Neighborhood change research has focused extensively on the implications for residential stability (Ellen and O’Regan, 2010; McKinnish, Walsh, White, 2009; Freeman, 2005; Freeman and Braconi, 2004); much less attention has been paid towards similar outcomes for local retail services.  If we consider local retail services as a neighborhood amenity, then it is reasonable to assume that the change in the volume and composition of such services would be correlated with shifts in local population characteristics and the built environment.  Furthermore, we can imagine that, like residential turnover, retail change can manifest itself in more volatile churn or more gradual (stable) evolution.  In this paper, we exploit uniquely fine-grained data to identify the relationship between the rate and nature of retail turnover and neighborhood change more broadly.  Are there thresholds in terms of neighborhood socioeconomic conditions after which retail services become more or less stable, more or less diverse?  Do shifts in socioeconomic neighborhood characteristics bring about changes in quality-of-life retail services, such as grocery stores or pharmacies, or do they tend to influence more recreational uses, such as bars and restaurants?  What, in aggregate, do these changes mean for neighborhood livability and vitality?   

To answer these questions, we rely on a longitudinal micro dataset on retail establishments across New York City, with location-specific information for over a two-decade period.  We consider neighborhood conditions and change broadly, and so we supplement this dataset with rich information on neighborhood economic, demographic and built characteristics, as well as point-level crime reports, building permits and property sales transactions. We anticipate that the findings will inform the local government's management of neighborhood transitions and of the risks (and benefits) associated with opening and operating retail businesses in changing urban landscapes.