Highway Penetration of Central Cities: Not a Major Cause of Suburbanization*

Nathaniel Baum-Snow has investigated the impact of the introduction of the interstate highway system in the U.S. on the decline of central city populations. He finds that it is has had a significant effect on central city population decline and increased suburbanization. He suggests that had the interstate highway system not been built, the “aggregate central city population would have [grown] by about 8 percent.†We offer a number of reasons to believe that the reported correlation is spurious. That is, we believe that central city populations would have declined even in the absence of the interstate highway system. We suggest, first, that suburbanization of cities is a long-standing and almost universal process. As incomes rise, most people want the range and choice offered by automobiles. Increased auto use, in turn, causes the further dispersal of destinations which increases the demand for auto use. This is a powerful cycle that can be observed in practically all places where incomes have been rising. Looking beyond Baum-Snow’s sample, we examine European cities that also experience significant suburbanization, and we find no evidence that a highway that pierces the central city makes any difference to central-city population change. We suggest that one possible source of spurious correlation is the initial existence of undeveloped “greenfield†areas in central cities in Baum-Snow’s study. In suggesting mechanisms to explain his findings, Baum-Snow points to the monocentric city model; we offer some criticisms of the relevance of that model. We find no fault with Baum-Snow's statistical work, but it is possible to get the statistical significances right and still be wrong.