One of the best predictors of how adult immigrants will do in the United States is how much education they have when they arrive. This makes perfect sense, since educational attainment is one of the best predictors of success among natives as well. A number of recent studies, such as this one and this one, show that the education level of new arrivals has increased dramatically in recent years, both in absolute terms and relative to natives. This of course should mean that new immigrants are working more, earning more, and avoiding poverty and welfare use compared to their counterparts from a decade ago. But in a new analysis, my colleague Karen Ziegler and I found this not to be the case. In short, relative to natives, immigrants are starting out as far behind in 2017 as they were a decade ago, even though they are coming with a great deal more education.
In the discussion that follows, I will focus on new arrivals who were between the ages of 25 and 65 and had resided in the country for five years or less at the time of the survey. These are people who for the most part received their education overseas. Of newly arrived immigrants, the share with at least a bachelor’s degree increased from 34 percent in 2007 to 49 percent in 2017. Over the same period, the share without a high-school diploma fell from 34 percent to 16 percent. In other words, the people who arrived in the five years prior to 2017 were much more educated than the people who arrived in the five years prior to 2007.
The education of natives also increased, but not nearly as much. What caused the dramatic increase in immigrant education? It was partly due to a decline in the share of new arrivals who are illegal immigrants, who tend to be the least educated, and a rise in the share from East and South Asia, who tend to be the most educated. There has also been a general increase in education in all the primary immigrant-sending regions, including Latin America.
The dramatic increase in immigrant education relative to natives reverses a longstanding trend that dates back to at least 1970. From 1970 to at least 2000 — and possibly all the way to 2007 — each new wave of immigrants was less educated, relative to natives, than the wave that had come before. (Education levels did increase in absolute terms for both immigrants and natives.) This decline caused a significant deterioration in the relative economic standing of new immigrants and of immigrants generally. Probably the best discussion of this decline and its consequences can be found in this study and in an earlier study, both by the National Academies of Sciences, Engineering, and Medicine.
The educational decline has ended, but the improvement has not led to economic successes. The share of new immigrants who were in the labor force (working or looking for work) was 73 percent in 2007 and 67 percent in 2017. Native labor-force participation also fell, from 78 to 76 percent, so the gap with new immigrants widened. The median income of new arrivals at arrival was about half that of natives in both 2007 and 2017. The share of adult new immigrants in poverty was slightly higher in 2017 than in 2007, and the gap with natives widened slightly. Overall, new immigrants remained about twice as likely to live in poverty as natives, despite their rise in education.
Another area where the dramatic increase in new immigrants’ education did not result in a dramatic improvement in their situation is welfare use. In general, welfare use has increased across American society over the last decade. But the increase among new immigrants has been one of the most profound. Consider Medicaid, the health-insurance program for the poor. In 2007, 6 percent of new adult immigrants were on Medicaid; by 2017 it was 17 percent. The share of natives on Medicaid increased from 7 percent to 13 percent. New immigrants are now more likely to use the program than natives. Technically, under federal law, new (non-refugee) green-card holders, temporary visitors, and illegal immigrants are not supposed to use this program, so use among new arrivals should be very low. But there is some fraud in the program; some exceptions exist under federal law; and some states on their own give Medicaid to the otherwise ineligible.
Most socioeconomic measures were no better in 2017 than in 2007 for both new immigrants and natives across all education levels. This is an indication that both immigrants and natives were impacted by the Great Recession and in some ways have not fully recovered. But it does not explain why the dramatic increase in immigrant education relative to natives did not result in a substantial narrowing of the gap between the two groups.
To be sure, educational attainment has been and remains one of the best predictors of success in modern America, for immigrants and natives alike. But the dramatic increase in the educational attainment of new arrivals did not result in a corresponding improvement in measures of socioeconomic well-being, because performance declined in each educational category. In other words, highly educated new immigrants do worse than highly educated new immigrants from a decade ago, and less-educated new immigrants underperform their less-educated counterparts from a decade ago as well.
Highly educated new immigrants do worse than highly educated new immigrants from a decade ago, and less-educated new immigrants underperform their less-educated counterparts from a decade ago as well.
Could these trends be explained by illegal immigrants’ and refugees’ underperforming their education, given their special circumstances? Probably not. All the evidence indicates that the number of new illegal immigrants arriving annually is lower now than a decade ago. As a result, illegal immigrants are a smaller share of new arrivals. (Of course, illegals are still coming — overstaying visas or sneaking across the border. But the numbers are not what they were.) It is also unlikely that an increase in refugee resettlement explains what happened. Refugees are only 3 percent to 6 percent of all new arrivals, and excluding the primary refugee-sending countries makes almost no difference in the data. While there has been an increase in the number of long-term temporary visitors allowed into the country, a large share are under age 25, such as foreign students, and are excluded from our analysis. In fact, the government estimates that the number of long-term temporary-visa holders in the country over age 25 was no higher in 2015 than in 2008, so it seems unlikely that the inclusion of temporary visitors in the data accounts for our findings.
Perhaps the most important factor that explains what happened is the economy. The lower share of new immigrants who are working or even looking for work may be an indication that the economy does not absorb new entrants as easily as it did a decade ago. Another factor that may help to explain what happened is that immigrants are not as educated as they seem to be. For example, the share of recent immigrants with a college education who were employed fell from 67 percent to 63 percent between 2007 and 2017, while it fell just one percentage point for native-born college graduates. And the share of college-educated immigrants in poverty was about 16 percent in 2017 and about 12 percent in 2012, while the share of college-educated natives in poverty was just one percentage point higher in 2017 than in 2012. This may be an indication that the actual marketable skills of new college-educated immigrants are not as high as they were in the past.
Another factor that likely accounts for some of the lack of convergence between new immigrants and natives is that the share of new arrivals who are women increased between 2007 and 2017. This does not seem to explain why the gap between new immigrants and natives did not narrow in terms of labor-force participation or welfare use. In fact, both male and female immigrants saw their labor force participation fall and their welfare use rise. However, it does help to explain the lack of convergence with regard to income. Immigrant men’s income did rise significantly relative to native men, while immigrant women’s income did not increase.
The causes for our findings are clearly in need of further research. In terms of public policy, the failure of increased immigrant education to translate into increased immigrant prosperity means that if we want immigrants who are likely to find jobs, have high incomes, pay a lot in taxes, and not use welfare programs, then selecting them based solely on their education may be insufficient to ensure this outcome. There are, of course, many possible goals of immigration policy, and a flow of immigrants who are able to earn high incomes or have low welfare use is not the only thing to consider when formulating policy. But if we want immigrants who are likely to be successful, we may have to look at other marketable skills, such as technical training and knowledge of English.