Spatial effects of automated driving: dispersion, concentration or both?

This paper studies possible effects of automated driving technology on the concentration and dispersion of the population and on residential land prices in the Netherlands. We perform simulations with the Dutch spatial general equilibrium model LUCA. The simulations account for two possible effects of automation: (i) self-driving cars allow drivers to use their time in the car more productively; (ii) self-driving transport offers faster and more comfortable door-to-door transfers than traditional busses, trams and metro. We find that more productive time use during car trips results in population flight from cities. Residential prices converge: they fall in cities and rise in non-urban areas. The efficiency gain in public transport has an opposite effect. It leads to further population clustering in urban areas and an increase in residential price disparity between cities and rural areas. A combination of these two components may result in concentration of the population in the largest most attractive cities and their suburbs at the cost of smaller cities and non-urban regions. These results are in particular relevant for countries where public transport has a considerable share in urban mobility. Neglecting the impact of vehicle automation on public transport may result in biased policy recommendations.