International Migration

The estimated income gains from international migration are large relative to most other interventions evaluated in the development economics literature that are designed to increase beneficiary income and employment opportunities. Moving from a developing to a developed country for work is associated with substantial income gains that are orders of magnitude larger than in situ efforts to raise domestic incomes in developing countries (Clemens 2011, Pritchett and Hani 2020), or the likely impacts of further liberalisation of international trade or capital flows.

Migration opportunities allocated via randomised lotteries – often implemented by governments to award scarce visas – provide compelling causal evidence of these gains. Gibson et al. (2018) study the impacts of New Zealand’s Pacific Access Category and Samoan Quota schemes, which randomly assign migration opportunities to Pacific Island residents. They find four- to fivefold income gains for lottery winners that persist up to a decade later. Mobarak et al. (2023) analyse a visa lottery for Bangladeshi workers to migrate to Malaysia, documenting a tripling of migrant incomes, as well as a broad range of improvements in well-being for both migrants and for their households remaining behind in Bangladesh.

International migration takes many forms around the world. Some citizens of developing countries migrate permanently to high-income nations either through employer sponsorship or on the strength of their skills. Others cross borders without formal documentation with the intention of staying temporarily. Sponsored migration offers paths to residency in some countries and, in such cases, workers often travel with their spouse and children, expecting to stay at the destination long-term. In many countries hosting large numbers of economic migrants, temporary work permits do not offer any legal path to citizenship, and workers often travel alone for a pre-specified contract period (Mobarak et al. 2023). Migrants will obviously share and remit differently if their nuclear family is left behind, and if they expect to return. There is therefore no one single estimate of returns to migration that would apply globally, and rigorous studies have produced very different estimates depending on the context, the conditions of migration, and on whom the effects are being measured.

The experimental evidence is reinforced by systematic comparisons of earnings for migrants versus non-migrants in nationally representative survey data. While these are not causal estimates of the effects of moving, they are helpful for calibrating the magnitude of potential gains in incomes. Clemens et al. (2019) estimate the “place premium”: the wage gain from working in the US rather than a developing country. The authors make wage comparisons for observably identical workers (e.g. born in the same non-US country with otherwise identical observable characteristics) in the US Census versus in one of 42 analogous origin-country surveys. For the median overseas country in their sample, moving to the US increases (PPP-adjusted) wages by a factor of 3.95, with a range from 1.7 (Morocco) to 16.4 (Yemen). For the median country, they estimate the lower-bound absolute annual income gain is US$13,600 (in 1999 PPP terms). Gains can be much higher for the most skilled. Gibson and McKenzie (2012) find the top academic achievers from Ghana and several Pacific Islands gain US$35,000–79,000 more per year from migrating, with the lifetime discounted gains from 30 years working abroad compared to at home ranging from US$532,000 to US$1.27 million. That is, migration is literally a million-dollar decision for many talented individuals. Clemens (2013) uses the randomised processing of US H-1B visas to examine why programmers earn far more in the US than in India. Given the quasi-random variation, wage differences are not primarily due to selection, but rather reflect the high value of location-specific factors such as institutions, infrastructure, and market conditions. These studies underscore how migration can dramatically increase workers’ earnings by enabling them to access more productive environments.

These income gains are not merely short-term, but represent persistent improvements in living standards. The magnitude of these effects — often several hundred percent increases in earnings — dwarfs the impacts of virtually any other development intervention. To put this in perspective, a randomised controlled trial of a development programme would be considered highly successful if it increased participant incomes by 10–30% (Banerjee and Duflo 2009, Banerjee et al. 2015, Bouguen et al. 2019). International labour mobility thus represents one of the most powerful anti-poverty tools available to the global community (Clemens 2011).

While the income gains are the easiest to quantify, moving from a developing to a developed country can benefit migrants in many other ways. For example, a survey of top academic achievers found migration to be in part due to broader career concerns such as the quality of opportunities to conduct research, work with other leaders in the profession, and learn from the best abroad (Gibson and McKenzie 2011). Opportunities for schooling for one’s children, health infrastructure, security, and lifecycle factors can all also be benefits that induce people to move, especially to destinations that welcome the entire family.

Future research should seek to broaden the evidence base on the economic benefits of international migration for migrants themselves in a wider range of settings. At the same time, it is important to conduct future research that quantifies the costs of international migration to migrants. In addition to the financial costs of migration, there are important non-pecuniary costs. Workers often face mistreatment and abuse in destination countries, particularly in lower-skilled occupations. How the connection between employees and employers is mediated also can matter. Unscrupulous middlemen sometimes tempt desperate migrants with fraudulent opportunities, and many migrants suffer large losses. Even in the absence of mistreatment, moving to a new place can be psychologically costly, involving considerable ex ante uncertainty, a loss of attachment to place, and separation from friends and some family members (Naidu et al. 2024, McKenzie 2024, Magaloni and Morten 2025).

There has been very little empirical research in economics that seeks to quantify the costs against the benefits of international migration. A rare exception is Naidu et al. (2024), who use an RCT to study the effects of international migration from India for construction work in the UAE. Workers’ incomes double, but they also bear substantial financial costs of migration, and experience physical and psychological costs from migrant work. Pecuniary and non-pecuniary costs together offset a substantial portion of the wage gains from migration. Future research should seek to expand this evidence base, systematically measuring the range of financial and non-pecuniary costs of international migration. Such costs must be gauged alongside wage gains and non-pecuniary benefits for a complete assessment of the effects on migrant well-being.

A second area for future research concerns the heterogeneity and uncertainty of the net gains from migration. Both the benefits and costs of moving can differ considerably across individuals and over time, and can be hard to anticipate. Batista and McKenzie (2023) find that risk and incomplete information are the key factors that reduce migration in incentivised lab experiments on decision choice. Additional work is needed to measure how risk and expectations affect migration decisions and the welfare gains of migration.

For full reference list see the end of the conclusion chapter.