National Review, December 3, 2015
The National Academies of Sciences, Engineering, and Medicine (NAS) recently published a nearly 500-page report on the assimilation of immigrants: The Integration of Immigrants into American Society. The New York Times, the Wall Street Journal, and others have cited the study as proof that today's wave of immigration is working out just fine.
Authored by many of the leading sociologists who study immigration, the report is frustrating to read because it suffers from a one-sided perspective that is unfortunately all too common among American academics.
If one had to summarize the authors' perspective, it would be that despite rampant discrimination, misguided efforts at immigration enforcement, and declining wages for unskilled workers, the nation's 78 million immigrants and their children, through their own Herculean effort, are doing reasonably well. Nonetheless, the report implicitly and explicitly implies that immigrants need more access to welfare, the U.S. should spend more on urban education, and we should give amnesty to illegal immigrants. For the authors, any suggestion that a lower level of immigration might help avoid problems in the future or facilitate assimilation is not worth considering.
The report focuses on such topics as the educational attainment, income, language, citizenship, crime rates, and geographic dispersion of immigrants. The study does not address important aspects of assimilation that might interest conservatives most, such as self-sufficiency and the use of welfare programs, as well as what John Fonte of the Hudson Institute calls "patriotic assimilation."
The problem with the study is both what it leaves out and how it chooses to present its findings. There is no question, for example, that immigrants' income and other measure of socioeconomic status rise the longer they live in the country. But they are still on average much worse off than natives, even when they have lived in the country for decades. The NAS study tends to emphasize progress, rather than the large, persistent gaps between immigrants and natives.
The authors also never discuss some of the things that immigrants do that might hinder their assimilation. Consider that the study shows that Americans of Mexican ancestry in the second generation (born in the United States to immigrant parents) and third generation-plus (born in the United States to American-born parents) remain poorer and less educated than non-Mexican natives. In fact, there appears to be no progress between the second and third generations. This is very important because roughly one-third of all children born today to immigrants (legal and illegal) are born to a Mexican mother.
One of the key explanations offered by the NAS study for the lack of progress is that "high levels of undocumented immigration across generations" and "racial and ethnic discrimination" and "poor quality schooling" are holding the group back. But the authors never seriously consider culture. Thomas Sowell and others have made the point that some groups have historically placed more emphasis on education. Even the New York Times has reported on the topic, analyzing in one piece, "What drives success?" While the role of culture in hindering or helping assimilation receives almost no attention in the NAS study, the word "discrimination" appears 63 times.
The NAS rightly devotes a good deal of attention to educational attainment, because it is a key predictor of economic performance. Furthermore, parental education is a good predictor of children's educational attainment. We should note that in general Asian immigrants and the small number who still come from Europe have as high or higher levels of education as natives do, so in these groups there is no education or even an income gap to close.
Many of today's adult immigrants who come from regions other than Asia, however, have low levels of education, and this partly explains their high poverty rates and welfare use. For these immigrants, the study makes the case for the "rapid educational integration" of the second generation — except for Mexicans. Take, for instance, Central American immigrants: Of men 25 to 59 years of age, the NAS reports that 48 percent have not completed high school and just 10 percent have a bachelor's degree. In contrast, just 7 percent of natives have not completed high school, and 36 percent have a college degree. But the education level of second-generation men (born in the United States to Central American immigrants) is similar to natives: Just 8 percent have not graduated high school, and 26 percent have at least a bachelor's degree. While encouraging, this may be misleading.
The education statistics reported in the NAS for second-generation adults ages 25 to 59 are, of course, not for the children of today's immigrants, who are still mostly children. The way the NAS reports the second generation's education levels indicates that most of their parents came before 1970. But in 1970, the Central American immigrant population was tiny and actually more educated than the native-born population. In 1970, 38 percent of Central American men had some college education, compared with only 29 percent of native-born men, and they were also more likely to have completed high school than natives — 40 percent versus 36 percent.
By the 1980s, however, the Central American population in the U.S. had exploded and became and remains among the least-educated. It is not at all clear that the children born to these more recent immigrants will do as well as the children born to those who arrived decades ago. To the extent that we can look at the children of today's Hispanic immigrants, things are worrisome. Second- and third-generation Hispanic students, for example, score well below the national average on standardized tests such as the Program for International Student Assessment.
Another problematic area of the study is its comparison of Great Wave immigrants (roughly those who arrived 1880 to 1920) to immigrants today. Based primarily on research by Joel Perlmann, the NAS authors argue that Great Wave immigrants were significantly poorer than natives. The clear implication is that immigrants in the past were poor and things turned out okay, so don't worry about all the poor immigrants today. There are a number of problems with this formulation.
First, as the report itself makes clear, times have changed. In the past "there was an abundance of jobs that paid a family wage for men." But today, unskilled workers face declining wages, and they often struggle. Second, there was no welfare state back then. Third, there is actually no systemic wage data from that time. The Census did not include an income question until 1940. Any conclusion about Great Wave immigrants' relative economic position is tentative at best.
But the most glaring omission in the NAS study with regard to the Great Wave is that in a study devoted to assimilation, the NAS never discusses the role that the post-1914 slowdown in immigration may have played in facilitating assimilation. WWI and then the restrictive legislation passed in 1921 and 1924 caused the immigrant share of the U.S. population to decline from almost 15 percent in 1910 to less than 5 percent by 1970. This reduced job competition for immigrants already here, and it prevented immigrant communities from being refreshed with new arrivals. Although arguably the most important event in American immigration history, the NAS study is entirely silent on its possible impact on assimilation.
We see more evidence that the study authors are strongly predisposed to favor immigration in how they describe the "need" for immigrant labor. The report states that today's unskilled immigrants face "declining wages for unskilled workers." But later, they argue that "demand for low-skilled labor remains high." The only way both can true is if the supply of workers is so large that it overwhelms the strong demand. The NAS study does not acknowledge this contradiction, but it does cite research showing that it is the illegality of illegal immigrants that puts downward pressure on wages. The clear implication is that amnesty would solve this problem.
However, legalized illegal immigrants would still account for roughly one-quarter of the nation's high-school dropouts. If one really wants to help the poorest workers, then enforcing the law and making a significant share of illegals go home would reduce the supply of unskilled labor and allow wages to rise.
The NAS report's whole discussion of illegal immigration tells us as much about how the authors think about the issue as it does about the issue itself. For the authors of the report, being illegal is a condition in which immigrants often "find themselves." But the passivity of that formulation is false; virtually all adult illegal immigrants knowingly overstayed a visa or purposefully snuck across the border.
On the related issue of immigration enforcement, the NAS discussion is also grossly incomplete. They cite at length and without question a 2013 Migration Policy Institute (MPI) study that mistakenly reported that $18 billion is spent annually on "immigration enforcement." Yet, as the very names of the agencies involved make clear, much of what they do is customs enforcement, not immigration enforcement: Immigration and Customs Enforcement (ICE); Customs and Border Protection (CBP). The MPI study was so flawed that the Washington Post published my colleague Mark Krikorian's critique of it.
The NAS study also states that "deportations have skyrocketed" without explaining that under President Obama, actual deportations (technically known as removals) from within the United States have declined steadily. To obscure this fact, the administration counted some people caught at the border, but processed within the United States, as interior removals. A 2014 Los Angeles Times headline summed things up nicely: "High deportation figures are misleading." As acting head of ICE John Sandweg made clear, "If you are a run-of-the-mill immigrant here illegally, your odds of getting deported are close to zero." But none of this comes up in the NAS study.
The NAS discussion of crime is also markedly incomplete. The study goes to considerable lengths to argue that immigrants do not commit a disproportionate share of crime, which is probably correct. But they ignore any evidence to the contrary, such as this analysis and this one and this one. They also make the mistake of citing studies based on the 2000 and 1990 censuses showing low immigrant-incarceration rates, unaware that the data are unreliable.
One of the biggest questions in studying crime is whom you compare immigrants and their children to. Racism and social problems in the black community make the group unique when it comes to crime. As a result, in 2013, the incarceration rate for all native-born men (ages 18 to 40) was 31 per 1,000, but it was only 18 per 1,000 when blacks were excluded. This compares with 23 per 1,000 for Mexican immigrants and 38 per 1,000 for U.S.-born Mexican Americans. But the NAS study ignores this, leaving the reader with an incomplete grasp of the issue.
The narrow way the NAS approaches immigration is front and center when the study argues that high poverty rates among immigrants are a particular concern "because many immigrants are barred from participation in social welfare programs." The impacts on schools and taxpayers, or even that it makes us a more unequal society, are apparently not a concern. Welfare restrictions on immigrants have only a modest impact. In 2012, 51 percent of immigrant households used at least one welfare program, compared with 30 percent of native households.
Despite all its problems, the NAS report still contains a great deal of valuable information. But it could have been a much more useful report if it had been presented in a less one-sided manner.