Ran Abramitzky, Stanford University | Leah Boustan, Princeton University

When large numbers of Italian immigrants came to the U.S. at the turn of the 20th century, they faced intense discrimination. But today, politicians from both the right and left have changed their tune, remembering Italians and other immigrants from Europe as dedicated members of American society who built the country, learned the language, and made it on their own.

Our understanding of how immigrants contribute to U.S. society is central to policymaking. But too often, that understanding is shaped not by hard evidence, but by pre-determined biases. This was true 100 years ago and remains true today.

For nearly ten years, we’ve been working on research that adds data to the debate. And now, in a new paper co-authored with Elisa Jácome of Princeton and Santiago Pérez of UC Davis, we use millions of father-son pairs spanning more than 100 years of U.S. history to show that immigrants today are no slower to move into the middle class than immigrants from 100 years ago.

Both today and in the past, many immigrants earn less than U.S.-born workers upon first arrival and do not completely catch up in a single generation. But their children do. No matter when their parents came to the U.S. or what country they came from, children of immigrants have higher rates of upward mobility than their U.S.-born peers. What’s more, their rates of mobility today are strikingly similar to rates of mobility in the past.

Importantly, the gap in mobility rates is much bigger for those growing up at the bottom of the income distribution. Children of first-generation immigrants growing up in the poorest 25 percent of the distribution end up near the middle as adults. These children of immigrants have rates of economic mobility that are 3–6 percentage points higher than their U.S. born peers. For those in the top quarter of the income distribution, the gap in mobility is about 1-5 percentage points.

Average adult income frank for children born to 25th percentile


Our research suggests that politicians crafting immigration policy shouldn’t be so short-sighted. Even immigrants who come to the U.S. with few resources or skills bring something that’s hugely beneficial to the U.S. economy: their children. Most importantly, our research suggests there’s room to be optimistic about the American Dream. For millions of families, coming to America can and does improve their children’s opportunities.

Our data

We compare three cohorts of immigrants. The first two cohorts account for four million immigrants—primarily from different parts of Europe—first captured by the 1880 or 1910 Censuses. Immigrants in the 1880 cohort are largely from Northern and Western Europe. Immigrants in the 1910 cohort are largely from Southern and Eastern Europe—a group thought to have faced greater initial disadvantages in the U.S. labor market. Using Census data, we’re able to follow their children for thirty years into adulthood. To study the mobility of these historical cohorts, we rely on occupation-based earnings computations.

The third cohort includes immigrants who first came to the U.S. around 1980, largely from poorer countries in Latin America and Asia. Thanks to data from Opportunity Insights, we are able to observe earnings data of fathers and sons in this cohort.

Children in the third cohort are the contemporaries of the Dreamers—the people who came to America as young children and are now seeking a path to citizenship–and entered the U.S. during an era of substantial migration policy restrictions. Importantly, however, our data on this cohort likely undercounts undocumented immigrants. The Opportunity Insights data includes only legal immigrants, and the GSS survey data we use undercounts undocumented immigrants to the extent that undocumented immigrants are less likely to respond to surveys.

Why are children of immigrants more upwardly mobile? Geography matters—a lot.

One way to interpret our findings is that the children of both poor and wealthy immigrants are more likely than their U.S.-born peers to achieve the American Dream. Why is this the case?

First, geography matters—a lot. Looking at our two historical cohorts, we find that immigrant parents are more likely than U.S. born parents to move to areas offering better prospects for their children. When we compare children growing up in the same U.S. region, the intergenerational gap between immigrants and the U.S.-born is reduced by more than 70 percent. When comparing children growing up in the same county, we no longer find an intergenerational gap between the children of immigrants and U.S.-born individuals. In other words, immigrant children did not earn more than others who grew up in the same location. Rather, their parents chose to live in parts of the Northeast, the upper Midwest, and some parts of the West that offered high mobility prospects to all. Immigrants in our historical cohorts were unlikely to settle in the South.

Secondly, we find it likely that many immigrant fathers—because of limitations in language, professional networks, or education—were “under-placed” in the income distribution. Their earnings likely didn’t fully reflect their abilities. But when their children are able to acquire the U.S.-specific labor market qualifications that their parents could not, they catch up quickly. In fact, we find the highest rates of upward mobility for children whose fathers immigrated to the U.S. as adults—and thus likely had more limitations in terms of language and professional networks, thereby giving their children more room to improve—compared to those whose fathers arrived as children.

Importantly, education does not explain the higher income mobility of second-generation immigrants in the past. The children of immigrants were not more educated than the children of U.S.-born individuals with comparable family income. Education may play a larger role in immigrant mobility today as the returns to schooling are now higher.

Additional reading

Between the two of us, and often with Katherine Eriksson of UC Davis, we have several additional studies on immigration that can add data to our policy debates:

In a study of workers that came to the U.S. during the Great Migration from 1880-1900, we find little evidence that the average immigrant, in the course of their lifetime, was able to move up the occupational ladder more rapidly than a U.S.-born worker. Even if there was some occupational growth for immigrants, whatever gap existed between immigrants and their U.S.-born peers when immigrants first arrived persisted over time (even if, in the case of some immigrants, immigrants held higher-paid occupation than their U.S.-born peers when they arrived). This pattern casts doubt on the conventional view that immigrants in the past who arrived with few skills were able to catch up with U.S.-born workers within a single generation.

In another study of the Great Migration, we find that men from wealthier families were somewhat less likely to migrate. We also compare the occupation-based earnings of Norwegian migrants to the occupation-based earnings of their brothers who remained in Europe. We estimate an economic return to migration of about 70 percent when compared to their brothers.

In a study of Norwegians who moved to the U.S. and back again to Norway from 1850–1913, we find evidence that even temporary movement to the U.S. improved the economic circumstances of the poor. The migrants who ultimately returned to Norway held lower paying jobs in the U.S. than other Norwegian migrants. Still, they were able to earn more after their return to Norway than Norwegians who never left, despite hailing from poorer backgrounds to begin with.

In a paper that examines the two main eras of U.S. immigration history—the Age of Mass Migration from Europe (1850–1913) and the recent period of renewed mass migration from Asia and Latin America—we present critical insights on who migrates, how they assimilate into the U.S. economy, and what effects immigration has on the U.S. labor market.

In a paper that studies the cultural assimilation of immigrants over the two main eras of U.S. immigration history, we document that immigrants assimilated into U.S. society at similar rates in the past and present. We measure cultural assimilation as immigrants giving their children fewer foreign names after spending more time in the U.S. and show that immigrants erase about one-half of the naming gap with natives after twenty years both historically and today. Immigrants from poorer countries choose more foreign names upon first arrival in both periods but are among the fastest to shift toward native-sounding names. We find substantial cultural assimilation for immigrants of all education levels.

For media inquiries, please contact Delaney Parrish (delaneyp@princeton.edu)