Senate Testimony Claims 61% Welfare Use in Undocumented Immigrant Households

A single statistic — 61% — has become one of the most politically charged numbers in the American immigration debate. Cited in Senate testimony and amplified across partisan media, the figure purports to show that nearly two-thirds of households headed by undocumented immigrants use at least one welfare program. But behind the headline number lies a methodological minefield, a decades-old policy paradox, and a debate that reveals as much about how we measure social reality as it does about immigration itself.

The Claim and Its Source

In February 2026, the Center for Immigration Studies (CIS) released a report analyzing data from the Census Bureau's 2024 Survey of Income and Program Participation (SIPP), finding that 61% of households headed by undocumented immigrants used one or more major welfare programs [1]. The figure quickly entered the political bloodstream, cited in Senate Budget Committee proceedings and by lawmakers pushing for tighter restrictions on immigrant access to public benefits [2].

The CIS report, authored by Steven A. Camarota, the organization's Director of Research, placed the 61% figure in a broader context: 51% of households headed by legal immigrants and 53% of all immigrant-headed households used at least one program, compared with 37% for U.S.-born households [3]. Among the specific programs, 44% of undocumented immigrant households collected food benefits — double the rate of U.S.-born households — and 44% used Medicaid, compared with 27% for the U.S.-born [1].

Camarota, who has testified before Congress more than 30 times on immigration's fiscal impact, argued that the data demonstrated that existing laws designed to restrict immigrant welfare access have been "relatively ineffective" [1][4].

The Paradox: Ineligible by Law, Counted in Practice

The central tension in this debate is a legal paradox that few headlines acknowledge. Under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), undocumented immigrants are categorically ineligible for virtually all federal means-tested benefits — including SNAP (food stamps), Medicaid, TANF (cash assistance), Supplemental Security Income (SSI), and the Children's Health Insurance Program (CHIP) [5][6].

So how can 61% of their households be using welfare?

The answer lies in their U.S.-born children. An estimated 5.3 million U.S.-born children of undocumented immigrants hold Social Security numbers and are, as American citizens, fully eligible for the same federal benefits as any other citizen [7]. When a household is headed by an undocumented immigrant but contains U.S.-born children receiving SNAP, Medicaid, or free school lunches, CIS counts that entire household as an "illegal immigrant household using welfare."

This is not fraud. As CIS itself acknowledges, these are legal benefits flowing to U.S. citizens who happen to have undocumented parents [1]. The question is whether attributing those benefits to "immigrant welfare use" is analytically sound — or politically misleading.

Source: Center for Immigration Studies / 2024 SIPP

Data as of Feb 4, 2026CSV

The Methodology War: Households vs. Individuals

The 61% figure sits at the center of an ongoing methodological dispute between CIS and the Cato Institute, a libertarian think tank, that has produced radically different portraits of immigrant welfare consumption.

CIS uses the household as its unit of analysis. If any member of a household headed by an immigrant uses any welfare program, the entire household is counted. This approach, CIS argues, captures the "total burden" immigrants place on the welfare system, including through their dependent children [3].

The Cato Institute's Alex Nowrasteh has published multiple rebuttals arguing this approach is fundamentally flawed [8]. Cato analyzes welfare at the individual level, attributing benefits to the person who actually receives them. Under this framework, benefits received by a U.S.-born child are counted as benefits to a U.S.-born citizen — not to an immigrant. Using this method, Cato finds that immigrants use welfare at lower rates than the native-born [8][9].

The gap between the two approaches is enormous. Cato has found that, relative to Department of Homeland Security estimates, CIS overstates native-born welfare use rates by an average of 95%, foreign-born rates by 173%, and non-citizen rates by 208% [8].

CIS responds that Cato's individual-level approach ignores an obvious reality: if immigrants were not in the country, their U.S.-born children would not exist, and the welfare spending on those children would not occur. "If we want to know how much extra burden is placed on the welfare system because immigrants are in the country," Camarota has written, "that calculation necessarily includes the welfare consumed by their dependent children" [10].

This is not merely an academic quarrel. The chosen methodology determines whether immigrants appear to be net welfare consumers or relatively modest users of public assistance — a distinction with direct consequences for legislation and public opinion.

The Broader Fiscal Picture

The welfare-use debate is one piece of a larger and equally contested fiscal puzzle.

The American Enterprise Institute has estimated the net cost of unauthorized immigration at the federal, state, and local levels at $150.7 billion annually, with each undocumented individual or their U.S.-born child costing approximately $8,776 per year [7]. AEI's breakdown attributes roughly $5.8 billion in SNAP costs and $1.43 billion in TANF costs to U.S.-born children of undocumented immigrants [7].

But these cost estimates exist alongside significant revenue contributions. Undocumented immigrants paid approximately $26.2 billion into the Social Security Trust Fund in 2023, despite most being unable to ever collect those benefits [6]. State-level analyses have found that immigrants in Arizona generate $2.4 billion in annual tax revenue against $1.4 billion in benefit usage, and Florida immigrants pay nearly $1,500 more per capita in taxes than they receive in benefits [11].

Source: GDELT Project

Data as of Mar 13, 2026CSV

The most comprehensive assessment may be the National Academy of Sciences' landmark 2017 report, "The Economic and Fiscal Consequences of Immigration." It found that first-generation immigrants are more costly to governments than the native-born, but their children — the second generation — "are among the strongest fiscal and economic contributors in the U.S." [12]. Over a 75-year horizon, the fiscal impact of immigrants was generally positive at the federal level, though negative at state and local levels where education and healthcare costs are concentrated [12].

A 2024 Cato Institute study went further, calculating that immigrants generated a cumulative fiscal surplus of $14.5 trillion between 1994 and 2023 [9].

The Source Under Scrutiny

Any analysis of the 61% claim must also reckon with questions about the source. The Center for Immigration Studies describes itself as an independent, nonpartisan research organization, but it has faced sustained criticism from across the political and academic spectrum.

The Southern Poverty Law Center designated CIS as an anti-immigrant hate group in 2016, citing the organization's circulation of material from white nationalist websites and its commissioning of analysts who had embraced what the SPLC termed "racist pseudoscience" [13]. CIS has vigorously disputed this designation.

Media Bias/Fact Check rates CIS as having a right-center bias [14]. Academic critics, including Jeffrey S. Passel, a senior demographer at the Pew Research Center, have raised concerns about "methodological problems" in CIS studies, particularly those related to small sample sizes [14]. The Cato Institute, Center on Budget and Policy Priorities, Urban Institute, and fact-checkers at PolitiFact and the Washington Post have all published critiques of specific CIS claims [14].

CIS defenders note that the organization's data is drawn from official Census Bureau surveys and that its household-level methodology, while debatable, is a legitimate analytical choice that answers a specific and important question.

The Political Context

The 61% statistic has entered public discourse during a period of heightened political conflict over immigration. In March 2026, the Senate Budget Committee held a hearing titled "Sanctuary Cities: The Cost of Undermining Law and Order," which featured clashes between Republican senators and witnesses from opposing policy camps [2].

David Bier, an immigration analyst at the Cato Institute, testified that immigrants tend to work at rates above the national average and access fewer government benefits, arguing that increased immigration could help address the national deficit [2]. His remarks drew sharp rebukes from Republican senators, including John Kennedy of Louisiana [2].

The hearing also saw a tense exchange when Sen. Bernie Moreno pressed former Biden advisor Brendan Duke on whether illegal immigration constitutes a crime — a question Duke struggled to answer directly [15].

Senate Republicans have used the welfare data, among other arguments, to push legislation barring states and localities from limiting cooperation with federal immigration enforcement. Committee Chair Lindsey Graham argued that sanctuary policies undermine federal law [2].

Meanwhile, GOP budget proposals have included provisions to cut benefits and raise fees for legal immigrants, extending the fiscal debate beyond undocumented populations to the broader immigrant community [16].

Source: Bureau of Labor Statistics

Data as of Mar 13, 2026CSV

What the Data Does — and Doesn't — Tell Us

Several facts are not in serious dispute. Undocumented immigrants themselves are overwhelmingly barred from federal welfare programs. Their U.S.-born children, as citizens, are legally entitled to those programs. Households headed by undocumented immigrants have higher poverty rates and lower incomes than U.S.-born households, which naturally correlates with higher rates of means-tested benefit usage among eligible household members.

What remains genuinely contested is whether this pattern represents a fiscal drain that argues for restricting immigration, a natural consequence of integrating low-income families that ultimately pays off in the second generation, or something in between.

The CIS household-level methodology answers one specific question: what is the total welfare footprint of households that exist because of immigration? The Cato individual-level methodology answers a different one: do immigrants themselves consume more welfare than natives? Both are valid questions. Neither alone tells the complete story.

Camarota himself has acknowledged a critical nuance that often gets lost: the high welfare rates "are not caused by an unwillingness to work." In fact, undocumented immigrant-headed households are more likely to have a worker present than U.S.-born households [4]. The issue is that many immigrant workers hold low-wage jobs that do not lift their families above the income thresholds for means-tested programs — a reality that reflects the structure of the American labor market as much as it does immigration policy.

The Numbers Behind the Numbers

The 61% figure is accurate within the framework CIS uses. It is also, by design, a measure that conflates the welfare eligibility of U.S. citizen children with the immigration status of their parents. Whether that conflation is clarifying or misleading depends on what question one is trying to answer — and, often, on what policy one is trying to advance.

What the debate ultimately reveals is not a simple story of immigrants exploiting a generous welfare state, but a more complex picture of a workforce that fills essential low-wage jobs, a benefits system designed to assist low-income American citizens, and a political system that has spent decades unable to reconcile those two realities.

The 61% statistic will continue to be cited in hearings, campaign speeches, and policy papers. Whether it illuminates or obscures depends entirely on whether the people citing it are willing to explain what it actually measures.