Will self driving cars mean more or fewer people in cities? | Netherlands

This study was done in collaboration with KIM Netherlands Institute for Transportation Policy Analysis

The effect of the automated and autonomous driving on where people live and work is twofold. On the one hand, densely populated urban areas become even more populated. On the other hand, non-urban regions attract population. We calculated these effects using the Dutch spatial equilibrium model LUCA; the above maps show the outcomes.

The two effects follow from two trends in the automation technology. On the one hand, shared automated vehicles bringing people cheap and quick from door to door, may replace the current public transport. This is especially true in urbanized areas where the high population density makes shared rides profitable. This trend attracts population to cities (see left map). On the other hand, in a self-driving private car, a long commute is less annoying as one can work or relax. This may make non-urban areas a more attractive place to live (see the map in the middle).

The map on the right combines the two trends. In Randstad, the conurbation in the West of the country, migration to the city prevails and the population grows. The impact of the population spread is more pronounced in the cities outside of the Randstad: these cities lose population, partly to the surrounding rural areas.

In developed countries, large attractive cities have recently been growing harder than the rest of the country. Our conclusions suggest that these cities may be facing additional challenges in accommodating population growth, especially in areas with a large share of public transport in the modal split.

July 2018


- Gelauff, G., Ossokina, I.V. and C.N. Teulings (2018) Spatial effects of automated driving: dispersion, concentration or both? , Discussion paper, revision submitted

- Gelauff, G., Ossokina, I.V. and C.N. Teulings (2017). English blog and Dutch blog , KIM Netherlands Institute for Transportation Policy Analysis.