The impact of improved transit supply, railways and road investments on car use and ownership in cites and rural areas

This project aims to answer three questions: whether i) improved public transport reduces car use and car ownership ii) investments in roads increase car use and car ownership iii) increased urbanization, density, reduces car use and car ownership.

We will regress temporal changes in car use (or car ownership) on temporal changes in public transport supply (including train) and temporal changes in road infrastructure.We reduce the methodological problems in the previous literature by applying micro panel register data covering all individuals in the Mälardalen region (a third of the Swedish population) 1997-2017, providing a large data set free from sample selection, covering dense and rural areas.

This implies that endogeneity problems and spurious correlation, present in most earlier studies, are substantially reduced because we can control for population changes and self-selection (spatial sorting). The data provided by SCB is a register database covering all adult citizens.

It includes income and other socio-economic variables, coordinates for households’ and jobs’ locations, car ownership and yearly car travel distance from car inspections. It does not include any measure of road accessibility or public transport supply.

We build such data, based on a detailed representation of the transport system at three points in time, 1997, 2006 and 2017, measuring travel times by all modes using the Trafikverkets’ transport model from these years.