Immigrants’ wages in the US

Published on September 12, 2019 Updated on December 9, 2019

on the September 12, 2019

Focus on Research

Immigrants’ wages in the US

Higher-educated foreign-born workers receive a significant wage premium on the US labor market, but unobserved individual-specific factors determine a huge variability in wages within each educational group. The resulting substantial overlap in the distribution of wages for high- and low-educated workers suggests that points-based system centered on education might fail to improve migrants’ quality.

Individuals that move across borders can face major difficulties in obtaining an adequate return for their human capital in the labor market of the destination country. The economics literature has explored the extent to which immigrants hold jobs for which they are over-qualified, a phenomenon that has been termed brain waste (Mattoo et al., 2008). Brain waste might reflect difficulties in ensuring the portability of immigrants’ education credentials or rather signal the low quality of the education that has been acquired prior to migration.

Bertoli and Stillman (2019) rely on data from various waves of the American Community Survey to analyze the relationship between (foreign) education and immigrants’ wages on the US labor market. Considering immigrants that are either (at most) high school dropouts or have at least four years of college education, they find that, for the vast majority of origin countries (95 out of 114 countries of origin), high-educated immigrants enjoy a sizeable and statistically significant wage premium. However, what is striking in the data is the extent to which this significant difference in mean wages co-exists with a very large variance in wages within each of the two educational groups. Large within-group variance leads the two wage distributions to overlap, often to a substantial extent.

Bertoli and Stillman (2019) propose the following thought experiment to measure the origin-specific extent of the overlap of the wage distributions of high- and low-educated immigrants in the US: for each country of origin, they randomly draw a high- and a low-educated immigrant, and then compare their wages. In the absence of overlap, the wage for the low-educated immigrant in this pair should always fall short the wage for the high-educated immigrant, while if the two distributions perfectly coincided, then the low-educated immigrant should receive a higher pay in 50 percent of the random draws. This thought experiment reveals that, on average, a low-educated immigrant receives a higher wage than an otherwise similar high-educated immigrant in 24.7 percent of the cases. The large share of pairs with an unexpected ranking of the two immigrants (notice that there is at least a 5 year difference in education between the two immigrants) with respect to their wages is also the case in origin countries with a large and significant difference in mean wages between the two groups. Furthermore, for 98 out of 114 countries, this measure of overlap exceeds the corresponding measure for the native workers, which stands at 13.8 percent.

What are the implications of this basic observation? The great deal of variability in wages within each education group that we find suggests that an increase in the extent to which immigrants to the United States are selected on the basis of their education might fail to increase their average productivity. Even though there is significant difference in average wages between high- and low-educated immigrants, the effects of an increase in education selectivity on immigrants’ quality are determined by the difference in the wages of the marginal worker in each of the two groups (Bertoli et al., 2016), i.e., the workers whose location decisions would be modified by the policy change. With a substantial overlap in the two wage distributions, an increase in education selectivity would often bring in a high-educated worker whose productivity (and hence wage) is below the wage of the marginal low-educated worker that is discouraged from migrating to the United States.

The importance of unobserved determinants of wages, i.e., factors that cannot be observed by an immigration officer, suggests that we should be cautious about the ability of points-based systems to raise immigrants’ quality, as measured by their wages.


Bertoli, S., Dequiedt, V. and Y. Zenou (2016). “Can selective immigration policies reduce migrants' quality?" Journal of Development Economics 119, 100-109.

Mattoo, A, Neagu, IC, Özden, Ç. (2008). “Brain waste? Educated immigrants in the US labor market.Journal of Development Economics 87(2), 255-269.