New Metrics for Measuring Employment Outcomes for HUD-Assisted Households

New Metrics for Measuring Employment Outcomes for HUD-Assisted Households

The provision of safe, decent, and affordable housing is central to HUD’s mission. Research has shown that stable housing can have a powerful, positive influence on a range of social outcomes, including reducing episodes of homelessness, improving health, decreasing incidents of domestic violence, reducing hunger, improving school stability for children, and promoting improved economic opportunity when children reach adulthood. Existing research does not show conclusively, however, what effect assisted housing has on increasing the earned income of recipients — a strategically important goal of HUD. To better understand the relationship between HUD programs and employment outcomes for assisted households, HUD has developed a set of four indicators that use existing administrative data to track and measure the extent to which individuals and households participate in work or training. The indicators, the methodology of their derivation, and preliminary results are detailed in a newly released report, “Assessment of Economic Opportunity Metrics for HUD-Assisted Renters.”

Generating New Insights From Existing Data

In recent years HUD has sought to assess how well assisted households are accessing economic opportunity in ways that support self-sufficiency and positive exits from housing assistance programs. HUD’s 2018–2022 Strategic Plan identified the proportion of positive exits among assisted households; the percentage of work-able households exiting assisted housing with low subsidy needs; and the median percentage change in income of individuals who complete an economic development program as three possible indicators of growing financial stability through subsidized housing. Because administrative systems are not designed to capture information about outcomes after program exit and cost-effective external data sources are limited, the four newly developed alternative indicators outlined in the report instead use the existing administrative data to create a richer picture of self-sufficiency progress needed for positive exits from assisted housing. Three sources in particular are used: the HUD Inventory Management System/Public and Indian Housing Information Center, the HUD Tenant Rental Assistance Certification System, and the American Community Survey. Although all data sources have their unique strengths and limitations, the four new indicators can supplement the picture of economic opportunity for assisted households and better inform data-driven policymaking. Specifically, two of the indicators use American Community Survey data to provide context about local labor markets, an innovation enabling HUD to examine intermediate outcomes of assisted households with better insight about the opportunities and challenges they face. The four indicators are:

• The ratio of HUD-assisted to overall population employment rates for work-able adults;
• The percentage of work-able, HUD-assisted households with stable employment for 3 years;
• The ratio between HUD-assisted and overall population rates of full-time employment for work-able adults; and
• The percentage of HUD-assisted, work-able individuals who are either working or in school.

The report details the data sources, methodology, and limitations of each indicator. Limitations result from the differences in the data definitions used in the HUD datasets compared with those in the American Community Survey or from difficulties in collecting certain kinds of data. The report includes values for each indicator for the years 2012 to 2016 for the nation’s metropolitan statistical areas, the nonmetropolitan balance of states, and national averages. To assess the differences among various types of assistance, the researchers calculated indicators for three types of HUD programs: public housing; the Housing Choice Voucher (HCV) program; and assisted multifamily programs, which aggregate other HUD rental assistance programs into a single category.

Results and Future Directions

Between 2012 and 2016, all four indicators showed improvement across all three families of assistance programs, with public housing and the HCV program consistently registering higher results compared with the multifamily programs. In addition, the first and third indicators, comparing the employed and full-time employed number of HUD-assisted households to the employed and full-time employed number of general population households, show faster increases among the HUD-assisted population than among the general population. Such results do not demonstrate that the programs cause better outcomes. For example, it is likely that assisted populations experienced the post-recession recovery at a later point in the business cycle than the general population. The inherent inability of performance indicators to establish causality is a major reason they are only one tool for evidence-building. Performance indicators provide information for program monitoring and adjustments in a way that complements random-control trials and other structured program evaluations.

With these new indicators, researchers and policymakers gain an additional tool in their efforts to promote increased earnings through work among HUD-assisted households. Relying on timely, robust, and contextualized data gives a more finely grained representation of the lives of people participating in HUD programs. Furthermore, by enabling comparisons among programs, comparisons across geographies, and consistent tracking over time, the new indicators provide a uniform basis from which to evaluate the effects of diverse policy approaches.