Social Safety Net Programs: What Evidence Shows About Effectiveness

The American social safety net is a complex collection of federal, state, and local programs designed to provide support to people experiencing poverty, unemployment, disability, food insecurity, housing instability, and other challenges. The programs vary widely in design, eligibility criteria, benefit levels, and evidence bases. Understanding what research shows about their effectiveness, and where significant evidence gaps remain, is essential for informed debate about reform. Medicaid, discussed in depth elsewhere, is one of the largest safety net programs by spending and enrollment. SNAP is the primary food assistance program. The Earned Income Tax Credit is the largest cash assistance program for low-income working families. Social Security Disability Insurance and Supplemental Security Income provide income support for people with disabling conditions. Unemployment Insurance provides temporary income replacement for workers who lose their jobs. Public housing and housing vouchers address housing affordability. Research on the cumulative effects of safety net programs on poverty is informative but complex. The official poverty rate, which measures poverty based on pre-transfer cash income compared to poverty thresholds, understates the poverty-reducing effects of the safety net because it does not count in-kind benefits like SNAP, Medicaid, and housing assistance. The Supplemental Poverty Measure, developed by the Census Bureau to address this limitation, produces a more comprehensive picture. Research using the SPM finds that safety net programs collectively lift many millions of people out of poverty annually, with Social Security, the EITC, SNAP, and other programs each contributing significantly. Long-term effects of safety net programs on child outcomes are a particularly rich research area. Studies that exploit geographic or timing variation in program availability have found that access to safety net programs in childhood has lasting positive effects on educational attainment, employment, health, and earnings in adulthood. The effects are particularly strong for access to Medicaid, the EITC, and food assistance in early childhood. These findings imply that safety net investments in childhood produce returns that extend far beyond the program period. Work incentives and disincentives in safety net design have been a persistent concern. The structure of many means-tested programs creates implicit marginal tax rates, the rate at which benefits are reduced as income increases, that can be very high for low-income workers moving from no income to low income. When benefits phase out quickly as income rises, workers face effective marginal tax rates that reduce the financial return to work. Program reforms that address these high implicit marginal rates, such as the EITC's structure of an earnings phase-in and a gradual phase-out, are associated with stronger work incentives. The behavioral effects of safety net programs have been extensively studied and are more nuanced than simple moral hazard arguments suggest. Research does not find that safety net programs systematically reduce work or create dependency for most recipients. Most recipients of means-tested programs cycle on and off programs as their economic circumstances change rather than relying on them indefinitely. For some programs in some populations, evidence of modest work disincentives exists, but these are typically small and vary significantly by program design. Administrative burden, the costs that individuals bear in applying for, documenting eligibility for, and maintaining benefits, is increasingly recognized as a significant determinant of who actually accesses benefits they are entitled to receive. Research by Pamela Herd and Donald Moynihan documents that administrative burdens, including information costs, compliance costs, and psychological costs, disproportionately affect people with less education, less time, and fewer resources. Programs with high administrative burdens reach smaller shares of their eligible populations and are less effective as safety nets. Integration across programs is a persistent challenge in the safety net. People with multiple needs must navigate separate eligibility systems, enrollment processes, and administrative requirements for each program they seek. Integration, whether through streamlined enrollment, benefit bundling, or case management that connects individuals to multiple programs, has shown promise in several demonstrations. The fragmented structure of the safety net creates inefficiencies and access barriers that an integrated approach could reduce. The design of safety net programs reflects both evidence and political choices about what behaviors to incentivize and what kinds of support are acceptable. Reform conversations that engage honestly with the evidence, including evidence about who is served and who is excluded by current designs, are better positioned to improve program effectiveness than those driven primarily by political ideology.