Over the past several decades, federal, state, and local governments in the U.S. have instituted a large number of programs that involve market-based incentive schemes to encourage private activities considered important to mitigating externalities or solving coordination problems. Many of these programs have been centered on the housing market, where externalities are pervasive. Such programs often take the form of tax credits, grants, or low-interest loans to households or developers for new housing construction, maintenance, and renovation projects.

These housing programs are often justified on the grounds that such projects have substantial positive spillovers on neighborhoods. Not only might one homeowner’s reinvestment activities increase property values of nearby homes, but they might also spur neighbors to reinvest themselves. Many rehabilitation programs are targeted at moderately and severely poor neighborhoods, where both increasing property values and instigating neighborhood revitalization efforts may be particularly important.  However, one major concern regarding these programs is that they could merely displace or crowd out unsubsidized private investment in housing. That is, some or even all of the investment subsidized by the government might have happened even in the absence of any government subsidy.

This paper examines the impacts of Missouri’s Neighborhood Preservation Act (NPA) on the state’s disadvantaged neighborhoods. The NPA offers tax credits to homeowners and developers that improve or expand the owner-occupied housing stock in low-income census block groups. Because there is a limit on the amount of credits that can be awarded, the state uses a lottery to determine which applicants receive the credits.

Taking advantage of the random assignment of NPA tax credits, I find only weak evidence that they increase the quantity of housing available in poor neighborhoods. However, I find that the tax credits induce changes in housing tenure and the composition of neighborhood residents, slightly increasing the share of of units that are owner-occupied and reducing the fraction of the local population that is disadvantaged. These results point to the potential for residential rehabilitation incentives to spur improvements in the quality, if not the quantity, of the local housing stock. Finally, I explore the nature and extent of externalities arising from housing investment induced by the program and find that spillovers associated with reinvestment are confined to at most the census block in which it occurs.