Long-run studies of daily travel: methodological review and convergence of distance traveled per capita across cities

The purpose of this paper is to review previous studies that have modelled daily travel using long-run variations of data (20 years or more), focusing on their theoretical (micro-founded) assumptions, and then to present a case study applying one of these assumptions. The review intends to clarify how these studies assume intergenerational differences in the (unobservable) mobility preferences of individuals, which determines how they describe the changes in the aggregate of individuals. This study shows that one group of studies describes these changes using changes in the behaviour of individuals representing all generations and by assuming their preferences over time. This assumption is useful, as it accommodates all types of data covering different periods, including the so-called aggregate data. This study also shows that the other study group describes them using the turnover of individuals of different generations, assuming that individuals born in different periods could have different preferences. This has been made possible by following birth cohorts over time using data from repeated surveys while enabling a focus on generation-specific characteristics in daily travel. Next, the case study provides an example of how most previous long-run studies on daily travel have employed theoretical assumptions. It applies the first assumption to examine a convergence hypothesis of distance travelled per capita across cities. This hypothesis is derived from the assumptions posed primarily on the production-side of a city’s transportation system, including diminishing returns under current transportation technologies. A linear regression model is employed to test the hypothesis with a city-level data set covering 39 developed cities in the world from 1960 to 2000. The results support the hypothesis, showing negative and statistically significant effects of per capita annual kilometres travelled on their subsequent average growth rates in every model with different sets of the control variables representing the surrounding conditions of the transportation network. The results of the case study suggest that current daily travel may be exhausting most of the quantitative potential of a city’s transportation network under the current technologies.