Did immigrants take the vast majority of new jobs in Texas? That’s the claim coming from the Center for Immigration Studies, an advocacy group for immigration reduction.


But Ray Perryman, head of an economics research firm based in Waco, Texas, argues that the paper has significant methodological problems, calling the eye-popping numbers in the study “highly suspect.” Perryman who has closely studied the economic impact of immigration in Texas and elsewhere, raises three basic questions about the study.

1) Does the study undercount the number of native workers getting jobs? First, Perryman believes that CIS’s methods may be overestimating the proportion of new jobs going to immigrant versus native-born workers. He points out that the study’s analysis of job growth doesn’t distinguish between “new jobs” that didn’t exist beforehand and openings that were created because of retirement, attrition, or other voluntary exits from the workforce.

Perryman points out that native workers were far more likely to leave the workforce because of retirement, for example, than immigrants who arrived after 2007--departures that would disproportionately reduce net job growth among native workers. For example, if 1,500 new immigrant workers who got jobs, and 1,500 native workers got jobs, but 500 native workers retired, the CIS study would have only counted 1,000 native workers as getting jobs. In other words, every immigrant worker who gets a job “is counted as a net new hire...while not every new [native] worker is counted” if there’s a native worker who leaves because of retirement, Perryman concludes.

Steven Camarota, one of the co-authors of the CIS study, says that he isn’t aware of any concrete evidence that native workers made up an outsized number of voluntary departures from the workforce. He adds that even if that were the case, newly arrived immigrant workers are still filling the new job vacancies in the state at a higher rate than native workers, even though immigrants only made up a third of the population. “Why don’t they get more of the jobs?” he asks.

Perryman acknowledges that there’s something to this: the CIS paper confirms earlier studies showing that non-native born workers are generally quicker to be rehired than native workers during times of recession (though immigration seems to improve employment rates for both groups in the longer run). “The basic notion that some percentage increase in immigrant hiring that exceeds overall averages has occurred would not be surprising,” he says.” But Perryman still insists that the CIS methodology exaggerates this gap.

2) Does it overestimate the proportion of illegal immigrants? Second, Perryman casts doubt on CIS’ figures on the number of illegal immigrants who’ve gotten jobs. Camarota and his co-author rely on the Current Population Survey to estimate the illegal immigrant population.: Because the Census doesn’t ask about immigration status, they statistically extrapolated from the data that the CPS does provide, assuming that illegal immigrants tend to fit a certain demographic profile. Perryman, however, says that it’s not reasonable to assume that illegal immigrants will respond to government surveys like the CPS in large numbers. “While it is certainly true that undocumented people at times submit government forms, it is highly unlikely that they do so in a manner that would allow valid statistical extrapolation,” he writes.

Camarota counters that CIS used the same approach that the Department of Homeland Security, the Pew Hispanic Center, and other major institutions use to estimate the illegal immigrant population. “All of them generally show that 90 percent of illegal immigrants respond to surveys of this kind,” he says.

3) Does the study reflect Texas’ immigrant population as a whole? Finally, Perryman points out that the study’s conclusions about newly arrived immigrants in Texas aren’t likely to hold true for the immigration population on the whole. By restricting its scope to immigrants who’ve arrived after 2007, the study doesn’t take into account any job losses by immigrants who came before 2007, he says. For example, if an immigrant who arrived after 2007 takes the job of a immigrant who came earlier, that still counts as a net gain in the study.

Camarota explains that it’s his very intention to focus just on the most recently arrived immigrants--those who are most closely linked to our current immigration policy. If the most recent arrivals have gotten more jobs than the rest of the immigrant and native population, that’s worth noting. “We’re looking at the net findings for an immigrant population who wouldn’t otherwise be here,” he concludes.