A Time Geography Approach to Understanding the Impact of Gasoline Price Changes on Traffic Safety

The impact of gasoline price changes on traffic safety has received increasing attention in empirical studies. However, this important relationship has not been explained within a conceptual or theoretical framework. In this study, we examine this relationship within a time geography framework in an attempt to understand the effect of time-varying fluctuations in gasoline prices and their relationship to traffic safety in a case study of Mississippi from April 2004 to December 2008. We further extend this work by examining the degree to which this relationship is differential in impact by age, gender, and race. The results suggest that changes in gasoline prices have immediate effects on reducing total traffic crashes and crashes of younger drivers, women, and whites. However, changes in gasoline prices do not affect total crashes of older drivers, men, or blacks. Within the theoretical framework of time geography, we understand gasoline prices as one type of capability constraint of the space-time path and spacetime prism. As gasoline prices increase (that is, as the capability constraint becomes stronger), traffic crash rates will decrease. However, the effects vary by age, gender, and race because the capability constraint of gasoline prices differs across demographic groups. INTRODUCTION Recent surges in gasoline prices have resulted in complex social and spatial implications for individuals, the public, transportation and land use planners, and decision makers. Populations that are more vulnerable to gasoline price increases may drive fewer miles and reduce gasoline consumption by reducing driving frequency and distance, shifting from personal vehicles to public transport, driving in a more fuel-efficient manner (such as driving more slowly and reducing sudden speeding and braking), switching to fuel-efficient vehicles, and relocating closer to workplaces. These changes in commuting behaviors and residential location could eventually lead to improved traffic safety. Although a large body of literature has empirically examined the impact of gasoline price changes on traffic safety (see Chi et al. 2010 for a review of the literature), a conceptual or theoretical framework has not been offered to explain the impact. In this study, we attempt to use time geography theory (Hägerstrand 1970) to understand the impact of gasoline price changes on traffic safety. Time geography theory has been utilized to study various perspectives of transportation accessibility, such as bus services (Lenntorp 1976), commute modes and times (Burns 1979), the transportation geographic information system (Miller 1991), disparities in gender accessibility (Kwan 1998), and information and communication technologies (Yu and Shaw 2007). It has also proven to be an important organizing principle for modeling travel behavior as a set of linked activity and travel decisions subject to constraints on resources such as time (Kitamura and Fujii 1996). We assert that the theory of time geography can also be useful for explaining the effects of gasoline price changes on traffic safety. Within the theoretical framework of time geography, we understand gasoline prices as one type of capability constraint. We expect that as gasoline prices increase (as the capability constraint becomes stronger), traffic crash rates will decrease. Furthermore, we expect that these capability constraints will be stronger for some and will manifest themselves in differential effects by age, gender, and race. The next section summarizes existing studies of gasoline price effects on traffic safety. Immediately following that is a section that reviews time geography theory and proposes a conceptual framework for understanding the relationship between gasoline prices and traffic Chi, Porter, Cosby, and Levinson 3 crashes within that context. Next, the data and methods section introduces the data that are related to traffic crashes by age, gender, and race in the state of Mississippi from April 2004– December 2008 and addresses the methods that direct our analyses. Finally, the results section reports our findings within the theoretical framework of time geography and is followed by a concluding section with a summary and discussion of our findings and their implications. PRIOR RESEARCH OF GASOLINE PRICE EFFECTS ON TRAFFIC SAFETY There exists an increasing body of literature that examines gasoline price effects on traffic safety (see Chi et al. 2010 for a review of the literature). Most of the studies, which have been undertaken with nearly a sole focus on fatal crashes (e.g., Grabowski and Morrisey 2004, 2006; Leigh and Geraghty 2008; Leigh and Wilkinson 1991; Wilson et al. 2009), found that increasing gasoline prices (or taxes) lead to higher traffic safety as measured by fewer traffic crashes. The effects were also found to vary by demographic characteristics of drivers and by crash types. These findings highlight a couple of important trends that have helped guide the current research. First, existing literature on the subject has consistently shown that the effects of gasoline prices on traffic crashes differ by age, gender, and race. The effects tend to be larger on younger drivers than on older drivers, larger on male drivers than on female drivers, and larger on black drivers than on white drivers (Chi et al. 2010; Grabowski and Morrisey 2004; Leigh and Wilkinson 1991). Second, gasoline price effects on traffic crashes differ by fatal crashes, injury crashes, property-damage-only crashes, and drunk-driving crashes (Chi et al. 2011; Huang and Levinson 2010). When compared to overall crashes, gasoline prices have larger effects on less severe crashes and drunk-driving crashes. Although these studies found a statistically significant association between gasoline prices and traffic crashes, a conceptual or theoretical framework has not been given to explain their association. In this study, we attempt to use time geography theory to explain the effect of gasoline prices on traffic crashes. UNDERSTANDING GASOLINE PRICE EFFECTS ON TRAFFIC SAFETY WITHIN TIME GEOGRAPHY THEORY Time Geography Theory Originally developed by Hägerstrand (1970), time geography theory studies the relationships between the behavioral possibilities of individuals and the various spatial and temporal constraints on those behaviors (Shaw and Yu 2009; Yu and Shaw 2007). The core notion of time geography is that an individual’s existence and activities are constrained by spatial and temporal attributes (Pred 1977). Time geography theory asserts that an individual can only participate in activities in a single location in space at one single time (Miller 1991). Three types of constraints on the activities exist: capability constraints, authority constraints, and coupling constraints (Hägerstrand 1970). Capability constraints refer to biological (e.g., sleeping and eating) and physical (e.g., vehicle ownership, time availability, maximum speed of travel) limitations that restrict an individual from participating in activities. Authority constraints represent limitations to accessing particular areas (e.g., military bases) or individuals that are classified by certain people or institutions. Coupling constraints indicate limitations for two or more individuals to participate in an activity in the same location at the same time interval. These three types of constraints dictate the spatial and temporal patterns of an individual’s movements. Cullen and Godson (1975) further refined the notion of constraints by noting that temporal constraints might vary in their rigidity by time of day, with constraints weakening later in the day as more discretionary activities are pursued. Chi, Porter, Cosby, and Levinson 4 Within this approach, an individual’s movements in space over time generate a space-time path, which can be understood as a series of lines in a three-dimensional system in which space is represented by a two-dimensional plane and time is represented by the vertical axis. However, any space-time path is only one of many possible space-time paths that can be taken by an individual in a given time interval. All the points that an individual can reach from an origin location to a destination location at a given time interval comprise a space-time prism. The prism is determined by the intermediate locations between the origin and destination points, the time required for participation in activities at those locations, and the travel time between each point. The space-time path and space-time prism are the two most essential concepts of time geography, and they provide a valuable measure of an individual’s accessibility within particular spatial and temporal constraints (Miller 1991). Conceptual Framework We understand gasoline prices as one type of capability constraint. Higher gasoline prices reduce an individual’s ability to afford the same amount of gasoline as that consumed at lower prices, causing an individual to drive less distance and less frequently, in turn reducing exposure to traffic crashes. Gasoline price increases reduce transportation accessibility by reducing the lengths and frequencies of economically feasible space-time paths and sizes of space-time prisms, which in turn leads to fewer traffic crashes. However, the capability constraint of gasoline prices varies by income and private vehicle ownership. An individual with higher income tends to have a higher capability of gasoline consumption and is less vulnerable to gasoline price increases. Yet the income influence on the gasoline price constraint is conditional on private vehicle ownership––gasoline prices affect automobile usage only if an individual owns a private vehicle. For those who do not own private vehicles, gasoline price changes are less likely to affect driving. For those who own private vehicles but have low income, the effect is substantial. For example, blacks and Latinos, who have relatively lower private vehicle ownership rates (Raphael and Stoll 2001), are less likely to be affected by gasoline price increases. In contrast, young drivers who own cars but lack discretionary income are likely to be substantially affected by increases in gasoline prices. We conceptualize gasoline prices as influencing traffic safety through at least five intermediate factors: trip frequency and distance for non-work trips, travel modes, driving behaviors, vehicle fuel efficiency, residential and work locations (Figure 1). First, rising gasoline prices could cause people to drive shorter distances by reducing trip frequency (fewer space-time paths) and distance (shorter space-time paths) as well as to make more multi-purpose (linked or chained) trips rather than single-purpose trips (for shorter total lengths of space-time paths) (Huang and Levinson 2010). Second, an increase in gasoline prices could cause some drivers to switch from personal vehicle use to other commuting modes, such as using public transportation (Lane 2010), carpooling, biking, or walking (fewer motor vehicle space-time paths and fewer destinations reached by motor vehicle). Third, surging gasoline prices could cause people to drive in a more fuel-efficient manner, such as driving more slowly and reducing sudden speeding and braking, which in turn likely lowers crash risk. These three paths through which gasoline prices affect traffic safety likely occur immediately after gasoline price increases. However, over the long term, gasoline prices may influence traffic safety through two additional factors—higher vehicle fuel efficiency and residential relocation or job change. An increase in gasoline prices could persuade some drivers to switch to fuel-efficient vehicles, which are generally lighter (and therefore more vulnerable if hit by a larger vehicle but likely to cause less Chi, Porter, Cosby, and Levinson 5 damage to another vehicle or a pedestrian in a crash) and equipped with better safety technologies (the net effects of lightness and safety have been found to be insignificant; Leigh and Wilkinson 1991). Increased gasoline prices could also induce workers, especially lowand medium-income automobile commuters who live far from their workplaces, to relocate closer to their workplaces (shorter space-time paths), and may encourage low-wage, younger, and parttime workers to find jobs nearer their residence (shorter space-time paths). FIGURE 1 A conceptual framework. The effects of gasoline prices on the intermediate factors eventually lead to greater traffic safety: rising gasoline prices reduce the total lengths of motor-vehicle-based space-time paths of an individual during a given time interval—i.e., an individual travels less and for shorter distances by car—which in turn reduces the likelihood of a traffic crash occurrence for that individual. Thus, we expect that as gasoline prices increase (as the capability constraint becomes stronger), traffic crash rates will decrease. Furthermore, as we expect capability constraints to be differentially impactful, we expect that significant variations will exist across demographic categorizations of the population. These conceptual links through which gasoline prices affect traffic safety are complex. They have not been tested in existing studies; this study likewise does not attempt to test these conceptual links. Measurement of the intermediate links might have potentially large errors, and the conceptual links might be imprecise. Moreover, the data for studying these conceptual links are unavailable and may be difficult to collect. Therefore, instead of modeling the effects by these links, this research directly models the effects of gasoline prices on reducing traffic crashes. The conceptual links are used only for the purpose of introducing the framework through which we interpret our findings and highlight the underlying mechanisms driving the findings reviewed in the literature section above. DATA AND METHODS Data and Variables This study makes use of data provided by the Mississippi Highway Patrol and analyzes monthly traffic crashes in the state of Mississippi from April 2004 until December 2008. Following the recent work of Chi et al. (2010), this analysis is unique in that most studies to date have only Gasoline prices Traffic safety Trip frequency and distance for non-work trips Commute modes Vehicle fuel efficiency Driving behaviors Residential and work locations Chi, Porter, Cosby, and Levinson 6 examined the effects of gasoline prices on traffic fatalities (e.g., Grabowski and Morrisey 2004; Leigh and Wilkinson 1991; Wilson et al. 2009) by making use of Fatal Accident Reporting System (FARS) data, which reports only information on fatal accidents. Our data is unique in that it contains information on vehicles, drivers, and passengers of all traffic accidents documented by the Mississippi Highway Patrol during the identified period. In addition, crashes per vehicle miles traveled (VMT) were computed using data collected from the Mississippi Department of Transportation. Data for regular-grade gasoline prices were obtained from the U.S. Department of Energy’s Energy Information Administration (EIA) (Figure 2). Given the location of Mississippi in the Gulf States region of the U.S., the data collected from the EIA were related to the average prices for all Gulf Coast states. Gasoline prices were also adjusted for inflation in June 2009 dollars. 0 50 100 150 200 250 300 350 400 45