Hysteresis & Urban Rail

Abstract

Cities are endowed with and accumulate natural and constructed assets based on their unique histories, which in turn define the choice set of the present. But, common practice is that current behaviour can be described without reference to past circumstances. This work departs from that practice by examining the effects of historical urban rail on current residential location and travel behaviour, from the era of horsecars (1865) and streetcars (1925) to the present in Boston. It uses tract level data to explore the hysteretical effects of past access to rail—the extent to which the urban system retains the impacts of rail even when it no longer exists. Current density and travel behaviour are measurably influenced by past access to rail. These findings are robust to a series of alternate causal, functional, and spatial specifications. The built environment and demographic patterns are found to be the strongest mechanisms for these persistent effects. Past access to rail has shaped the city, and that shape has, in turn, affected travel behaviour. For density and auto ownership there is also a residual measurable effect of past access unexplained by the built environment or demographic patterns. This research shows that past rail access continues to reverberate in current residential location and travel behaviour. These findings of hysteresis add to an understanding of the long-term impacts of rail infrastructure, and suggest that if higher density and lower levels of auto ownership are desirable, policymakers should focus on reuse of areas that were built around rail.