Change in visible impervious surface area in southeastern Michigan before and after the “Great Recession:” spatial differentiation in remotely sensed land-cover dynamics

Abstract

Rates of urban growth and change are spatially differentiated within any given region and are often associated with the social and economic characteristics of communities as well as with preexisting patterns of development. We examined how those patterns of association were affected by the “Great Recession” in southeastern Michigan, a region with a low overall growth prior to the recession but with a high degree of internal heterogeneity and social segregation. We used an innovative application of available Landsat imagery and spatial autoregressive methods to relate the rates of change in visible impervious surface area (VISA) before (i.e., 2001–2005) and after (i.e., 2006–2011) the Great Recession to four factors that characterize tract-level variability in socioeconomic characteristics based on the Census data. Then, using a difference-in-differences approach, we modeled relationships between changes in these rates to examine the effects of the Great Recession on spatial differentiation in the rates of land-cover change. In both periods, VISA increased at higher rates in wealthier and less-deprived communities. Comparing the differences in rates of land-cover change from before to after the recession, we found that the rate of VISA increase declined most rapidly in wealthier and less-deprived communities. Therefore, the patterns of land-cover change indicate that the recession served to reduce observed differences in the rates of change across communities with different socioeconomic compositions. These observed patterns are consistent with both reduced private-sector investment in new residential development and increased public-sector investment in infrastructure.