Skip to main content

Research Note: Economic Security Programs Significantly Reduce Poverty in Every State

Government programs play a powerful role helping people in retirement, after a job loss or family breakup, or when employers pay wages too low to meet families’ needs. While far from adequate for many, economic security (or “safety net”) programs such as Social Security, food and housing assistance, and tax credits kept 40 million people a year above the poverty line in the years before the COVID-19 pandemic, including 10 million children, and cut poverty by half nationwide, according to a comprehensive examination of poverty data.[1] Our analysis uses a version of the Supplemental Poverty Measure (SPM), a more comprehensive measure of poverty than the “official” measure.[2] This research note describes these programs’ impact on poverty for each state and Washington, D.C.

Government benefit payments (minus taxes) reduced poverty by more than 40 percent in 49 states.Government benefit payments (minus taxes) reduced poverty by more than 40 percent in 49 states. In Wisconsin, for example, benefit payments cut the poverty rate by 62 percent: from 21.8 percent before counting benefits and taxes to 8.3 percent after counting them. (These figures reveal the importance of government policies for reducing poverty within states, but caution should be used when making cross-state comparisons, as explained below.) Economic security programs lifted many thousands of people above the poverty line in every state, ranging from 64,000 in Wyoming to 4.3 million in California. Not surprisingly, states with large overall populations had the largest numbers of people lifted above the poverty line. (See Appendix Tables 1 and 2 for state-by-state figures.[3])

Economic security programs also reduced racial and ethnic disparities in poverty. In Louisiana, for example, the poverty rate for Black children was 35 percentage points higher than for white children before counting government assistance and taxes but 14 percentage points higher than for white children after counting them. Nevertheless, Black and Latino individuals across the country are still much more likely to live in poverty than white individuals. These large, long-standing disparities largely reflect public and private discrimination (both past and present) in areas such as employment, housing, and education.[4] (See Appendix Tables 3, 4, and 5 for state-by-state poverty figures by race and ethnicity.)

Analysis Uses Broader Poverty Measure, Adjusts for Underreporting

To provide a fuller picture of the public income support system for low-income families, this analysis blends the SPM with corrections for underreporting of key government benefits in survey data. We average together data for five years (2013 to 2017) to increase the reliability of state estimates.[5] We end with 2017 since that’s the most recent year for which the corrections for underreporting are available.

Whereas the official poverty measure only counts pre-tax cash income when assessing a family’s ability to make ends meet, the SPM also counts non-cash benefits (like SNAP and rental assistance) and refundable tax credits (like the Earned Income Tax Credit). The SPM also subtracts federal and state income taxes, federal payroll taxes, and certain non-discretionary expenses (such as out-of-pocket health costs and child care) from income when calculating what resources a family has available to buy basics such as food, clothing, and shelter.[6] Thus, the SPM provides a more complete picture of the resources available to low-income families than the official poverty measure.

Under the SPM, a family is considered to be in poverty if its resources are below a poverty threshold ($27,005 for a two-adult, two-child family renting in an average-cost community in 2017) that accounts for differences in family composition and geographic differences in housing costs.

Along with using the SPM to provide a more accurate picture of state-level poverty rates, we correct for underreporting of key government benefits from SNAP, Supplemental Security Income, and Temporary Assistance for Needy Families using the Department of Health and Human Services/Urban Institute Transfer Income Model. Household surveys depend on participants’ recollections over many months and typically fail to capture some income; Census data are no exception. Correcting for underreporting reveals that economic security programs are more effective at reducing poverty than figures based on uncorrected data would suggest. (See Appendix Table 6 for state-by-state data on the impact of correcting for underreporting.)

These corrections make a particularly notable difference for SNAP, the largest of the three programs we correct for underreporting. The corrections roughly double the program’s estimated impact on poverty, relative to the figures that Census regularly provides. Appendix Table 7 shows how much SNAP reduces poverty by state, overall and for children, after these underreporting corrections.

Caution should be used when making cross-state comparisons of anti-poverty figures in this analysis. Unlike official poverty thresholds, SPM thresholds vary across the country based on geographic differences in housing costs. For example, the average SPM threshold for a two-adult, two-child renter family in 2017 was $32,988 in California and $23,979 in Alabama. (See Appendix Table 8 for the thresholds in every state.)

Under the SPM, economic security programs may appear less effective in states such as California, which have higher housing costs, than in states such as Alabama, which have lower housing costs. Most federal anti-poverty programs are designed to provide the most assistance to households near or below the official poverty threshold, where needs are higher. The official poverty threshold for a family with two adults and two children — $24,858 in 2017 — is well below the SPM poverty thresholds in many high-cost places. In those high-cost areas, therefore, lifting a family above the official poverty line is not necessarily enough to lift them above the SPM poverty line.

Census Bureau researchers have raised concerns about whether the current SPM geographic adjustment understates the amount of income needed for families’ well-being in lower-housing-cost states such as Alabama, which might provide fewer amenities — such as high-quality schools — than higher-cost states.[7] Areas with lower housing costs might also be more costly to live in due to lack of nearby stores or services, higher transportation costs, minimal internet service, or less support from local charitable agencies, costs not considered in the SPM’s geographic adjustment. Researchers continue to investigate the topic. (See Appendix Tables 9 and 10 to compare the impact of geographic adjustment on anti-poverty figures by state.)

End Notes

[1] Economic security programs examined in this research note include: Social Security, unemployment insurance, workers’ compensation, veterans’ benefits, Temporary Assistance for Needy Families, state General Assistance, Supplemental Security Income, the Supplemental Nutrition Assistance Program (SNAP), the National School Lunch Program, the WIC nutrition program, rental assistance such as Section 8 and public housing, home energy assistance, the Earned Income Tax Credit, and the Child Tax Credit.

[2] Many state-level estimates of poverty, such as those derived from the U.S. Census Bureau’s American Community Survey (ACS), are based on the official definition of poverty. For instance, state-level estimates of poverty from Table B17001 of the ACS are based on the official definition of poverty.

[3] For Appendix Tables 1-10 referenced in this paper, please see

[4] For an analysis that focuses on racial and ethnic disparities in national-level poverty trends and the impact of economic security programs on poverty by race and ethnicity, see Danilo Trisi and Matt Saenz, “Economic Security Programs Reduce Overall Poverty, Racial and Ethnic Inequities,” Center on Budget and Policy Priorities, January 28, 2021,

[5] Using five years of data increases the reliability of estimates by increasing sample sizes. Even with five years of data, however, we are not able to produce reliable estimates for all racial and ethnic groups in every state. To ensure the reliability of our estimates, we suppress data for groups with fewer than 150 unweighted observations in poverty.

[6] These calculations include federal and state income taxes and federal payroll taxes, but not sales, property, or other taxes due to data limitations. The joint effect of government assistance and taxes on reducing poverty would look smaller if it included these other taxes, partly because many state and local taxes heavily burden lower-income households.

[7] Trudi Renwick, “Incorporating Amenities into Geographic Adjustments of the Supplemental Poverty Measure Thresholds,” U.S. Census Bureau, December 2018,