More Evidence on Third-Generation Outcomes

Grandchildren of Mexican immigrants compared to grandchildren of European immigrants

By Jason Richwine on May 28, 2019

Will the children and grandchildren of low-skill immigrants eventually rise to the same socioeconomic level as natives? In a report published last fall, I investigated this question using the NLSY-97, a survey of people born between 1980 and 1984 that includes their grandparents' places of birth. The grandparent information helps identify a true "third generation," meaning U.S.-born people who have two U.S.-born parents but at least one foreign-born grandparent.

Because the largest and most consistently low-skill immigrant group has come from Mexico, my report compared the grandchildren of Mexican immigrants to a reference group of white Americans from the "fourth-plus generation" – meaning U.S. born with two U.S.-born parents and four U.S.-born grandparents. The results indicated that Americans with at least one Mexican-born grandparent lag significantly behind on measures of education and income. In other words, assimilation of this initially low-skill group is still not complete by the third generation.

After the Great Wave of immigration ended in the 1920s, Americans developed some romantic notions about assimilation. No matter where immigrants come from, no matter what skills they bring with them, no matter what circumstances they find themselves in upon arrival, their children and grandchildren will supposedly converge to the socioeconomic level of the pre-existing population. Desirable as that outcome may be, the convergence is often incomplete. The results for third-generation Mexican Americans described above are perhaps the starkest illustration.

Differential levels of assimilation are also evident when comparing the grandchildren of immigrant groups who arrived in the same time period. After the U.S. and Mexico, the most common grandparent place of birth in the NLSY-97 is Europe. (Unfortunately, no specific countries in Europe are identified in the data.) This post provides the results of a new analysis comparing two third-generation groups -- the grandchildren of immigrants from Mexico, and the grandchildren of immigrants from Europe.

Based on parental data from the NLSY-97 and year-of-arrival data from the 1970 census, most grandparents of the NLSY-97's European third generation arrived in the U.S. between 1910 and 1950. Unlike Mexican immigrants, who were almost uniformly low-skill, European immigrants in that time frame were more mixed. They include largely low-skill Southern and Eastern European immigrants who arrived before the 1924 restriction, but also some educated refugees from Central Europe during the Nazi period, along with both skilled and unskilled immigrants from the post-war era.

The table below compares the grandchildren of Mexican immigrants and the grandchildren of European immigrants on measures of educational attainment, test scores, work time, and income. Although the two groups graduated from high school at about the same rate, the grandchildren of European immigrants have more than double the rate of college completion. They also scored higher on the AFQT, which the military uses to assess math and verbal skills. Similarly, although weeks worked are roughly equivalent for both groups, the grandchildren of European immigrants significantly out-earn their counterparts with Mexican-born grandparents.

On most measures, the European third generation even slightly outperforms the reference group of fourth-plus generation whites. Clearly, not all immigrant groups end up in the same place by the third generation.

For details on the data set and the calculations, please see my report from last fall. Also note that for simplicity and sample size considerations, the ethnic and cross-sectional samples of the Mexican third generation are combined in the table above.

Topics: Education