Meta-Analysis of Elasticity of Travel Behavior (Mode Choice) Regarding to the Urban Form

Modern world faces a multitude of challenges regarding to displacement including high fuel consumption, air pollution, congestion in streets, decrease in physical activities, insecurity of passers-by and etc. Most of them are resulted by automobiles and vehicle-stricken urban forms. Planners and public health officials try more to develop land planning and urban designing for reducing automobile use and social-environmental costs. A growing number of empirical studies have investigated the relationship between the built environment and pedestrian behaviors. Most of them are based on the assumption that travel is a derived demand, or that the travel demand is derived from the demand for activities and movements. In other words, they assume that individuals travel in order to reach a destination or participate in an activity; they do not travel simply for the sake of travel. According to derived demand theory, these studies provided evidence of a correlation between the built environment and pedestrian behaviors. Many studies found that residents living in traditional neighborhoods (characterized as high density, high accessibility, mixed land use, rectilinear street network, and so on) would drive less and walk more than those living in suburban neighborhoods. Also, several studies showed that there are not significant relationship between urban form and travel behavior. The results of such studies on travel demand have been summarized in recent years. Many individual studies have been carried out in various countries with different objectives and methods in order to analyze urban relationships and travel behavior (type of vehicle), which have achieved various results and which are not generalizable. This paper aims to generalize the result of previous studies using meta-analysis approach. The present study has initially analyzed the dimensions of the study with a detailed review of empirical studies using a descriptive-analytical method to organize a summary of any literature to change them into a typology, but there is no best rationale for doing, yet. These studies can be usefully organized in any number of ways, for example, by travel purpose (joumey-to-work travel vs. shopping vs. trip chains, etc.), by analytical method (simulations vs. regressions, etc.), by the characterization and measures of urban form (trip ease vs. street layout vs. composite measures of density, accessibility, or pedestrian features, etc.), by the choice of other explanatory variables (travel costs vs. travel opportunities vs. characteristics of the built environment or of travelers, etc.), or by the nature and level of detail in the data. This process would be useful to choose appropriate studies as sample case. In fact, in order to combine the result of studies, these studies should have several things in common. As they analyze effects of the built environment on travel choices, all these studies are controlled statistically for confounding influences on travel behavior and sociodemographic influences in particular. They use different statistical methods because the outcome variables differ from study to study. Almost all them are based on sizeable samples and have been carried out with a disaggregate approach. After choosing the study analysis model, composition and generalization of the results of study has been conducted in the area of city form and travel behavior by using a meta-analysis method and by calculating Elasticities (amount of sensitivity). Elasticities were obtained from the individual studies in our sample case in different ways including copping them from published studies where they were reported explicitly, calculating them from regression coefficients and the mean dependent and independent values. Most commonly, they were obtained directly from the studies of Ewing and Cervero (2001, 2010). Elasticities have been computed for individual studies and pooled them to produce weighted averages. In addition; in the studies which Elasticity was not available, significant sign of regression has been used. The most important findings of the study indicate that there are variety relationships between urban form variable and mode choice. As mentioned, density, diversity and design are three components of urban form. For every component some indicators have been defined. Among different indicators, “population density”, “distance to shopping centers” and “intersection density” have more positive and generalizable effects on walking or cycling. Distance to shopping centers has the strongest relationship. Also block size has a negative effect on walking or cycling. “Population density”, “percentage of intersections on crossroads” and also “distance to nearest station” have more effects on using public transportation. Percentage of intersections on crossroads has different effect on walking or cycling and using public transportation. It influences on walking or cycling negatively but influences positively on using public transportation. Moreover, it can be said that “population density”, “gross residential and employment density” and “land use mix” have significantly negative effects on using automobiles (p<.001).