Immigration policy has emerged as a focal point of political debate in many advanced economies. Proposals to curtail immigration are frequently justified on both cultural and economic grounds. Advocates of restrictions often express concerns about national identity (New York Times 2025), and social cohesion. But economic arguments (The Economist 2025) are also prominent: immigrants are said to increase competition in local labour markets, depress wages for native workers (The Telegraph 2025), and strain public resources. Recent work has sought to evaluate the consequences of immigration restrictions on economic growth (Medici et al. 204) and labour market outcomes of native-born workers (Clemens et al. 2018, Abramitzky et al. 2023). This work has improved our understanding of the short- and medium-run labour market effects of immigration policy. Yet we know less about its implications for the next generation of native-born individuals. This gap in perspective matters, since immigration restrictions are persistent – once enacted, they tend to remain in place for decades.

Our recent paper (Feigenbaum et al. 2025) examines the long-run consequences of the US immigration quota acts of the early 1920s for the intergenerational mobility of US-born men. These laws abruptly ended the Age of Mass Migration – a period between 1850 and 1920 during which more than 30 million Europeans migrated to the US – and replaced it with a restrictive national-origins quota system that remained in force for over 40 years (Abramitzky and Boustan 2017). The quotas sharply curtailed immigration from Southern and Eastern Europe while preserving relatively high allowances for Northern and Western European countries. This policy shift generated a large, sudden, and geographically uneven change in both the scale and composition of immigration across the country, creating a useful setting to study the long-run effects of immigration restrictions.

Our analysis exploits the fact that immigrant communities were unevenly distributed across US counties before the quotas were introduced. Counties with larger pre-1920 populations of immigrants from heavily restricted origins experienced a larger decline in immigration after the laws were enacted. We summarise this variation using a county-level measure of ‘quota exposure’, which combines historical settlement patterns with predicted immigrant inflows in the absence of the policy change. Figure 1 shows the geographic distribution of this exposure measure after removing broad regional differences, so that the map highlights variation across counties within the same part of the country – the source of variation we rely on in the analysis. While exposure was higher in the Northeast, the Midwest, and parts of California, it varied substantially even within these regions, reflecting fine-grained differences in local immigrant settlement patterns.

Figure 1 Immigration quota exposure

Notes: The figure plots the immigration quota exposure, defined as the number of predicted missing immigrants, scaled by the 1900 county population, partialling out Census region fixed effects.

To study how these shocks affected economic mobility, we combine the quota exposure measure with newly linked individual-level census data. We link US-born sons from US-born parents observed in childhood in the 1900 and 1920 full-count censuses to their adult outcomes in 1920 and 1940, respectively. These data allow us to follow individuals across generations and locations while observing both family background and adult occupational outcomes. Our primary outcome is absolute intergenerational occupational mobility – i.e. the probability that a son attains a higher-status occupation than his father. We measure occupational status using harmonised occupation-based scores that proxy for long-run economic standing and adjust for differences across cohorts, races, regions, and nativity.

Figure 2 maps intergenerational mobility rates for US-born white and Black men who were children in 1920 and adults in 1940, after accounting for broad regional differences. By construction, the figure highlights variation in mobility across counties within the same region. Among white men, upward mobility is relatively high across the Midwest and Great Plains and in parts of the South, notably Texas, while it is lower in many industrial counties in the Northeast and Mid-Atlantic. For Black men, mobility is lower overall and especially low across large parts of the South, consistent with the constraints imposed by segregation and Jim Crow institutions (Althoff and Reichardt 2024). At the same time, the maps reveal substantial heterogeneity within regions for both groups.

Figure 2 Intergenerational mobility rates, 1920-1940

(a) US-born white men

(b) US-born Black men

Notes: The figure plots the raw absolute intergenerational mobility for the cohorts of US-born white (Panel A) and Black (Panel B) men who were children in 1920 and adults in 1940, after partialling out Census region fixed effects. Intergenerational mobility is defined as the probability that the son has a higher occupational rank than the father.

Our empirical approach compares changes in intergenerational mobility before and after the introduction of the quotas across counties with different levels of exposure. We ask whether children who grew up in places that lost more immigrants due to the quotas experienced different mobility outcomes than similar children other counties in the same region. We find that immigration restrictions reduced intergenerational mobility among US-born white men. In counties more exposed to the quotas, white sons were less likely to surpass their fathers in occupational status. A five-percentage-point increase in quota exposure – roughly the difference between a lightly and a heavily affected county – reduced the probability of upward mobility by about 1.9 percentage points, or nearly 4% relative to the mean. This magnitude is comparable to well-documented geographic differences in upward mobility across US regions and to the gains experienced by Black migrants during the Great Migration (Derenoncourt 2022).

For US-born Black men, the pattern is different. Quota exposure is associated with a positive but imprecisely estimated increase in intergenerational mobility. While these estimates are imprecise, the contrast in sign relative to white men is consistent across specifications and aligns with historical evidence on labour market competition (Collins 1997). Black workers and European immigrants – particularly those from Southern and Eastern Europe – were close substitutes in many low-skill urban occupations. By limiting immigrant inflows, the quotas may have modestly reduced competition in these segments, improving occupational prospects for Black men.

To shed further light on these long-run effects, we turn to the 1940 Census, the first to record individual earnings and hours worked in a systematic way. We examine labour market outcomes for the same cohort of men who were children when the quotas were introduced. The patterns mirror the mobility results. US-born white men from more exposed counties earned significantly lower wages in adulthood. A five percentage point increase in quota exposure reduced weekly wages by about 2.6% and hourly wages by roughly 2.3%. Given that real wages for white men rose by around 10–20% between 1920 and 1940 (Carter et al. 2006), these losses represent a non-trivial share of overall wage growth. For Black men, wage and employment effects are again small and imprecisely estimated, and generally positive in sign. While we do not emphasise these estimates on their own, the overall pattern is consistent with the intergenerational mobility results.

Why would restricting immigration slow mobility for native-born whites? We consider three broad mechanisms. First, immigration restrictions might have induced substitution from other sources – such as internal migrants or immigrants from unrestricted countries – thereby offsetting the loss of European labour. We find evidence of such substitution, particularly in urban areas, but it was incomplete at the county level and cannot explain the magnitude of the effects. Second, restrictions might have reduced incentives to invest in education by lowering returns to skill. We find that exposure to the quotas reduced educational attainment among white men, but the magnitudes are modest and can explain only a limited share of the observed wage losses.

The evidence points to a third mechanism: immigrant–native complementarities. A large body of research shows that immigrants can raise the productivity of native workers by enabling specialisation and occupational upgrading (Ottaviano and Peri 2012). We show that the negative effects of the quotas were larger in counties that lost more immigrants who were less English-proficient or more concentrated in manual and industrial occupations – groups that historically complemented native-born white workers. When these complementary workers disappeared, productivity, wages, and mobility declined.

Our findings indicate that immigration policy can have long-lasting effects on economic opportunity for native-born populations – and that these effects are unevenly distributed. While much of today’s debate focuses on short-run wage competition, our results suggest that restricting immigration can reduce productivity and slow upward mobility for groups that benefit from complementarities with immigrant labour. At the same time, the experience of Black Americans in the early 20th century reminds us that immigration policy also reshapes the distribution of opportunity across groups. Policies that harm some may leave others unaffected or even slightly better off, depending on patterns of substitution and complementarity in local labour markets. In a diverse society like the US, immigration policy can shape who gains access to opportunity – and who does not – over generations. Understanding these long-run and distributional effects is essential for informed policy design today.

References

Abramitzky, R and L Boustan (2017), “Immigration in American Economic History,” Journal of Economic Literature 55 (4): 1311–1345.

Abramitzky, R, P Ager, L Boustan, E Cohen, and C W Hansen (2023), “The Effect of Immigration Restrictions on Local Labor Markets: Lessons from the 1920s Border Closure,” American Economic Journal: Applied Economics 15(1): 164–91.

Althoff, L and H Reichardt (2024), “Why history continues to shape racial inequality in the US”, VoxEU.org, 2 April.

Carter, S B, S S Gartner, M R Haines, A L Olmstead, R Sutch, and G Wright (2006), Historical Statistics of the United States: Earliest Times to the Present, Millennial Edition, Cambridge University Press.

Clemens, M A, E G Lewis, and H M Postel (2018), “Immigration Restrictions As Active Labor Market Policy: Evidence from the Mexican Bracero Exclusion,” American Economic Review 108(6): 1468–87.

Collins, W J (1997), “When the Tide Turned: Immigration and the Delay of the Great Black Migration,” The Journal of Economic History 57(3): 607–632.

Derenoncourt, E (2022), “Can You Move to Opportunity? Evidence from the Great Migration,” American Economic Review 112(2): 369–408.

Feigenbaum, J, Y J Hung, M Tabellini, and M Tomasella (2026), “Immigration Restrictions and Natives’ Intergenerational Mobility: Evidence from the 1920s US Quota Acts,” NBER Working Paper 34775.

New York Times (2025), “A Progressive Future Depends on National Identity,” 19 June.

Medici, C, N Qian and M Tabellini (2024), “The impact of the Chinese Exclusion Act on the economic development of the Western US”, VoxEU.org, 1 December.

Ottaviano, G IP and G Peri (2012), “Rethinking the Effect of Immigration on Wages,” Journal of the European Economic Association 10(1): 152–197.

The Economist (2025), “Your Guide to the New Anti-Immigration Argument,” 13 March.

The Telegraph (2025), “Surge in Immigration Risks ‘Pushing Down Wages’ for Low-Paid Britons,” 15 April.