Hypercongestion, autonomous vehicle, and urban spatial structure

Abstract: This paper examines the effects of hypercongestion mitigation by perimeter control and the introduction of autonomous vehicles on the spatial structures of cities. By incorporating a bathtub model, we develop a land use model where hypercongestion occurs in the downtown area and interacts with land use. We show that hypercongestion mitigation by perimeter control decreases the commuting cost in the short-run and results in a less dense urban spatial structure in the long-run. Furthermore, we reveal that the impact of autonomous vehicles depends on the presence of hypercongestion. Introduction of autonomous vehicles may increase the commuting cost in the presence of hypercongestion and cause a decrease in suburban population, but make cities spatially expanded outward.This result contradicts that of the standard bottleneck model. When perimeter control is implemented, the introduction of autonomous vehicles decreases the commuting cost and results in a less dense urban spatial structure. These results show that hypercongestion is a key factor that can change urban spatial structures.