Transit-oriented development and the frequency of modal use
Journal of Transport and Land Use
Transit-oriented development (TOD) is assumed to be effective at changing the travel behavior of individuals, particularly in increasing the use of public transit, but also by increasing walking activity while decreasing driving. The analysis presented here examines the frequency of using these modes of travel for those living in TODs and proximate to a train station versus those living more distantly. Household survey data was collected for the area surrounding eight rail stations in the state
... of New Jersey in the United States. The models developed include factors that control for attitudes about the neighborhood where respondents reside and how long they have lived at their current residence. Other control variables to represent the urban design of the neighborhood are also included. A factor analysis of the attitudinal variables produces five factors that are used as controls in an ordered structural equation model of frequency of using public transit, walking, and driving. Results suggest that those living in TODs and closer to the train station use public transit and walk more frequently than those living farther out; they also drive less frequently than those living farther out. 22 JOURNAL OF TRANSPORT AND LAND USE 8.2 this approach. SEM models allow for the simultaneous control of various interactions between variables and are considered to show causal effects. We begin with a brief literature review that focuses on studies that have analyzed TODs and travel behavior. While we recognize the large literature on associations between travel behavior and urban design, density, and diversity, we mainly cover the relevant literature on TODs. Our data collection process and strategy is then presented followed by our data analysis, beginning with our theoretical framework and a discussion of our attitudinal variables, followed by the SEM analysis. Literature review The literature reports considerable evidence that TOD residents drive less than those living in nontransit-supportive locations Cervero 2010, Arrington et al. 2008). The density, mix of uses, and relative accessibility allows residents as well as workers and visitors to fulfill at least some of their daily needs without driving. Cervero (2004) found evidence that many TOD ridership gains were a result of self-selectionindividuals who choose a residential location in order to be close to transit. Analyzing travel diaries from California residents, it was found that nearly 20 percent of those who lived within 0.5 miles of transit used transit to travel to work, compared to less than 9 percent of those living more than 0.5 miles from a station (Cervero 2004) . A key question is whether those who desire to drive less choose to live in transit-accessible locations. If this is the case, then this self-selection into certain neighborhoods could potentially bias statistical results. However, recent research that has sought to control for self-selection bias has found that it plays a small role and that statistical associations are still substantial (Cao et al. 2009b ). Thus, despite any self-selection bias, TODs can still lead to net reductions in vehicle travel. Chatman (2009) suggests that those households that value accessibility most may be most set in their ways; that is, they take transit already, while those who value accessibility least are more likely to change their behavior if they live in a TOD. Thus, self-selection could lead to underestimates of the impact of TODs on travel (Chatman 2009). One potential impact on travel behavior is the effect of TODs on vehicle miles traveled (VMT). Reductions in VMT result in lower energy demands as well as reduced demands on the road network. Lower energy demands have economic and environmental benefits while reduced demands on the road network can result in lower long-term investment requirements (Cambridge Systematics Inc. 2006). Cervero and Arrington (2008) found that residents of 17 TODs averaged 44 percent fewer vehicle trips than estimates based on the trip generation manual published by the Institute of Transportation Engineers. The authors noted that vehicle trip rates were particularly low in Washington, DC, and Portland, Oregon, where regional and corridor planning for TOD has been adopted, and that neighborhood densities and vehicle trip rates were inversely proportional. That is, as densities rise, trip rates fell. These findings suggest that attempts to model the effects of TODs must account for regional planning practices and neighborhood land-use characteristics . Looking only at commute trips, Arrington et al. (2008) found that residents living near transit were five to six times more likely to commute by transit than others living in the region . While residential density is a key feature of TODs, the mix of land uses within TODs can also reduce vehicle travel. Bartholomew and Ewing (2010) noted that increases in proximity and convenience allowed TOD residents, workers, and visitors to complete their daily activities within a smaller area, which results in shorter travel distances, lower average vehicle trip rates, and lower VMT (Bartholomew and Ewing 2010, Bartholomew and Ewing 2011). A recent review disentangled the effects of the "5 Ds" associated with smart growth-density, diversity, design, destination accessibility, and distance to transitnoting that diversity tends to lead to more walking than most other effects, with the exception of design features (Ewing and Cervero 2010).