Nonprofit organizations play a critical role in U.S. housing policy, a role typically justified by the claim that their housing investments produce significant neighbor-hood spillover benefits. However, little work has actually been done to measure these impacts on neighborhoods. This paper compares the neighborhood spillover effects of city-supported rehabilitation of rental housing undertaken by nonprofit and for-profit developers, using data from New York City. To measure these benefits, we use increases in neighboring property values, estimated from a difference-in-difference specification of a hedonic regression model. We study the impacts of about 43,000 units of city-supported housing completed during the 1980s and 1990s, and our sample of property transactions includes nearly 300,000 individual sales.

We find that both nonprofit and for-profit projects generate significant, positive spillover effects. This finding in itself is significant, given the widespread skepticism about the impact of subsidized housing on neighborhoods. We also find some differences across sectors. First, the impact of nonprofit housing remains stable over time, whereas the effect of for-profit housing declines slightly with time. Second, while large for-profit and nonprofit developments deliver similar benefits, in the case of small projects, for-profit developments generate greater impacts than their nonprofit counterparts. These differences are consistent with theoretical predictions. In particular, in the presence of information asymmetries with respect to housing quality, the non-distribution constraint should lead nonprofits to invest more than for-profits in developing and maintaining features that benefit the broader community. Meanwhile, the fact that scale makes a difference to nonprofit impacts may reflect the capacity constraints often faced by smaller nonprofits