Association for Public Policy Analysis & Management

Federal housing spending as a whole—including the mortgage interest deduction used by many taxpayers—intensifies economic inequality by concentrating benefits for those at the higher end of income distribution, as researchers have shown. But decision-making about the allocation of limited federal resources for low-income housing programs and how it may affect inequality among those who are income eligible has received less attention. Only one in four households eligible for participation in low-income housing programs receives assistance. This raises the questions of what mechanisms distinguish the lucky 25% and how rationing procedures affect stratification among the poor.

Using multiple data and analytic methods, I first characterize the architecture of low-income housing programs and then examine who is more likely to benefit from it. Using the administrative plans that local housing agencies in Michigan use to administer the HCV program (collection rate, 94% = 59/63), I examine two important program parameters related to managing waitlists under the discretion of local Public Housing Agencies to characterize resource allocation processes in the current low-income housing programs: waitlist preferences and rationing algorithms. Results suggest that local housing agencies universally use a rationing algorithm that regularly purges applicants from the waitlist who are unreachable to check their continued interest in and verify their eligibility status for program participation. This systematic rationing based on each of contact along with geographically bounded eligibility for waitlist preference that 40 percent of local PHAs in Michigan use could promote the selective attrition of applicants experiencing a higher level of residential mobility from the waitlist.

Using the first two waves of SIPP 2014 Panel data (2013-2014), I then examine whether more economically marginalized potential applicants are more likely to move, and thus to be rationed from the waitlist, than their higher income counterparts. To fully utilize the panel data that contains monthly information, I use a multi-spell discrete time proportional hazard model with time-varying covariates, with the risk onset of moving starting from the first month in each residence spell (8,410 persons aged 16 years or older have at least one income-eligible renter spell; together they contribute 10,308 such spells and 138,018 person-month observations). Results suggest that those who are in deep poverty (less than 50% of Federal Poverty Line) have a higher hazard ratio of moving by 70% than income-eligible applicants with higher income (100-125% of FPL), which exposes them to a greater risk of being purged from the waitlist due to non-response to PHA’s attempt to contact them or losing geographically bounded eligibility for waitlist preference. This paper argues that those with a lower chance of attritions are more likely to be those with a higher income, which suggests that current low-income housing programs preserve and even deepen economic stratification amongst the poor by shifting limited federal housing resources away from the least advantaged.