Nate Cohn of The Upshot observes that a growing number of Hispanics are identifying as white. But I’m not sure he’s giving us a clear picture of the phenomenon of “ethnic attrition,” which we’ve discussed in this space.
Specifically, Cohn is drawing on new Census data which finds that an estimated net 1.2 million of the 35 million Americans who identify as Hispanic changed their racial identification from “some other race” to white between the 2000 and 2010 Census. The Census treats Hispanic identity as an ethnocultural category that exists in parallel to its racial categories: white, black, Asian, American Indian or Alaskan native, and native Hawaiian or other Pacific islander. One can be a white or black Hispanic, or for that matter an Asian Hispanic, etc., though a large majority of self-identified Hispanics identify as white, black, or as belonging to “some other race,” an identification that could reflect an Amerindian or mestizo racial identity separate and distinct from other racial identities, rooted in the history and the ethnocultural politics of various Latin American societies, with their richer, more complex racial typologies. The number Cohn cites is a net number because while 2.5 million Hispanic Americans shifted from some other race to white, 1.3 million switched from white to some other race — a switch that merits at least as much attention, I should think, as the switch in the other direction.
Cohn makes the usual points about this shift — that it complicates the notion that America will ever become a “majority-minority” nation as Hispanics come to identify as white, etc. It is worth noting, however, that the phenomenon he identifies is not in fact ethnic attrition, when people who could plausibly claim Hispanic identity choose not to identify as Hispanic at all. That is the phenomenon that Brian Duncan and Stephen Trejo have described in detail. Essentially, people with, say, only one grandparent with origins in Latin America are less likely than people with three grandparents in Latin America to identify as Hispanic at all, which makes intuitive sense. This is quite different from the fact that some Hispanic identifiers say that they are white rather than members of some other (mestizo) race, as Hispanic identity itself has political and cultural salience.
Moreover, Cohn’s broader conclusion strikes me as confused:
White identifiers are likelier to be second- and third-generation Hispanics than foreign-born and noncitizen Hispanics. They also have higher levels of education and income. The researchers’ data did not show the country of origin of the families of those people who shifted their identification.
The results are a strong sign that fears of a unique “Hispanic challenge,” where Hispanic immigrants might remain as a permanent Spanish-speaking underclass, are overblown.
But wait a second. Cohn notes that white identifiers have higher levels of education and income and that they are likelier to be second- and third-generation Americans. What he doesn’t note, per Duncan and Trejo, is that they are also likelier to be the products of intermarriage. These phenomena are related. Cohn is persuaded by the thesis that contemporary Hispanic immigrants are broadly similar to the Irish and Italian immigrants of earlier eras. The central problem with this thesis is that modern America has seen the rise of assortative marriage, in which people marry other people with similar educational credentials. During the first decades of the twentieth century, European immigrants, including immigrants from southern and central and eastern Europe, had levels of educational attainment that were comparable or only somewhat lower than those of native-born Americans; in the first decades of the twenty-first century, in contrast, the educational attainment of native-born Americans is substantially greater than that of the average among Hispanic immigrants, and second- and third-generation Hispanic Americans lag behind non-Hispanic native-born Americans. Hence it is more educated Hispanics who tend to marry non-Hispanics while less-educated Hispanics are less likely to intermarry and integrate. That is, the fact that we are seeing ethnic attrition among upwardly mobile Hispanics — the ones who have levels of educational attainment comparable to native-born Americans, who are least likely to live in segregated communities, and who are most likely to intermarry — means that we have a large residual population of Hispanics that is culturally and economically isolated. In 2012, Heather Mac Donald, writing in City Journal, observed the following:
Nationally, 42 percent of Latino children entering kindergarten are in the lowest quartile of reading preparedness, compared with 18 percent of white children, reports UCLA education professor Patricia Gándara in her 2009 book The Latino Education Crisis. By eighth grade, 43 percent of whites and 47 percent of Asians nationally are proficient or better in reading, compared with only 19 percent of Latino students.
Later on, she adds:
Hispanics made up nearly 60 percent of California’s poor in 2010, despite being less than 38 percent of the population. Nearly one-quarter of all Hispanics in California are poor, compared with a little over one-tenth of non-Hispanics. Nationally, the poverty rate of Hispanic adults drops from 25.5 percent in the first generation—the immigrant generation, that is—to 17 percent in the second but rises to 19 percent in the third, according to a Center for Immigration Studies analysis. (The poverty rate for white adults is 9 percent.) That frustrating third-generation economic stall repeats the pattern in high school graduation and college completion rates as well.
This backsliding from the second to the third generation is a staple of the literature on Hispanic assimilation, and it is important to keep in mind. While Cohn is convinced that the rise of white identifiers means that the “Hispanic challenge” is overblown, my sense is that white identification is only part, and a small part, of the story of Hispanic integration.
In that regard, the census numbers are not new: There is mounting evidence that Hispanics are succeeding in American society at a pace similar to that of prior waves of European immigrants.
What is the mounting evidence Cohn has in mind? The link is to a 2013 piece by Cohn’s editor, David Leonhardt, which doesn’t actually say what Cohn seems to think. For example, Leonhardt notes the progress between the first and second generations:
Immigrant Latino households have a median income that trails the national median by $24,000 (or more than 40 percent). Among second-generation Latino households, the gap is only $10,000, according to a recent Pew Research Center report. Similarly, only 7 percent of Latino immigrants marry someone of a different ethnicity; a whopping 26 percent of the second generation does. “It’s a very reassuring set of metrics,” said Paul Taylor, the Pew center’s executive vice president.
Yet Leonhardt neglects the widely-respected work of the UCLA sociologists Edward Telles and Vilma Ortiz, who find that “while Mexican Americans make financial strides from the first to the second generation, economic progress halts at the second generation, and poverty rates remain high for later generations.” Remarkably, Telles and Ortiz find that “educational attainment peaks among second generation children of immigrants, but declines for the third and fourth generations.” Suffice it to say, if Telles and Ortiz are right, Hispanics are not succeeding in American society at a pace similar to prior waves of European immigrants. In fairness, the size of the third and fourth generations is relatively small. It could be that U.S. institutions for promoting upward mobility — the pubic schools serving children in high-poverty neighborhoods, the criminal justice system, etc. — will help third- and fourth-generation Latinos close the gap with whites, and that the economic conditions of the next century will be more favorable to the interests of less-skilled workers than those that prevailed in postwar America (U.S. economic hegemony, war deaths in the First and Second World Wars that contributed to tighter labor markets, etc.), despite ongoing technological advances. But does that seem plausible to you?
And on family structure, Leonhardt writes:
Even one alarming trend among the children of Latino immigrants highlights their increased American-ness: younger Latinos are having more children outside marriage than their parents did, just as whites and African-Americans are.
This certainly speaks to segmented assimilation, i.e., the phenomenon in which skilled and less-skilled, authorized and unauthorized immigrants, experience different “modes of incorporation.” Low-income, less-skilled Latino immigrants and their children have family structures very similar to low-income, less-skilled non-Latinos. But if anything, it suggests that Latino immigrants will have a more difficult time climbing the economic ladder than earlier waves of immigrants — exactly the opposite of the point Leonhardt is trying to make.
Leonhardt’s column is in part an implicit argument for granting unauthorized immigrants legal status. The literature on segmented assimilation clearly demonstrates that unauthorized status is a real barrier to upward mobility, yet this is just as much a case for stronger immigration enforcement measures as it is for legalization. But it is also an argument that we shouldn’t be overly concerned about the fact that Hispanics lag so far behind other Americans:
To be an impoverished immigrant who does not speak English and has few labor-market skills is not easy. Over time, the specific challenges — legal, cultural and educational — have changed. Yet the core parts of the story have not, including its trajectory.
This argument strikes me as wrongheaded. Yes, the trajectory is upward. But if you consider family structure to be a core part of the story (I do — America has fewer intact families now than in the past, and the advantages that accrue to children raised in stable two-parent households relative to other children are arguably greater than they were in the past), the broader economic conditions shaping labor market outcomes for less-skilled workers (they were better in the 1920s and the 1950s than in the 2010s), skill-biased technical change, the rise of mass incarceration (second-generation Mexican-American males are incarcerated at almost the same rate as native-born African-American men), then no, the core parts of the story have changed and they’ve changed quite a lot. Another core part of the story is that the U.S. is a more egalitarian society now than we were in the first decades of the twentieth century. Large ethnic and racial disparities that were tolerated then are less likely to be tolerated now. The fact that Italian immigrants and Americans of Italian origins were poorer than other Americans was not considered a pressing issue that threatened the legitimacy of Americans politico-economic order. It is not clear that persistent Latino poverty will meet with the same acquiescence in the decades to come.
Elsewhere, Leonhardt has suggested that the fact that the U.S. median household income is somewhat lower than that of the Canadian median household income should be a source of serious reflection (an argument that overlooks the role of housing prices, but we’ll leave that aside). Surely the fact that less-skilled Latino immigrants and their children are struggling to close the gap with other Americans in a knowledge-intensive, service-oriented economy ought to be cause for reflection as well. As David Frum notes, Canada has an immigration policy that seeks to raise the average skill level of their workforce; the U.S. has an immigration policy that, in effect, reduces the average skill level of the workforce.
My view is that U.S. policymakers should do everything they can to facilitate upward mobility for all Americans, including Hispanic Americans. One way to achieve this goal is to shield less-skilled foreign-bon Hispanics currently residing in the U.S. from competition from new arrivals. As an added bonus, this will tend to encourage assimilation and intermarriage.
I wouldn’t say that America faces a “Hispanic challenge” because Hispanics are an integral part of American society. I would say, however, that Hispanics face unique challenges that are in many respects more daunting than those facing Italian immigrants who settled in the less-skilled America of the early twentieth century, when the relative labor market position of less-skilled men was (perhaps) the strongest its been in human history.