Intergenerational mobility of immigrants in 15 destination countries

As immigration continues to shape the demographic makeup of developed countries, understanding the factors that promote successful integration across generations becomes increasingly important. This column uses data from 15 immigrant-receiving countries to reveal large income gaps for first-generation immigrants that diminish in the second generation in many destinations. Lower rates of mobility for sons than daughters of immigrants suggest a need for gender-specific integration strategies, while the fact that parental country of origin does not drive cross-country differences in mobility suggests that immigration selection policies focused on country of origin may be less important for integration than pathways to success for second-generation immigrants.

The integration of immigrants into the labour markets and societies of developed countries is a pressing policy challenge. In 2019, 10-30% of the populations of most OECD countries were born abroad, and a similar range of children aged 0-14 had at least one foreign-born parent (OECD/EU 2023).

Social scientists and policymakers have long debated whether immigrant families experience economic mobility across generations, how this mobility compares to that of local-born families, and what factors influence these outcomes (Borjas 2006). Previous research on immigrant economic integration has largely been limited to single-country studies (e.g. Abramitzky et al. 2021, Jensen and Manning forthcoming, Razin 2018), making it difficult to identify which patterns are universal and which are specific to particular destinations. Algan et al. (2010) and Bucca and Drouhot (2024) provide comparative evidence across a limited set of Western European countries, but neither study includes destination countries outside Europe, nor do they contain direct parent-child links.

In a new study (Boustan et al. 2025), we compile and harmonise administrative and survey data from 15 immigrant-receiving countries, providing the first comprehensive comparative perspective on the economic integration of immigrants and their children across generations. Our data include 11 European and four non-European countries, representing 44% of global immigrants and 68% of immigrants in high-income countries. Following the methodological approach of Chetty et al. (2020) and Abramitzky et al. (2021), we focus on children born in destination countries between 1978 and 1984 and examine their labour market outcomes approximately 30 years later. Our findings reveal significant cross-country differences in immigrants’ intergenerational mobility, with important implications for immigration policy and labour market integration.

Differences in income and rates of mobility

Our harmonised data from 15 destination countries allow us to establish two fundamental facts about the economic integration of immigrants and their children.

• First, Figure 1 shows large income gaps for first-generation immigrants that diminish in the second generation in many destinations. The median income rank gap across destination countries in our data is -5 rank points. The median gap between second-generation immigrants (children of immigrants) and the children of the local born is much smaller, at less than 1 rank point.
• Second, we find evidence of gender differences in income gaps. Daughters of immigrants experience smaller income gaps than the sons of immigrants in all destination countries. The median rank gap is -3 points for sons and zero points for daughters.

Figure 1 Income rank gaps between immigrants and local-born, first generation (parents) and second generation (children)

Notes: This figure reports the mean difference in income ranks between immigrants and local-born, as well as between their children. Children are born in 1978-1983. Immigration status is determined by father’s country of birth. Child income is measured in 2014-2015, and parental income in 1994-2000. Income ranks, 0-100, are determined within child birth cohorts.

To further understand the dynamics behind these findings, we decompose the cross-country income gaps between the children of immigrants and children of local-born into three components: (1) a part due to differences in parental income; (2) a part due to difference in absolute mobility (that is, gaps in income for children raised at the bottom of the income distribution); and (3) a part due to relative mobility (the correlation between the income of parents and children). Our decomposition exercise reveals the following:

1. On average, around half of the cross-country variation in second-generation income gaps can be explained by parental income differences (Figure 2). Children of immigrants tend to be raised in poorer households than the children of local-born parents. Thus, countries with a smaller first-generation income rank gap, such as the US and Canada, also have smaller second-generation income rank gaps.
2. After accounting for parental income, remaining income gaps for the children of immigrants are driven by differential rates of absolute mobility (these differences are shown in Figure 3). Differences in relative mobility play a much smaller role in explaining income gaps between children of immigrants and locals.
3. In most countries, daughters of immigrants exhibit higher absolute mobility than daughters of local-born parents (Figure 3). Sons of immigrants only enjoy this advantage in non-European countries with long histories of immigrant incorporation (Australia, Canada, Israel, and the US), as well as in the UK. As a result, daughters of immigrants have higher income than daughters of local-born parents raised at the same point in the income distribution in most destination countries, while sons of immigrants often have lower income

Figure 2 Oaxaca-Blinder decompositions of differences in child income ranks

Notes: This figure plots results from a Oaxaca-Blinder decomposition of the difference in mean income rank between children of immigrants and children of local born, using children of local born as the reference group. Specifically, the dark grey bars plot the difference in mean income ranks between the children of immigrants and children of local born. The light grey bars plot the gap in income that cannot be explained by parental income. Children are born in 1978-1983. Immigration status is determined by father’s country of birth. Child income is measured in 2014-2015, and parental income in 1994-2000. Income ranks, 0-100, are determined within child birth cohorts. 95% confidence intervals indicated.

Figure 3 Differences in absolute mobility between children of immigrants and children of locals (higher for immigrants = above zero)

Notes: This figure plots estimates differences in absolute mobility between children of immigrants and children of locals for each destination country. Children are born in 1978-1983. Immigration status is determined by father’s country of birth. Child income is measured in 2014-2015, and parental income in 1994-2000. Income ranks, 0-100, are determined within cohorts.  95% confidence intervals indicated.

Potential drivers of cross-country differences

After establishing that cross-country differences in the mobility of children of immigrants across destinations is primarily driven by gaps in absolute mobility with the children of locals, we explore potential drivers of these gaps. We emphasise that this exploration is based on cross-country comparisons and, as such, we can only provide suggestive rather than causal evidence for these mechanisms.

We first explore differences between immigrant and local-born parents, beyond measured income. Differences in parental attributes – including parental country of origin – cannot explain cross-country variation in the absolute mobility gap. First, for most countries, other parental characteristics (i.e. parental wealth, geographic location, and industry of employment) cannot account for the remaining gap between the children of immigrants and the local born. Second, differences in the composition of parental sending countries do not help explain variation in absolute mobility across destinations. For example, China is a large sending country in Canada and Turkey is a large sending country in Austria. However, controlling for parental sending country does not affect our estimates of destination country differences in absolute mobility.

Given that parental attributes cannot account for cross-country differences in absolute mobility, we turn as an alternative to associations with destination country attributes. First, in Figure 4, we document that the mobility gaps for sons are higher in countries with lower income inequality. Sons of immigrants may be excluded or choose not to participate in equality-enhancing institutions such as vocational training, apprenticeships, or union membership. Indeed, the mobility gap in income for sons is strongly correlated with a mobility gap in employment rates (extensive margin), which can be depressed by weak school-to-work transitions. Daughters of immigrants are less sensitive to destination-country inequality. Second, we find that both sons and daughters of immigrants enjoy higher mobility in countries with access to citizenship for the second generation and positive attitudes toward immigrants.

Figure 4 Association between mobility gaps inequality in destination countries

Notes: This figure plots estimates of absolute mobility difference between children of immigrants and children of locals for each destination country on the y-axis and the country-level 2014 Gini coefficient on the x-axis (from OECD data explorer: https://data-explorer.oecd.org). Children are born in 1978-1983. Immigration status is determined by father’s country of birth. Child income is measured in 2014-2015, and parental income in 1994-2000. Income ranks, 0-100, are determined within cohorts. 95% confidence intervals indicated.

Policy implications

Our findings have important implications for immigration and integration policies across developed economies. First, the substantial narrowing of income gaps between the first and second generations in most countries suggests that economic integration does occur over time, though the speed and extent vary across contexts. This highlights the importance of taking a long-term, multigenerational view when considering the success of immigration policies.

Second, the different rates of mobility for sons and daughters of immigrants suggest a need for gender-specific integration strategies. Programmes could target young men of immigrant background in continental European countries, where they face the largest disadvantages. These could include enhanced school-to-work transitions and improved access to vocational training and apprenticeships, although more research is needed to establish their effectiveness.

Third, policies that promote citizenship access for the second generation may be particularly effective in fostering economic integration. Countries with easier pathways to citizenship show consistently stronger economic outcomes for children of immigrants – this is in line with causal evidence (Govind 2021, Vonnahme and Gathmann 2022).

Finally, the fact that parental country of origin does not drive cross-country differences in mobility suggests that immigration selection policies focused on country of origin may be less important for long-term integration than policies that provide opportunities and pathways to success for second-generation immigrants once they are in the destination country, such as language training and placement policies (Hasager et al. 2022).

As immigration continues to shape the demographic makeup of developed countries, understanding the factors that promote successful economic integration across generations becomes increasingly important. Our findings imply that with appropriate policies and institutional arrangements, the children of immigrants can achieve substantial economic convergence with the children of natives – though current approaches tend to be more successful for daughters than for sons, and more effective in some national contexts than others.

Source: cepr.org