to the distance from the stop or station. Rules of thumb suggest 1/4 mile is the outer limit within which most people are willing to walk to a bus stop and 1/2 mile is the rough equivalent for rail transit. Survey-based studies have shown these rules to be reasonable but not quite comparable. Research in the Miami-Dade area of Florida and Orange County, California, has found a 1/4 mile distance to encompass 80 to 90 percent of persons walking to bus service, while 1/2 mile is barely past the median walking distance to stud- ied West Coast rail transit stations. The ultimate in deliberate placement of dwellings close to transit stops and stations is found in good Transit Oriented Development (TOD) design. It has been amply demonstrated that transit prime mode shares in TODs and most transit-adjacent developments exceed the transit shares found in nearby non-TOD areas. Walk mode-of-access shares in TODs are mostly within the 70 to 100 percent range, with 90 to 100 percent most common. There are essentially no reported empirical attempts to isolate changes in transit use or walk or bike mode of access in response to transit access improvements, except in the rather obvious (but important) case where breaching an access barrier has extended a station’s tributary area. One may 16-27

either infer effect from the strong inverse relationship to access distance or model the effect using behavioral models developed for the purpose. Orange County, California, found that when a res- idential area was 80 to 100 percent within 1/4 mile of a bus stop, as measured along roadway cen- terlines, the work commute bus mode share was 7.9 percent. This dropped to 3.1 percent when the coverage was 40 to 80 percent, and 0.5 percent when the coverage was 10 to 20 percent. The researchers concluded that providing better pedestrian connectivity would increase coverage and thus bus shares. Application of access-choice behavioral models produced estimates that walking to Chicago- region Metra commuter rail stations could be increased by 7.2 percent if pedestrian system improvements were implemented. It was also estimated that bicycle parking and access enhance- ments would raise the Metra bike access mode share from 2.1 to 3.2 percent for home-based trips originating within 2 miles of the station, with one-half the added bicycle access coming from the drive-and-park mode. More dramatic increases were estimated for bicycling to Chicago Transit Authority subway/elevated stations, but with some 80 percent of the shift coming from the walk access mode. Revealed preference modeling has identified relatively limited importance for socio-economic characteristics in choice of access mode to rail transit, except as expressed in auto availability, which diminishes likelihood of choosing either walk or bike access. Larger numbers of auto park- and-ride spaces at stations have also been shown to dampen walking. Walk mode-of-access shares have been found to universally decline with distance, and except for blocks particularly close to stops and stations, the same pattern is seen for bike access. One or two built-environment mea- sures of neighborhood pedestrian/bicycle friendliness have shown significance for increasing walk access share in 4 out of 5 walk mode of access models. Nearby presence of streets with higher posted traffic speeds entered one bike access model (and its companion walk access model) as a negative factor. Transit access survey results from 14 U.S. cities and national transit share data suggest that bicy- cle trips taken in conjunction with transit use constituted, circa the year 2000, on the order of 1/10 of 1 percent of all trips taken in the United States. That would be roughly 1 for every 10 bike-only trips. Relatively little empirical study has been done on this aspect of travel. A stated preference experiment focusing on transit access identified bike lockers as a significant incentive to bike-and- ride instead of driving to transit or all the way. Lockable covered parking was 40 percent as effec- tive. Relative to bike lanes, lockers were 3 times more important for frequent cyclists, but slightly less important for infrequent cyclists. Bus and rail systems that offer full-scale bike-on-transit programs have found that 3/4 or more of riders arriving by bicycle are taking advantage of the bike-on-transit service. The Phoenix, Arizona, area bike-on-bus program is in this category. It started with a 1991 demonstration, by FY 2000–2001 already served 2,400 weekday bike boardings (a 1.9 percent share of annual passenger boardings), and in FY 2008–2009 served over 4,600 weekday bike boardings (a 2.2 percent annual share). The median share for U.S. bike-on-transit programs circa 2000 was 0.7 percent of passenger boardings. Among the most heavily used U.S. bike-on-transit services is that of Caltrain commuter rail on the San Francisco Peninsula, reflecting a need for many of its commuters to reach Silicon Valley jobs not within easy walking distance of suburban stations. In a 2007 AM peak period, 924 cyclists boarded with their bikes, 7 percent of all Caltrain passengers. Bike-on-transit, aside from a higher- income component on urban rail systems, is a mobility option heavily used by the transportation disadvantaged, including students. In a survey covering three Florida “Bikes-on-Bus” programs, with 0.25 to 1.61 percent bike-on-bus shares, 78 percent of users reported annual incomes below 16-28

$30,000 in 2004 dollars. The median access distance via bike was 1 mile, while the median egress distance by bike was 1/4 mile. Point-of-Destination Facilities. Point-of-destination facilities are provided at a workplace, school, shopping area, or other attraction to make it more feasible or easier to use non-motorized trans- portation. The obvious example is bicycle parking. Quantitative empirical data on impacts is extremely limited. Nevertheless, an overall importance of destination facilities for engendering more utilitarian bicycling—and also walking—is apparent. When Portland, Oregon, created four “Bike Central” locations offering showers, changing facili- ties, and bicycle storage for a modest fee, a before-and-after study found users of the service increased their average frequency of commuting by bicycle from 3.1 days per month before to 15.5 after. They drove, or rode transit, less. “Bike stations” perform a similar function. A 2009 survey covered eight stations in seven cities, ranging in capacity from 40 bikes (Auckland) to 300 (Chicago). Average percent occupancy, where known, ranged from 28 percent (Seattle) to 88 per- cent (San Francisco, Caltrain terminal). A Riverside, California, company with about 650 employ- ees, subject to a trip reduction ordinance, installed bike lockers, provided changing facilities, offered access to tools for cycle repairs, and also offered financial incentives to bicycle commuters worth about $2.00 per day cycled. The 10 percent bicycle commute mode share achieved was 10 times the regional average. Research utilizing stated preference experiments offers additional insight. One such study, based in Edmonton, Canada, estimated large effects for secure bicycle parking provisions (equivalent to a reduction of en route cycling time of 27 minutes) and smaller effects for showers (equivalent to 4 minutes). A study combining U.K. National Travel Survey and stated preference data estimated that with a starting workplace commute trip bicycle mode share of 5.8 percent, the bike share would increase to 6.3 percent with outdoor bike parking, 6.6 percent with indoor secure parking, and 7.1 percent with that plus showers. An empirical study of workplace destination amenity effects on combined walk and bike work trip mode shares found that measures such as high walking accessibility to convenience services, high appearance of safety around the workplace, and high workplace and vicinity aesthetic appeal, were each associated with NMT work trip shares higher by 0.7 to 1.5 percentage points. This was an environment where observed overall NMT mode share averages were within or close to the range of 2 to 4 percent. San Francisco travel demand modeling found an urban vitality measure to be, for both work and other trip purposes, an indicator of higher mode shares for walking, walk- transit combinations, and (“other” trip purposes only) bicycling. Bikesharing, involving the shared use of a publicly available bicycle fleet, may be considered both an origin and a destination facility and service. Implementation of this relatively new development appears to be following a typical technology adoption curve and is presently in the “innovators” or “early adopters” phase of market penetration. Estimates of impact vary widely, ranging from 44 percent cycling increases (Lyon, France, first year) to 70 percent (Paris, France) and even a tripling where initial shares were small (Barcelona, Spain, first year). These were major, compre- hensive programs, and it is not altogether clear to what area coverage the mode shift reports apply. In Minneapolis, 1/3 of first-year subscribers previously rode a bicycle less than once a month. If the service had not been available, 46 percent would have walked or used their own bike, 20 per- cent would have used transit, 19 percent would have driven, 6 percent would have ridden in an auto or taxi or made other arrangements, and 9 percent would not have made the trip. About 1/3 used the bikesharing service to access public transit, the same order of magnitude (in terms of pro- portion) as first- and second-year reports for bikesharing in Paris. 16-29

Pedestrian/Bicycle Friendly Neighborhoods. Some of the more notable travel impacts of urban land use structure and design relate to their effect on use of active transportation, specifically, the decision to walk, to cycle, or to do one or both in conjunction with taking public transportation. A meta-analysis of over 50 studies found that the built environment descriptors most closely related to walking were intersection and street density (measures of connectivity), land use diversity, and local access to jobs. Land use mix, neighborhood design measures, and distance to a transit stop were important to transit use. Elasticities were derived, and all were in the lower inelastic range when each was examined in isolation. However, the combined effect on active transportation use of an array of supportive built environment characteristics could be quite large. (For a brief expla- nation of elasticities see Footnote 12 in the “Response by Type of NMT Strategy” section—“Street Crossings” subsection. For a full explanation, see Appendix A in Chapter 1, “Introduction.”) Other research has found greater density, higher mix of land use, aesthetics, street connectivity, enhanced accessibility or proximity, traditional neighborhood design, and related infrastructure and conditions such as sidewalks and safety to be positively correlated with walking or with both walking and bicycling activity. For children the list is shorter, with distance to school critical. Closeness of schools to homes correlates significantly with walking and cycling to school, with some 20 studies finding distance to school inversely related to choice of active transportation for school access. Overall, the relationships of neighborhood characteristics to walking and bicycling to school appear to be either logical, or insignificant, while generally weaker (distance to schools excepted) than those for adults. Partially conflicting studies have, on balance, found street connec- tivity and destination proximity to be positively related to physical activity of children. Disaggregate studies that separately account for intensity of transit service and other density- related parameters find a weak association between density of development in and of itself and propensity to walk or ride transit. This effect may be taken to imply that when other factors are ana- lytically controlled for, the presence of more residents or jobs per unit area only slightly increases walk and transit mode shares. However, and importantly so, density produces conditions that are themselves strongly conducive to use of active transportation. Suitably organized dense develop- ment leads to and supports higher levels of transit service, brings activities into closer proximity, and fosters land values that induce priced parking, all characteristics that lead to additional walk- ing, bicycling, transit use, and lower vehicle miles of travel (VMT). Low walk and transit use elas- ticities for density do show that density not well integrated into the urban fabric, such as apartments in the middle of auto-oriented suburban sprawl, will not have large beneficial effects on either walk- ing or bicycling for transportation or on transit use. Diversity and design are both more strongly related to prevalence and mode choice of walk trips than density per se. Where there are more local opportunities to meet daily needs there will typi- cally be more walking, with the relationship strongest and more often identifiable in the case of utilitarian walking. There is a general lack of consistent evidence that destination proximity is asso- ciated with recreational walking. Choice of walk versus auto access to transit is an aspect of travel behavior particularly sensitive to land use characteristics, and it has been found to be highly responsive to land use mix in particular. Positive land use mix elasticities for walking to transit have been estimated to lie in the elastic range, at +1.1, very sensitive. Bicycling appears to be an individual choice only moderately associated with the local land use and design environment, although shared use trail proximity and certain commercial use group- ings had significant positive relationships in Seattle area research. Companion studies found the neighborhood environmental measures most related to walking to be closeness to grocery stores, restaurants, and retail; lack of office building dominance; and density of the individual’s home par- cel. Numerous studies support the importance of proximity of retail stores to higher rates of util- 16-30

itarian walking, for example, walk trips were determined to be more than twice as likely for Minneapolis-St. Paul households less than 1/8 of a mile from the nearest retail as compared to those greater than 3/8 of a mile from retail. Various neighborhood walkability scores have been shown in the Puget Sound Area and in Canada to be positively related to walking activity, includ- ing walk mode shifts observed in longitudinal panel survey observations of response to changed residential location and neighborhood environment. NMT Policies and Programs. Major exemplary illustrations of translating policy into city-wide bicycle or NMT programs in the United States are provided by Portland, Oregon; Davis, California; and Boulder, Colorado. Brisbane, Australia, provides an additional “new world” example. The NMT policies and programs in Portland, Oregon, have been heavily focused on bicycles, although pedestrians have benefited. The City of Portland has since the mid-1970s pursued policies designed to reduce auto use, particularly in the central area. Portland Bicycle Master Plan implemen- tation did not move full steam ahead, however, until the 1990s. Results, including a 10-year tripling of work trip bicycle shares and a 17-year quintupling of central-area bridge crossings, were described previously under “Pedestrian/Bicycle Systems and Interconnections.” Also focused heavily on bicycling have been the NMT policies and overall program in Davis, California, a university town long known as the U.S. bicycle capital. Today Davis has systems of bicycle lanes and separated shared use paths that total close to 50 miles each, in an area of 10 square miles. Davis provides a unique sequence of lessons. The bicycle program reached its zenith in the latter half of the 20th Century, and now effects of program maturity and even decline may be observed, as citizen involvement and numerous university support programs have withered and/or disappeared. Increased in-commuting is also a factor. The circa 1970 Davis bicycle mode share for trips to work neared or may have exceeded 30 percent, but stood at 14 percent in the year 2000. The student share for commuting to campus approached or reached 80 percent circa 1970. Student bicycling shares to campus were 48 percent in 2007, but with much of the drop having been taken up by use of free bus service. Boulder, Colorado, is also a university town but, as a suburb of Denver, it serves as well as a home for major employers. The policy and program focus differs from Portland and Davis in that it addresses in one Transportation Master Plan (TMP) goal the enhancement of all active transporta- tion modes: pedestrian, bicycle, and transit. Off-road shared use paths and pedestrian/bicycle undercrossings of highways have been major NMT development program components. Travel mode shares of residents shifted, between 1990 and 2006, from 18.2 to 18.9 percent walk, 9.1 to 13.6 percent bike, and 1.6 to 4.0 percent transit, for an overall growth of 26 percent in the active transportation share. For employees working in the city, the active transportation share for com- mute trips has increased by 16 percent and for midday-trips by 30 percent. In Brisbane, Australia, the “Brisbane Active Transportation Strategy” derives from national pedestrian and cycling strategies. Relevant components of Brisbane’s strategy, together with state government agency internal initiatives preceding strategy and master plan adoption, have under- girded metropolitan area NMT facility investment initiatives. Again, results were described in the “Pedestrian/Bicycle Systems and Interconnections” summary. Urban area walk trip mode shares for work trips to the CBD and its fringe have close to tripled between 1986 and 2006, while corre- sponding bicycle shares have sextupled. Comparisons with European programs raise issues of transferability to American situations, but are nevertheless instructive. Most north-central European countries reported circa 1995 walk and bike mode shares that, combined, were 5 to 6 times higher than found for 1995 in the United States, 16-31

including walk shares 3 to 5 times higher and bicycle shares 10 to over 25 times higher. North- central European NMT shares reflect major gains from low points set in the 1970s, with reversals of earlier declines correlating well with shifts in policy and funding toward substantial support for walking and bicycling. Perhaps the most direct countering of arguments that higher bicycling shares in north-central Europe relative to the United States are primarily attributable to higher urban densities and correspondingly shorter trips is provided by comparison of mode shares strat- ified by trip length. For example, bicycle shares for trips 1-1/2 miles or less in length during the 2000–2005 period were 2 percent in the United States (and United Kingdom), 14 percent in Germany, 27 percent in Denmark, and 37 percent in the Netherlands. (An unstudied aspect of this comparison is that U.S. trip tour lengths may be longer, even when individual trips are short, with corresponding mode choice effects.) The dominant school-focused active-transportation program effort in the United States is the Safe Routes to School (SRTS) program. Quantified results for infrastructure improvement approaches are limited and mainly from California. Counts taken before and after sidewalk improvements on the approaches to California elementary schools showed a weighted-average five-site 46 percent increase in schoolchild walking. Similar counts at intersection signalization projects indicated a weighted-average two-site 24 percent increase. Results for other crossing improvements were inconclusive. Other SRTS approaches involve various forms of encouragement. An intensive pilot program of outreach, encouragement, and aid to schools and parents in Marin County, California, is reported to have resulted in a 21-month, 64 percent increase in walking at surveyed schools and more than a doubling of bicycling. (Results exhibited some anomalies and 12-month results averaging a 17 percent increase for walking and 54 percent for cycling offer a more conservative perspective.) A walking increase of 6 to 12 percent was achieved with a possibly less-intensive pilot program in three Arlington, Massachusetts, schools. Two programs in England that depended mostly on coordination and encouragement proved inef- fective. Daily tracking of student walking and cycling, with recognition and perhaps awards, appears to produce results. A school in Brampton, Ontario, Canada, achieved a 1/4 reduction in auto drop-offs in this manner and a primary school in Dorset, England, obtained a 16 percent increase in walking/cycling rates. A Boulder, Colorado, elementary school more than doubled walking and bicycling and reduced school-area traffic by 36 percent with a trip-tracking challenge combined with walking school buses (WSBs) and other actions. Research in Dorchester County, England, estimated a 26 percent shift from auto use among case study WSB participants, while a Seattle inner-city school achieved a school-wide 37 percent increase in walking with three WSBs. Participant turnover proved a major consideration for Dorchester WSBs and Nelson, New Zealand, cycle trains (CTs) as students “graduated” from supervised walking in WSBs to supervised cycling in CTs, and from both to independent walking and cycling. Walking/Bicycling Promotion and Information. Results of a broad mass-market walking and cycling information and promotion program in England were inconclusive, similar to the outcome of a larger number of studied public-transit mass-marketing efforts in North America. However, “Ride to Work Day” events may attract the uninitiated: In Melbourne, Australia, high first-time female cyclist participation was achieved and over 1/4 of first-time riders reported riding to work at least once in the course of a survey week 5 months later. One group-targeted information and promotion program achieved approximately a doubling in walk trips, while another increased time spent walking by 64 minutes per week, both as measured against control groups in the short term. Information and promotion marketing focused on cycling via a new NMT facility, an obscure rail-trail in Western Sydney, Australia, was accompanied by a significant increase in overall 16-32

cycling (from 17 to 28 minutes per week) by persons living nearby who already had a bike avail- able. In transit marketing, similar programs offer only very limited evidence of longer-term gains. “Individualized marketing” delivers information on environmentally friendly travel modes tailored to needs of individual participants willing to receive project outreach. Household-based dialogue marketing applications encompassing personalized encouragement and addressing all purposes of travel are the most common. Evidence from England and Portland, Oregon, indicates that the travel mode shifts encouraged are in fact larger for discretionary and other non-work travel than for commute trips. Statistics from an audited large-scale application in South Perth, Australia, provide a representative scaling of target area population involvement. Among some 18,600 target- area households, 72 percent proved possible to contact by telephone and also agreed to the initial interview. Of households thus interviewed, 5 percent were regular users of environmentally friendly modes with no need for further information and another 12 percent were regular users desirous of additional walk, bike, or transit facility/service guidance. Regular users received small rewards as encouragement along with any information requested. All other “interested” house- holds accounted for 46 percent. They received targeted information and incentives to try shifting modes. That left 37 percent of interviewed households that were not interested. Representative target area results based on averages for projects in six sectors of Perth, Australia, the combined outcomes of the Federal Transit Administration’s four-city Individualized Marketing Demonstration Program (IMDP) projects in the United States, and an average of two of the annual programs in sectors of Portland, show a range of 1 to 4 percentage points gain in walk trip mode share. They also show a 1 to 2 percentage points gain in bicycle share and also in transit share. Corresponding auto driver or drive-alone mode share declines are in the 3 to 6 percentage points range. Overall average U.K. results are encompassed by these same ranges. The mode shifts obtained in the IMDP projects translate into relative gains of 20 to 25 percent, in their U.S. context, for each of walk, bike, and transit use. Australian studies have determined that individualized marketing effects tend to be greater for large-scale programs involving more than 5,000 households per application. Small-sample studies of programs in cities with poor transit service and non-motorized transportation facilities suggest that impact may be reduced by half in such circumstances, but contrary results have been seen. Nine projects done in conjunction with rail transit improvements in Portland (1 project) and Germany (8 projects) have averaged first-year transit ridership gains of 48 percent in individual- ized marketing target areas as compared to about one-half that in control areas receiving only the transit improvements. A number of long-term surveys across three continents have found substan- tial retention of mode shifts after 1 to 4 years, indicating durability of impact. Despite some controversy, the national government transportation agencies in the United Kingdom and Australia have each concluded that the benefits of household-based individualized marketing programs are sound and cost-effective. Employer-based and school-based individualized marketing has been less studied and so far shows less promise than the household-based approach. Assessments of U.S. programs with expanded information and activity menus have not allowed robust con- clusions on effectiveness of the add-ons. Nevertheless, an augmented large-scale program in Bellingham, Washington, is notable for obtaining absolute shifts of +4 percentage points for walking, +3 percentage points for cycling, and −6 percentage points for driving alone, paired with a slight tran- sit mode share increase and a −1 percentage point auto passenger decrease. External evidence in the form of pedestrian and bicycle count and transit ridership data lend sup- port to the significance of household-based individualized marketing outcomes. The 2005 project in Portland estimated a 7 percent relative gain in walk mode share based on survey analysis and 16-33

Next: Response by Type of NMT Strategy »
Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities Get This Book
×
 Traveler Response to Transportation System Changes Handbook, Third Edition: Chapter 16, Pedestrian and Bicycle Facilities
Buy Paperback | $98.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Transit Cooperative Research Program (TCRP) Report 95: Traveler Response to Transportation System Changes Handbook, Third Edition; Chapter 16, Pedestrian and Bicycle Facilities examines pedestrian and bicyclist behavior and travel demand outcomes in a relatively broad sense.

The report covers traveler response to non-motorized transportation (NMT) facilities both in isolation and as part of the total urban fabric, along with the effects of associated programs and promotion. The report looks not only at transportation outcomes, but also recreational and public health outcomes.

TCRP Report 95, Chapter 16 focuses on the travel behavior and public health implications of pedestrian/bicycle area-wide systems; NMT-link facilities such as sidewalks, bicycle lanes, and on-transit accommodation of bicycles; and node-specific facilities such as street-crossing treatments, bicycle parking, and showers.

The report also includes discussion of the implications of pedestrian and bicycle “friendly” neighborhoods, policies, programs, and promotion.

The report is complemented by illustrative photographs provided as a “Photo Gallery” at the conclusion of the report. In addition, PowerPoint slides of the photographs are available for download..

The Traveler Response to Transportation System Changes Handbook consists of these Chapter 1 introductory materials and 15 stand-alone published topic area chapters. Each topic area chapter provides traveler response findings including supportive information and interpretation, and also includes case studies and a bibliography consisting of the references utilized as sources.

READ FREE ONLINE

  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!