After running a cash register for a few years at a New Jersey grocery store, my brother, who is no policy wonk, observed: “I thought immigrants couldn’t get welfare, yet so many people on food stamps have a foreign accent. Something’s not right.” The public has long sensed that there is a problem with immigrants’ use of welfare. New research published by my organization, the Center for Immigration Studies, shows that the public is right.
In 2012, 51 percent of all immigrant households (legal and illegal) used Medicaid or received cash, housing, or food assistance. No single program accounts for immigrants’ higher overall welfare use. Using data from the Census Bureau Survey of Income and Program Participation (SIPP), my colleagues and I calculated that the share of immigrant households that use food programs (40 percent) and Medicaid (42 percent) is much higher than that of native-born households (22 percent and 23 percent). Immigrant use of cash programs is somewhat higher than that of natives (12 percent vs. 10 percent); immigrant and native use of housing programs is comparable (about 6 percent).
These disparities might not create a fiscal problem if immigrants paid a lot more in taxes than they actually do. But we estimate that, on average, immigrant households pay roughly 11 percent less than native households in federal taxes.
There is, of course, no reason to see high welfare use among immigrants as a moral failing on their part. Rather, it reflects the fact that many of them arrive with modest levels of education. About one in four adult immigrants have not graduated from high school, a rate triple that of natives. Because immigrants earn wages that reflect their education levels, many are poor. This allows them or their children to qualify for welfare. Even among immigrant households without children, the share that uses welfare is significantly higher than that among their native-born counterparts (30 percent vs. 20 percent).
The SIPP data show that welfare use among less educated immigrants is extraordinarily high. In 2012, 76 percent of households headed by an immigrant who had not graduated from high school used one welfare program or more, as did 63 percent of households headed by an immigrant with only a high-school education.
The overwhelming majority of immigrant households receiving welfare have at least one worker. But working does not always provide self-sufficiency. For example, a household of two children and a single parent earning $20,000 a year qualifies for almost every non-cash welfare program. If the parent’s income is low in some months, the family may also receive cash assistance.
All of this means that employers’ use of immigrant labor to fill low-wage jobs in agriculture, meat processing, hotels, restaurants, retail, construction, and other industries creates very large costs for taxpayers. It would make far more sense to draw some of the 30 million working-age native-born Americans who have no more than a high-school education and are currently not working back into the labor market by reducing immigration and letting wages rise. Of course, less educated natives also use a lot of welfare, but they are already here and do not add a new cost.
It might be countered that we should continue to welcome immigrant workers but also “build a wall around the welfare state.” The 1996 welfare-reform legislation tried to do this, however, and failed.
Advocacy groups’ efforts to scale back welfare restrictions, state and federal bureaucracies’ resistance to enforcing them, and subsequent changes to the 1996 law by Congress itself have all undermined restrictions that were not very strong to begin with. As a result, they apply to only a modest share of immigrants. Some programs are simply not restricted; there are numerous exceptions to restrictions on others; states often provide welfare with their own money and so are not bound by federal restrictions; and some provisions go entirely unenforced.
To give just one example, consider the requirement that new immigrants have sponsors who will provide financial support if they need it. An “indigent exception” allows them to receive welfare if they are poor (defined as making less than 130 percent of the federal poverty level), thereby negating the whole point of the sponsorship requirement, which was to make their welfare use unnecessary.
The most important reason that the 1996 welfare restrictions did not have their intended effect is that non-citizen immigrants can receive benefits on behalf of their U.S.-born children, who, as citizens, have access to all welfare programs. (They can receive welfare themselves if they are refugees, asylees, or have been here for more than five years.) Otherwise-ineligible immigrants, including those here illegally, can even live in public housing if they have a U.S.-born child. There is a native-born child present in 86 percent of illegal-immigrant households that receive welfare.
People often assume that the problem involves illegal immigrants only. But based, again, on the SIPP data, we estimate that 49 percent of legal-immigrant households receive welfare — a share not very different from the 51 percent, mentioned above, of legal- and illegal-immigrant households combined.
Legal immigrants also benefit from a greater variety of welfare programs. Illegal-immigrant adults can receive benefits via the Women, Infants, and Children program, and sometimes through Medicaid; illegal-immigrant households can do the same, but they also make extensive use of food programs and sign their children up for Medicaid at very high rates. Legal-immigrant households heavily use not only these programs, but also of those that provide cash assistance. They account for three-fourths of all immigrant households that receive welfare.
A common response to these facts is that immigrants in poverty use slightly less welfare than do poor natives. While this may be true for some programs (and not others), the argument ignores the fact that the poverty rate among immigrant families is about 50 percent higher than among natives. It also ignores the central policy question: Why allow in so many less educated immigrants in the first place, regardless of how they compare with native-born citizens who would be here anyway?
Another response is that, while immigrant households receive more welfare than native households, the difference in receipt of all government programs is smaller when Social Security is taken into account. It’s true that native households are currently more likely to receive Social Security than immigrant ones, but the data show that almost everyone — immigrant and native alike — eventually collects Social Security upon reaching retirement.
Still another reply is that immigrant households receive smaller average welfare payments than do native households participating in the same programs. But data from the SIPP show that this is simply not true: If anything, immigrants receive larger average benefits. For example, immigrant households getting benefits through Temporary Assistance for Needy Families received, on average, about $2,900 in cash from the program in 2012, compared with roughly $2,500 for native households. Immigrant households on food stamps and Supplemental Security Income also received larger annual payments than did native households on these programs.
Looking at welfare alone cannot establish whether immigrant households are a net fiscal drain. But on average, immigrant households have lower incomes, pay less in federal income and payroll taxes, and, of course, make more extensive use of welfare programs. And while estimating all tax payments and all costs for immigrant households is difficult, owing to limited data, calculations that have attempted to do so — such as those of the National Academy of Sciences (1997) and the Heritage Foundation (2013) — have found that immigrant households are a significant net fiscal drain.
Perhaps the most honest reply is that, as Bryan Caplan has argued in the Cato Journal, it is simply “morally impermissible” to limit immigration. One is certainly free to believe this, but one then must accept, as its correlate, that American citizens are obliged to pay the costs of an open-borders policy. Stating all of this forthrightly is better than pretending that immigration creates no fiscal burden for taxpayers.
Under the current system, some 20 to 25 million new legal immigrants will settle in the country in the next two decades. The overwhelming majority will be allowed to come because they have a relative here, not because of their work skills. Accordingly, if we want to avoid high immigrant welfare use in the future, we will have to implement a more selective legal-immigration system that admits, primarily, skilled immigrants who are unlikely to need welfare programs. We will also have to start enforcing laws against illegal immigration. Otherwise, we must accept without complaint the welfare costs that an unchanged system will continue to impose.
– Mr. Camarota is the director of research at the Center for Immigration Studies.