JEL Classification

The number of migrants worldwide has more than tripled since 1975, reaching 258 million in 2016 (United Nations 2017; Zlotnick 1998). Today, about 3.5 per cent of the world’s population are migrants. The migrant population has grown especially rapidly in Europe, North America and Australasia. In Europe, the population share of migrants increased from 2.7 per cent in 1975 to 10.5 per cent in 2017—a nearly fourfold increase over little more than 30 years. Over the same period, the share of migrants in the USA more than doubled to 15.3 per cent, reaching levels last seen at the beginning of the twentieth century. Other regions, such as Latin America, experienced much lower increases in migrant numbers or even a decline in the migrant population share (see Ferrie and Hatton 2015, for an overview of trends and developments in international migration in the last two centuries).

The increased importance of immigration in many developed countries has sparked a heated policy debate on the economic effects of immigration in immigrant-receiving countries.Footnote 1 These policy debates, along with the increased demographic importance of migrants, have also fuelled academic interest in the economic determinants and consequences of immigration. This chapter first summarises the core topics and debates in immigration economics, and then discusses how economic history can inform these debates.

Core Topics in the Economics of Immigration

Issues in the economics of immigration that have received particular attention include the decision to migrate; the integration of migrants into the economy of host countries; and the effect of immigration on native workers’ economic outcomes. Much of the economics literature on immigration—and the policy issues they revolve around—focus on the labour market, which is also the focus of this chapter.

The decision to migrate is a natural starting point for studying the economics of immigration. Economists and other social scientists have long highlighted that international migration is a selective process. Some residents choose to leave their country of birth, whereas others choose to stay. One key question then is who moves from origin to destination—and how the selection process depends on the characteristics of origin and destination countries. The key insight of the literature’s workhorse model , the Roy model, is that the selection of immigrants depends on the relative return to labour market skills in origin and destination countries (see Bansak et al. 2015; Borjas 2014, for textbook treatments; and Borjas 1987, 1991, for seminal contributions).

In particular, the Roy model predicts that migrants will have higher skills than residents who stay behind—and are thus positively selected in terms of their skills—if the destination country exhibits higher relative returns to skills and thus greater levels of income inequality than the origin country. In contrast, immigrants will be negatively selected if the destination country offers lower relative returns to skills and is therefore more equal than the origin. Numerous studies have tested the predictions of the Roy model, generally with mixed success (see, for instance, Belot and Hatton 2012; Brücker and Defoort 2009; Grogger and Hanson 2011). In fact, many studies find contemporary migration flows to be positively selected, regardless of the relative returns to skills in origin and destination.

The selection of immigrants has direct implications for their economic integration in host economies, which in turn is central for the economic costs and benefits of immigration for receiving countries. Starting with the seminal contribution of Chiswick (1978), a large literature in economics studies whether newly arrived immigrants fare worse than natives do—and, if so, whether they catch up over time (see, for instance, Card 2005; Lubotsky 2007; Borjas 2015 for recent contributions).Footnote 2 The answers to these questions depend on immigrants’ pre-existing skills but also on their investment into new skills after arrival in the host country. The skills of immigrants not only affect their own labour market success in host countries, but also that of their children (Dustmann and Glitz 2011). Since a significant share of migrant children come from a more disadvantaged family background than their native peers, the persistence of disadvantages across generations is one of the central impediments to the successful integration of second generation migrants (see Black and Devereux 2011, for a survey of the growing literature on the persistence of economic status across generations).

The skill level of immigrants relative to that of natives also determines the effect of immigration on wage and employment outcomes of native workers —at least in theory. In a simple textbook model of a competitive labour market, immigration will shift out the supply curve if immigrants and natives have the same skills and are perfect substitutes. Consequently, immigration decreases wages of competing native workers. In contrast, immigration will increase native wages if immigrants and natives have different set of skills and are complements in production. Consequently, immigration increases wages of complementary native workers.

The most common empirical approach to test these predictions is to correlate native wages and immigration across regions of a country (see Dustmann et al. 2016, for a discussion of alternative empirical approaches and their interpretation). Studies using this so-called spatial correlation approach compare native earnings in areas with a high immigrant share to native earnings in areas with a low share—and often find little or no evidence for an adverse wage effect of immigration (see Kerr and Kerr 2011, for a survey). However, immigrants might self-select into high-wage areas, severely complicating the identification of the true effect of immigration on wages in spatial correlation studies (see, for instance, Dustmann et al. 2008, for a discussion). A positive correlation between immigration and wages, for instance, might be evidence of a positive wage effect of immigration—but it might also indicate that migrants move into high-wage regions.

The Role of History for Current Debates in Immigration Economics

Migration is all but a new phenomenon. For thousands of years people have left their region of origin to improve their standards of living or to seek shelter from war and conflict . Can the analysis of historical migration episodes inform current debates in immigration economics—and, if so, how? At least three aspects come to mind: Economic history can highlight the importance of context for the applicability of economic models of immigration, provide a long-term perspective on immigration and serve as a laboratory for empirical studies in migration research.

First, historical episodes can provide the context for testing economic models of immigration. Abramitzky et al. (2012), for instance, study the selection of immigrants from Norway to the USA during the “Age of Mass Migration” (1850–1913). This is a particularly interesting episode to study migrant selection because the USA maintained a nearly open border for immigrants at the time. Migration decisions were thus unaffected by legal barriers, which govern much of contemporary migration flows. In line with the assumptions of a simple Roy model, the Age of Mass Migration thus allows researchers to test the selection of immigrants in the absence of selective migration policies. Abramitzky et al. (2012) show that Norwegian migrants to the USA were negatively selected from the sending population. This finding is consistent with the predictions of the Roy model, since income inequality was higher in Norway than in the USA during the period. Stolz and Baten (2012) come to similar conclusions. Using a comprehensive panel of 52 source countries and 5 destination countries, they show that the selectivity of migrants in the mid and late nineteenth century is in line with the predictions of the Roy model.Footnote 3 Today’s often-restrictive immigration policies towards the low skilled could thus explain why the Roy model is less successful in predicting the selection of contemporary immigrant flows.

Second, a historical perspective allows us to study the long-term effects of immigration, whereas much of the existing literature on immigration has focused on the short-term effects of immigration. Clearly, short- and long-term effects could differ significantly in both magnitude and sign. Recent studies that offer such a long-term perspective include Hornung (2014) and Sequeira et al. (2017). Both studies document significant positive effects of immigration on host countries in the long run. Hornung (2014) analyses the long-term effects of Huguenot refugees to Prussia on the productivity of textile manufactories. He shows that Prussian towns with higher refugee shares in 1700 experienced higher output and productivity levels in 1802. Sequeira et al. (2017) study the effect of immigration from Europe to the USA during the Age of Mass Migration on a wide range of contemporary economic outcomes. They find that counties with higher historical immigration are generally more prosperous today.

Bauer et al. (2013) take a long-term perspective on the economic integration of forced migrants, studying the economic integration of the eight million forced migrants from Central and East Europe who arrived in West Germany after World War II. They show that even a quarter of a century after displacement, first generation migrants performed significantly worse than native West Germans with similar socio-demographic characteristics. Differences in economic performance shrink, but are still visible, in the second generation.

Third, historical events can sometimes serve as a laboratory to study causal effects of immigration that are otherwise difficult to quantify. The literature on the labour market effects of immigration is a case in point. As discussed, immigrants typically self-select into booming labour markets, making it difficult to separate correlation from causation. Unique historical events can create regional variation in immigrant inflows that is independent of local economic conditions. Such “natural experiments of history” might clarify the direction of causality between immigrant inflows and native labour market outcomes (see Boustan et al. 2010; Braun and Mahmoud 2014, for recent examples).

The so-called Mariel Boatlift is one of the most famous examples of a natural experiment in immigration economics. On 20 April 1980, Fidel Castro announced that Cubans wanting to leave to the USA could do so from the port of Mariel. By the end of October 1980, 125,000 Cubans had reached Florida. The majority settled in Miami where the labour force increased—almost overnight—by seven per cent. Importantly , the inflow of Cubans to Miami was not driven by expected labour market opportunities, but by geographic proximity to the origin and the settlement of earlier Cuban migrants in Miami. Card (1990) uses the Mariel Boatlift as a natural experiment to estimate the effect of immigration on native wages and unemployment. He compares labour market conditions in Miami—relative to those of four similar cities that were unaffected by the inflow—before and after the Mariel Boatlift. He does not find any evidence for adverse labour market effects of immigration although his results are controversially discussed until today (see Borjas 2017; Peri and Yasenov 2015, for recent contributions).

Economics, History and Policy

Immigration economics addresses many of the questions that we encounter in today’s heated policy debates: Do immigrants displace natives in the labour market—or do they improve labour market outcomes for native workers? How quickly do immigrants integrate into the labour market—and which policies can foster the integration process? Who chooses to migrate—and how do labour market conditions in source and origin countries affect the selection process? A key learning outcome from introducing economic history into a course in immigration economics is that none of these questions are new, and all were already relevant in the past. Comparing research findings from different migration episodes can teach students the importance of context for the applicability of economic models and, more generally, the relevance (and irrelevance) of economic history for today’s immigration debates. Abramitzky and Boustan (2017) is an excellent starting point in this regard; they review side-by-side the historical and contemporary evidence on immigration into the USA and discuss how the direct comparison of different migration episodes in US history can shed light on current debates.