Targeting Housing Aid to the Homeless: Evidence from the National Survey of Homeless Assistance Providers and Clients

This study combines data from the Survey of Income and Program Participation with the National Survey of Homeless Assistance Providers and Clients to estimate how well units of subsidized housing are targeted toward households at risk of being homeless. Additional funding for subsidized housing is a commonly proposed solution to the problem of homelessness. However, research on the effect of subsidized housing in reducing homelessness is far from conclusive. Using data from a 1987 study of the homeless, Early (1998) found nearly 95 % of households in subsidized units would not be expected to be homeless if their subsidies were removed. Some have argued that those results are no longer valid since more recent selection criteria examine carefully a household’s risk of being homeless. The results presented here suggest that subsidized housing is still not targeted toward those most at risk of being homeless. Therefore, a simple expansion of existing housing programs will have little effect on the number of homeless. In fact, the estimates suggest that those in subsidized units in 1996 are even less likely to be homeless than the subsidized population in 1987