VoxEU Column Migration Politics and economics

The seeds of ideology: Historical immigration and political preferences in the US

Immigration to Europe and the US has met with a heated political backlash in recent decades, but the long-term impact of immigration on political ideology not well understood. This column focuses on the migration of millions of Europeans to the US between 1900 and 1930, and finds that the historical presence of European immigrants encouraged a more liberal political ideology and stronger preferences for redistribution among the native-born population. The difference is due in part to inter-group contact, which allows for the transmission (or ‘melting’) of immigrants’ experiences and ideas.

In recent years, immigration has become a potent political and social issue around the world, a trend reflected in the rise of anti-immigrant, right-wing parties that paint the lack of immigrant assimilation as a vital threat to host societies. While a large and growing body of work has examined the short run political impact of immigration (Dustmann et al. 2019, Halla et al. 2017, Tabellini 2020), much less is known about its effects on the native population’s political ideology in the long run. And yet, the long run effects of immigration on natives’ political preferences can vastly differ from their short run counterpart. First, consistent with the ‘contact hypothesis’ (Allport 1954), natives can change their attitudes towards minorities after prolonged interactions, which may gradually eliminate initially negative stereotypes. Second, although the immigration literature typically views the process of assimilation as one sided – with immigrants converging towards the new, local culture (Abramitzky et al. 2020, Fouka 2020, Lazear 1999) – it is possible for immigrants’ culture to spill-over into the native population, eventually creating a diverse or ‘melting pot’ society.

Our recent work (Giuliano and Tabellini 2020) explores these ideas by studying the long run effects of the 1900-1930 migration of millions of Europeans on the political ideology of US-born individuals today, which we measure using nationally representative survey data from the Cooperative Congressional Election Study (CCES). Focusing on this historical period, also known as the Age of Mass Migration, has a number of advantages.1 First, much like today, natives’ concerns about the lack of assimilation loomed large (Abramitzky and Boustan 2017, Higham 1955). Second, the composition of immigrants changed dramatically during these 30 years, allowing us to leverage variation in the cultural background and political preferences of different European groups. Third, by focusing on the Age of Mass Migration, we can estimate the effects of immigration on American ideology over more than a century.

The 1900-1930 decades also represent an almost ideal ‘quasi-natural’ experiment to causally identify the effects of immigration (Abramitzky and Boustan 2017, Abramitzky et al. 2019a, Tabellini 2020). Between 1910 and 1930, immigration from different European countries was differentially impacted by nationwide shocks – WWI and the Immigration Acts – that were arguably unrelated to cultural, political, or economic conditions prevailing in individual US counties.2 Because immigrants tend to concentrate in areas with larger ethnic enclaves, the differential effect of these shocks across European countries generated significant variation in the number as well as the ‘cultural mix’ of immigrants received by US counties between 1910 and 1930. Such variation is displayed in Figure 1, which plots the average immigrant share across counties between 1910 and 1930.

Figure 1 Average 1910-1930 immigrant share

Notes: 1910-1930 average share of European immigrants (over county population).

Source: Authors’ calculations from IPUMS sample of US Census (Ruggles et al. 2020).

Our main findings, reported in Figure 2 below, indicate that US-born respondents who today live in counties with higher historical immigration are significantly more likely to vote for the Democratic Party and to support more generous welfare spending. These effects are quantitatively large, and comparable to those of key determinants of political preferences in the US. For instance, the impact of a five percentage point increase in the average immigrant share is roughly equivalent to the estimated impact in other work for the effects of race or of moving from an income of $100,000 to $20,000 US dollars per year (Alesina and Giuliano 2011). We obtain similar findings for several other proxies of left-leaning political preferences and preferences for redistribution, such as party identification or support for an increase in the minimum wage.

Figure 2 Effect of historical immigration on political ideology

Notes: The y-axis reports individuals’ response to whether they voted for a Democratic candidate in the last presidential elections and their preferences for redistribution, respectively. The x-axis refers to the instrumented 1910-1930 average fraction of immigrants in the county. The scatterplots pool observations into 30 bins. Each point in the scatter diagram represents the residuals of the two variables, after partialling out state and survey wave fixed effects, historical county controls, and individual characteristics.

Source: adapted from Giuliano and Tabellini (2020).

Our results run counter to those of most studies on the short run effects of immigration and diversity (Alesina et al. 1999, Dahlberg et al. 2012, Tabellini 2020), which find a negative relationship between immigration or diversity and a native population’s preferences for redistribution and left-leaning ideology. We examine the possible reasons for this difference in the second part of the paper, in which we explore the mechanisms through which historical immigration moved natives' political ideology to the left. We start by ruling out the possibility that our findings are due to the characteristics of the counties where European immigrants initially settled, which may have persisted, influencing natives’ preferences in the long run. Next, we examine a number of economic channels. First, we argue that direct income effects are unlikely to explain our results: since European immigration increased income per capita in both the short and long run (Sequeira et al. 2020), if anything, this should have reduced natives’ preferences for redistribution (Meltzer and Richard 1981). Second, we verify that none of the immigrants’ economic characteristics – such as their education, skills, occupation, income, or historical intergenerational mobility in the US – influences our findings. Moreover, to the extent that immigrants were ‘selected’, one would expect that the more individualistic moved to the US and eventually remained there (Abramitzky et al. 2019b, Beck Knudsen 2019).

Having ruled out a number of economic mechanisms, we turn to our most preferred interpretation. We argue that immigration left its footprint on American ideology via cultural transmission from immigrants to the native-born, as immigrants brought with them their preferences for the welfare state, which in turn were transmitted to individuals born in the US. To test this hypothesis, we construct an index of exposure to historical social welfare reforms in immigrants’ countries of origin that counts the number of years elapsed between the introduction of the reform and the year in which an individual emigrated. After validating with modern data from the European Social Survey that this index indeed proxies for preferences for redistribution, we exploit variation in the ‘cultural mix’ of immigrants received by different counties over time. Consistent with our conjecture, after controlling for the direct effect of immigration, we find that higher exposure to social welfare reforms in the immigrants’ countries of origin is predictive of stronger preferences for redistribution and a more liberal ideology today. As shown in Figure 3, when splitting the sample in counties with historical exposure to welfare reforms above and below the median, we find the effects of immigration are significantly stronger in counties with higher values of the index.

Figure 3 Effect of immigration, by exposure to education reforms

Notes: The bars report the 2SLS coefficient on historical immigration (with corresponding 95% confidence intervals) on political ideology and preferences for redistribution of US born individuals, split for counties with exposure to education reforms above (resp. below) the sample median in orange (resp. blue).

Source: adapted from Giuliano and Tabellini (2020).

If preferences for redistribution of US born individuals were, at least partially, driven by immigrants, results should be stronger in counties where immigrants had more frequent contacts with the native-born population. To corroborate this interpretation, we show that the effects of immigration on natives’ ideology are stronger in counties where, historically, intergroup contact – proxied by inter-marriage and residential integration – was higher, and where immigrants’ preferences could thus be more easily transmitted to those born in the US.

In the last part of the paper, we trace out the dynamics of the political realignment induced by the arrival of immigrants. We show that the 1910-1930 average immigrant share in a county is not significantly associated with the Democratic vote share prior to 1928. However, a large upward jump occurs in the presidential elections of that year, when Alfred Smith – a Roman Catholic with an immigrant background – was the candidate for the Democratic Party, and the Republican candidate, Herbert Hoover, ran on a platform of ‘rugged individualism’ (Bazzi et al. 2017). The 1928 increase in support for the Democratic Party persists, suggesting that the initial political mobilisation of immigrants (Andersen 1979) was an important factor behind the positive association between historical immigration and left-leaning ideology today.

Findings in this paper highlight the importance of distinguishing between the short and long run effects of diversity and immigration on political preferences and ideology in receiving countries. Although immigrants might be opposed, generate backlash, and reduce the native population’s preferences for redistribution upon arrival, they can eventually lead to higher social cohesion and stronger desire for generous government spending. Moreover, our results indicate that immigrants’ assimilation is not a one-sided process. Instead, immigrants’ preferences might spill over and be transmitted to the native population, thereby contributing to a diverse and complex culture, and to the development of a ‘melting-pot’ society.


Abramitzky, R, L P Boustan and Katherine Eriksson (2020), “Do Immigrants Assimilate More Slowly Today Than in the Past?”, American Economic Review: Insights, 2(1): 125-141.

Abramitzky, R et al. (2019), “The Effects of Immigration on the Economy: Lessons from the 1920s Border Closure”, NBER Working Paper 26536.

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

Abramitzky, R, L P Boustan and K Eriksson (2019), "To the new world and back again: Return migrants in the age of mass migration", ILR Review, 72(2): 300-322.

Alesina, A, R Baqir and W Easterly (1999), “Public Goods and Ethnic Divisions”, Quarterly Journal of Economics, 114(4): 1243-1284.

Alesina, A and P Giuliano (2011), “Preferences for redistribution”, Handbook of social economics, Elsevier, Vol. 1: 93–131.

Allport, G W (1954), The Nature of Prejudice, Oxford: Addison-Wesley, 1954.

Andersen, K (1979), The creation of a Democratic Majority 1928-1936, University of Chicago Press: 1979.

Bazzi, S, M Fiszbein and M Gebresilasse (2017), “Frontier Culture: The Roots and Persistence of ‘Rugged Individualism’ in the United States”, NBER Working Paper 23997.

Dahlberg, M, K Edmark and H Lundqvist (2012), “Ethnic diversity and preferences for redistribution”, Journal of Political Economy, 120(1): 41–76.

Dustmann, C, K Vasiljeva and A P Damm (2010), “Refugee migration and electoral outcomes”, Review of Economic Studies, 86(5): 2035–2091.

Fouka, V (2020), “Backlash: The Unintended Effects of Language Prohibition in US Schools after World War I”, The Review of Economic Studies, 87(1): 204-239.

Giuliano, P and M Tabellini (2020), “The Seeds of Ideology: Historical Immigration and Political Preferences in the United States”, NBER Working Paper 27238.

Goldin, C (1994), “The political economy of immigration restriction in the United States, 1890 to 1921”, The regulated economy: A historical approach to political economy, University of Chicago Press: 223–258.

Halla, M, A F Wagner and J Zweimüller (2017), “Immigration and voting for the far right”, Journal of the European Economic  Association, 15(6): 1341–1385.

Higham, J (1955), Strangers in the Land: Patterns of American Nativism, 1860-1925, Rutgers University Press.

Knudsen, A S B (2019), “Those Who Stayed:  Selection and Cultural Change during the Age of Mass Migration”, Working Paper.

Lazear, E P (1999), "Culture and language", Journal of political Economy, 107(S6): S95-S126.

Meltzer, A H and  F Richard (1981), “A rational theory of the size of government”, Journal of political Economy, 89(5): 914–927.

Sequeira, S, N Nunn and N Qian (2020), “Immigrants and the Making of America”, Review of Economic Studies, 87(1): 382–419.

Tabellini, M (2020), “Gifts of the Immigrants, Woes of the Natives:  Lessons from the Age of Mass Migration”, Review of Economic Studies, 87(1): 454–486.


1 The Age of Mass Migration is typically defined as the period between 1850 and 1920. However, the largest migration flows occurred between 1900 and 1914, and the immigrant share of the US population peaked at 14% in 1920 (Abramitzky and Boustan 2017).

2 Most notably, the Immigration Act of 1921 introduced country-specific quotas that were based on the population from each country living in the US in 1910. With the National Origins Act of 1924, the ‘baseline year’ was moved to 1890, when very few immigrants from Eastern and Southern Europe were living in the US, with the goal of restricting immigration from these regions even further (Goldin 1994).

5,249 Reads