Dynamics of Automobile Ownership Under Rapid Growth

Little research has focused on how factors influencing travel behavior change in rapidly developing and motorizing cities. We examine the case of household motor vehicle ownership, focusing on potential variations in preferences revealed through vehicle choice models estimated for Santiago de Chile in 1991 and 2001, and including measures of relative location, metro proximity, residential density, and land use mix. The results indicate that preferences have changed from 1991 to 2001, suggesting that as incomes rise and vehicle ownership becomes increasingly affordable, demographic and land-use and other contextual variables change in their apparent influence. The results vary across land use and locational variables; most notably the relationship between vehicle ownership and land use mix appears to weaken over time, while the distance to CBD effect strengthens, and the residential density effect varies in the apparent direction of change, depending on the vehicle ownership category. An effect of proximity to Metro apparently emerges by 2001 for the household decision to own 3 or more vehicles. This research shows that although income and motorization rates rapidly increased in Santiago, certain elements of the built environment influence household vehicle ownership, and these influences change over time. Future research should focus on potential market segments, such as suburban versus urban, aim to control for self-selection regarding the land use and locational characteristics, and better understand the implications for travel forecasting. INTRODUCTION The growth in motor vehicle fleets arguably poses as the single most important factor influencing developing countries’ mobility and accessibility. Most persons want the convenience, status, and comfort of private motorized travel, but in aggregate these desires translate into well-documented problems. The potent force of global motorization raises numerous interesting questions. Does a foreseeable ceiling exist for motorization (vehicles per capita) in cities of the developing world? Will resource constraints or the accumulation of externalities attenuate motorization? Do the dynamics of vehicle ownership structurally change cities? Do the dynamics of urban growth change vehicle ownership preferences and trends? We set out to shed some light on the last of these questions, examining the case of Santiago de Chile, during the 1990s. International comparative evidence shows a strong relationship between income and motorization (1). In much of the developing world, the vast majority of the population is still at income levels well below the minimal vehicle ownership threshold. This situation is changing, however, as incomes rise and vehicle prices fall, spurred in part by entrance into the international motor vehicle production market of companies from countries like China and India as well as international trade of used vehicles (2). Within a country or city, the same income-ownership relationship exists, but at this finer resolution the influences of local policies and other more nuanced factors appear. In Chile, for example, household vehicle ownership data suggest that cities within “free trade zones” (import-tariff-free) – Punta Arenas, Iquique, Arica – have the lowest share of zero-car households in the country (3). Other local policies, not originally aimed at affecting vehicle ownership per se, also play a role. For example, Mexico City’s “Hoy No Circula” program, a restriction on driving by certain vehicles (based on license plate numbers) during high pollution days, created the perverse impact of promoting the purchase of an additional second hand vehicle (apparently imported from other parts of the country) by many families (4). The government more recently adapted the ban to create an incentive for purchasing cleaner vehicles, an approach also adapted in the case of Santiago, which uses a similar restriction policy. Evidence also suggests a relationship between household automobile ownership and relative transportation levels of service and urban form and design. In fact, increasing motorization and its attendant impacts may further induce motorization (5). For example, motorization fuels spatial decentralization, which in turn drives motorization. Further, while motorization exacerbates congestion, congestion may then create the perverse incentive of increasing automobile ownership and use. Increasing congestion further encumbers main arteries, slowing buses and other surface transit relative to private cars which can choose less congested alternative routes and/or destinations (5). Zegras (6) finds, for the Santiago case in 2001, that household vehicle ownership can be partly explained by the