Why Are Immigration Advocates So Quick to Play the Race Card?

By Reihan Salam

Friday, July 01, 2016

Recently, Adam Ozimek, an economist and immigration advocate, made a provocative claim about the politics of immigration. While acknowledging that not all critics of high levels of less-skilled immigration are racists — he singles out Ramesh Ponnuru, Garett Jones, and yours truly as exceptions — he maintains that racism accounts for much if not most of the opposition to high immigration levels. Moreover, he asserts that if we were to somehow eliminate racism, opposition to immigration would drastically decline (“by more than 50%”), though he allows that the decline could be somewhat smaller.

What should we make of Ozimek’s contention? My guess is that if immigration policy were not viewed through a racial lens, opposition to immigration would in fact increase substantially. Many people who would otherwise be skeptical of the virtues of mass immigration can’t stand the thought of being racist. So when influential voices insist that opposition to immigration is racist, they find plenty of citizens who take those claims at face value. I have no more hard evidence for this claim than Ozimek has for his. But I do have at least some suggestive evidence.

If racism played a large role in driving opposition to immigration, non-Hispanic whites would, one assumes, be more favorably disposed toward immigrants of European origin than toward immigrants of Mexican origin. The political scientists Morris Levy and Matthew Wright suggest otherwise in a new paper that they’ve ably summarized in the Washington Post. Levy and Wright conducted an online poll of non-Hispanic whites in California in June 2015. All respondents were read a short vignette about a hypothetical program that would grant legal status to illegal immigrants, and then they were asked whether a hypothetical immigrant ought to be included in the program. One-third were asked about a Mexican immigrant (“Juan”), another third were asked about a Chinese immigrant (“Yuan”), and the final third were asked about a German immigrant (“Johan”). In every case, respondents were told that the immigrant in question had lived in the U.S. for two years. But in only half of them, they were also told that he spoke English and had held a steady job for the duration of his time in the U.S.

Levy and Wright posit that if anti-Hispanic bias were at work, respondents would discount the positive information in the case of Juan while taking account of it in the case of Yuan or Johan. The results were revealing. In the absence of information about English-language fluency or work history, respondents were seven to eight percentage points less likely to believe that Juan should be granted legal status. This clearly suggests some degree of bias. When the positive information was included, however, this gap disappeared. Essentially, Levy and Wright’s respondents were operating under the assumption that Mexican immigrants to the U.S. tend to be less educated than German and Chinese immigrants to the U.S., and so, lacking additional evidence, they assumed that Juan would be needier than Yuan or Johan. Once they knew that Juan spoke English and had been working steadily, they were as inclined to help him as to help his fictional counterparts.

So haven’t Levy and Wright proven Ozimek’s point? Is the real reason for opposition to immigration this false notion that Latino immigrants don’t work? Not quite. Levy and Wright’s findings are in keeping with the work of the political scientists Jens Hainmueller of Stanford and Daniel Hopkins of the University of Pennsylvania, who’ve also surveyed Americans on their attitudes toward different kinds of immigrants. Hainmueller and Hopkins found a broad consensus: Americans strongly prefer educated immigrants in high-status jobs over other immigrants, and this preference varies little according to education, partisanship, labor-market position, or ethnocentrism.

With this preference in mind, many aspects of the U.S. immigration debate come into clearer view. Why might the non-Hispanic whites surveyed by Levy and Wright be more inclined to assume that a Mexican immigrant relies on safety-net benefits than does a German or Chinese immigrant? Like most voters on both sides of the immigration debate, they are presumably relying on anecdotal evidence, which tells them that the incomes of Mexican immigrants are lower than those of German and Chinese immigrants. Anecdotal evidence can be highly misleading, and when combined with racial prejudice, it can seriously distort our perception of what is going on in the world. In this case, however, the belief that households headed by Mexican immigrants have lower incomes than those headed by Chinese and German immigrants is correct. In 2014, the median household income for households headed by Mexican immigrants was $37,390, well below the median for all immigrant-headed ($49,487) and native-headed households ($54,565). For households headed by Chinese immigrants, the median household income in 2014 was $57,000. For households headed by French, Swedish, and Danish immigrants, median household incomes were $81,000, $79,000, and $76,000 respectively, and the number for European immigrants overall was $60,000. As in most affluent market democracies, the U.S. safety net is designed — albeit imperfectly — to transfer resources to households with below-average incomes. So it should come as no surprise that immigrant-headed households with low incomes really are more reliant on safety-net benefits than are those with high incomes.

Drawing on data from the Census Bureau’s Survey on Income and Program Participation, the Center for Immigration Studies has found that while 73 percent of households headed by an immigrant from Central America and Mexico are enrolled in at least one safety-net program, the same is true of 32 percent of households headed by East Asian immigrants, and of 26 percent of European immigrants. Does this gap in program participation reflect that immigrants from Central America and Mexico are less inhibited about relying on safety-net benefits than European or East Asian immigrants are, or does it reflect some other difference? One thing we do know is that, on average, immigrants from Mexico have lower levels of formal education than do immigrants from Germany and China, to draw on Levy and Wright’s experiment. This difference in educational attainment is, as we’ve discussed elsewhere, compounded when we adjust years of schooling for the quality of schooling, as the quality of schooling in Germany tends to be somewhat higher than that in Mexico.

Why do we see such pronounced differences in immigrants from different countries? One driver is domestic inequality in the countries that send immigrants to the U.S., as George Borjas of the Harvard Kennedy School has found. One striking pattern he has identified is that immigrants originating from low-inequality societies tend to earn higher entry wages than do immigrants originating from high-inequality societies. This is true even when we compare countries that are at similar stages of development. To explain this pattern, Borjas posits that, when the payoff to acquiring skills is relatively high, as it is in high-inequality societies, emigration is a less attractive option to skilled professionals than when the payoff to acquiring skills is relatively low, as in low-inequality societies.

Although a moderately skilled worker might have better economic prospects in egalitarian South Korea than in the U.S., living in a highly egalitarian society could be more frustrating for an ambitious skilled professional who knows she could command a much higher wage in the U.S. The situation is reversed in an inegalitarian society, such as Mexico’s, where skilled professionals can lead lives that are in many respects more comfortable than the lives they would lead in the U.S., whereas middle-income workers can greatly improve their circumstances when they move north of the border. Moreover, if it is extremely difficult for people of modest means to acquire valuable skills, as is often the case in inegalitarian societies, emigration might be the most attractive option for less-skilled people from poor families looking to climb the economic ladder. Mexican immigrants are not “worse” than European and East Asian immigrants in some moral sense. Rather, they are more likely to be drawn from the ranks of the relatively poor in their native country than are their European and East Asian counterparts, and so they are more likely to suffer from disadvantages that limit their upward social mobility in the U.S.

Might the fact that skilled immigrants to the U.S. fare better than less-skilled immigrants have implications for U.S. immigration policy? Much depends on outcomes in the second generation. The chief way that immigration shapes the American future is via the children of immigrants. If the children of skilled and less-skilled immigrants have identical experiences in the U.S., we could safely dismiss the case for more-selective immigration policies. As it turns out, however, the experiences of the children of skilled and less-skilled immigrants are in fact markedly different. Nathan Joo and Richard V. Reeves have observed that second-generation Asian Americans earn substantially higher incomes than second-generation Hispanics and native whites in part because Asian immigrants from many countries have higher levels of educational attainment than both the median person in their native countries and the general U.S. population — a phenomenon that Jennifer Lee and Min Zhou have dubbed “hyper-selectivity.”

In contrast, Mexican immigrants, who represent 63 percent of all Hispanic immigrants, have a lower level of educational attainment than does the U.S. population and the Mexican population — that is, Mexican immigration to the U.S. has been defined by “hypo-selectivity.” To put this differently: The reason the Asian-American second generation tends to be more affluent than the Mexican-American second generation is not that “Asian values” are superior to “Mexican values.” It is simply that differences in education and work experience in the first generation powerfully shape the lives of the second generation. This has nothing to do with race. Had Mexican immigration to the U.S. followed a “hyper-selective” pattern, the Mexican-American second generation would be much more like the Asian-American second generation than is currently the case.

Given the widely held preference among Americans for educated immigrants over less-educated immigrants, and for immigrants working in high-status occupations over those working in low-status occupations, it should come as no surprise that many Americans favor reducing immigration levels. If more Americans knew about the composition of the immigrant influx — if they knew about the number of less-skilled immigrants we admit relative to skilled immigrants — it seems likely that more Americans would favor a reduction in immigration levels, or at the very least a concerted effort to change the composition of the immigrant influx. For example, we might welcome more Latino immigrants with post-graduate degrees and fewer with a high-school education or less. The evidence we have to date suggests that Americans would greatly prefer this approach, even if it did not change the ethnic character of immigration to the U.S.

Far from being rooted in racism, opposition to immigration in the U.S. seems to be rooted in concerns about the ability of less-skilled immigrants to support themselves without Medicaid, SNAP, the earned-income tax credit, and various other supports. The voters I have in mind might not have a sophisticated grasp of public finance, but they intuitively understand that low-wage workers generally need more public assistance than high-wage workers, hence their apparent preference for a more selective immigration policy. One suspects that these concerns will grow more pronounced as more low-wage jobs grow susceptible to automation and offshoring, and as increases in the minimum wage lead employers to substitute skilled workers for less-skilled workers.

I can’t that imagine any of this will persuade Ozimek or others who are convinced that immigration skepticism is rooted in racism. Much depends on what exactly we mean by “racism.” The U.S. has a long and tragic history of racial conflict, which is still with us today. Americans are therefore very sensitive to charges of racism, and for good reason. Political entrepreneurs are by and large aware of the power of accusing others of bigotry, and so there is a strong temptation to level this accusation to advance any number of political objectives. To some, the belief that we have a greater obligation to those who currently reside in the U.S. than to those who do not is racist in itself. Suffice it to say, this definition of racism is not widely embraced outside America’s elite universities and editorial boards.

I have no doubt that those who make liberal use of the racism charge in political debates do so sincerely. That doesn’t change the fact that, in the case of the immigration debate, at least, racism plays a far smaller role than they’ve led millions of Americans to believe.