Responding to a Critic of CIS's Welfare Study

The Center for Immigration Studies released a report this week showing that 51 percent of immigrant-headed households used at least one welfare program in 2012, compared to 30 percent of households headed by the native-born.

The report has garnered significant media attention, and along with that attention have come the inevitable attempts at rebuttals from advocates of high immigration. One of the most frenetic critics has been the Cato Institute's Alex Nowrasteh. We usually don't respond point-by-point when so many of the points are merely distractions or frankly just silly — it grants an undeserved credibility to the distractions — but for people who like bullet points, here are our responses:

Nowrasteh: Household-level analysis "isn't useful or used much anymore by scholars who study this issue."

• First, as we explain in detail in our report, households are the best way to evaluate welfare use for the simple reason that welfare eligibility is typically determined by the characteristics of households or families, not by individuals alone. Moreover, welfare benefits can often be consumed by all members of the household, such as food purchased with food stamps. Even when the government provides food or health insurance only to children, it creates a benefit to adult members of the household who will not have to spend money on those things.
• Cato published a report by one of its researchers, Mike Tanner, that measured a number of welfare benefits at the household level.
• When researchers have focused on the fiscal impact of immigrants or their welfare use, households have been the most common way to examine the issue. The National Research Council did a household-level analysis looking at the fiscal impact of immigrants. Deborah Garvey and Thomas Espenshade in their study of New Jersey also used households because, as they observed, "households come closer to approximating a functioning socioeconomic unit of mutual exchange and support." The U.S. Census Bureau, has also used the household as the basis for studying immigrant welfare use. A 2014 Census Bureau study looked at public assistance by household. The most detailed study on the fiscal impact of immigration was released by the Heritage Foundation in 2013, which also used households as the unit of analysis.
• A person-level analysis is needlessly confusing — is free school lunch a welfare benefit for the parent, the child, or both? — and it necessarily excludes benefits for U.S.-born children of immigrants.
• Which leads us to a crucial point: The children of immigrants, whether they are U.S.-born or not, must absolutely count toward immigrant welfare usage. Consider an immigrant who comes here, chooses to have children, and then turns to the taxpayers to help support those children. The idea that this should not count as immigrant usage of the welfare state is, frankly, absurd.
• For those who are curious about welfare use in households without children, we included that as Table A9 in our report. Welfare use is much reduced for both groups when there are no children, but immigrant households still use welfare at substantially higher rates than natives for almost all programs.
• What about U.S.-born adults in immigrant households? Don't they inflate immigrant welfare totals? No. The very first table in our report shows what happens when we limit immigrant households to those without any U.S.-born adults. The welfare rates are virtually unchanged.
• Finally, do larger immigrant households contribute to their higher welfare rates? Of course. Does that bias the analysis? No. Immigrants have larger households because they choose to have more children. As we already discussed above, having children is a decision immigrants make that can be quite costly to taxpayers. And to the extent that immigrant households have more adult members, any bias cuts both ways. More adults means there are more potential welfare recipients, but it also means more potential workers who can increase the household's income above the eligibility thresholds for welfare receipt.

Nowrasteh: "When immigrants consume welfare, the dollar value of the benefits is typically far lower for them than it is for natives — sometimes substantially so. CIS could have included the value of welfare benefits consumed, but they did not."

• No, we could not. Most of the non-cash benefits are not assigned dollar values in the SIPP, which is the dataset we used because it provides the best picture of welfare participation.
• Second, the claim that immigrant welfare users consume fewer welfare dollars than native welfare users rests on an analysis of the Current Population Survey (CPS) that Cato published in 2013, which looked only at those with low incomes. While a useful survey, the research we cite in our study makes clear that the CPS significantly understates welfare use. Further, as shown in Table A1 in the appendix of our report, that understatement is particularly pronounced with regard to immigrant welfare use. Therefore, it is likely that the CPS is also undercounting welfare dollar consumption for immigrant users relative to native users.
• Third, we know from Table A7 of our report that, once on welfare, immigrants tend to stay on it slightly longer, 10.2 months vs. 9.8 months.
• Fourth, while some of the data in the SIPP is reported only at the household or family level, where individual-level participation data is available it shows immigrant households have more users on average, which almost certainly translates into higher costs per household. As discussed in end note 25 of our report, of households on SSI in January 2012, natives had 1.15 users on average and immigrants 1.17. For native households with at least one person on Medicaid, native households had 1.95 users on average compared to 2.19 users for immigrant households. For TANF-using households, it was 2.16 users for natives and 2.68 for immigrants. For households using SNAP, native households had 2.5 users on average and immigrant households had 3.09 users. And for WIC, it was 1.32 users for immigrants and 1.4 for natives. With the exception of WIC, immigrant households using welfare tend to have somewhat more people using the program.

Nowrasteh: "Social Security and Medicare should be included because they are the largest programs in the welfare state."

• First, Social Security and Medicare are not means-tested anti-poverty programs, which is what we mean by welfare.
• Second, as we show in Figure 9, immigrant households do not pay more in Social Security taxes than native households, even though there are more workers on average in immigrant households and fewer retirees. This again is because so many immigrants have modest levels of education and earn lower average wages.
• Third, low earners (immigrant or native) generally receive more in benefits than they contribute in taxes to these programs. So, although working-age, low-skill immigrants are currently paying into the system, they will eventually take it all back and then some.
• The 2013 Cato welfare study mentioned above was entirely focused on the welfare system and dealt with cash programs, SNAP, WIC, energy assistance, and Medicaid, but did not include Social Security or Medicare as welfare.
• As we noted before, the Cato Institute has condemned Social Security as a Ponzi scheme. Under that view, each new layer of the pyramid — i.e., a younger generation of participants — digs a deeper hole. Simultaneously characterizing Social Security as a Ponzi scheme and trumpeting the contributions of new participants is fundamentally inconsistent. It's like prosecuting Bernie Madoff while encouraging him to find new marks.

Nowrasteh: CIS doesn't support non-welfare-using immigration either, so it's disingenuous for CIS to complain about immigrant welfare use.

• CIS supports the approach of the bipartisan Jordan Commission, which concluded that a more modest level of immigration would better serve the interests of the American people. We support the unlimited migration of the spouses and minor children of U.S.-citizens and some level of truly high-skilled immigration (people who will be much less likely to use welfare). But mass immigration, especially the unlimited immigration favored by Cato, is problematic in a modern, post-industrial society. The high use of welfare by immigrant households is one strong piece of evidence demonstrating this fact.

It is perhaps worth mentioning that no matter how much immigrants grow the welfare state, or lower the wages of low-skill natives, or generate social tensions, or pose a national security risk, or create any other problem, Cato would still have an ideological commitment to unlimited immigration. They believe foreigners have a right to come to the United States even if the American people through their elected representatives choose otherwise. Or as Cato Institute adjunct scholar Bryan Caplan wrote in the Cato Journal, limiting immigration is "morally impermissible", even if all the objections raised by those that favor limits are right.