International Migration

The substantial income gains from international migration generate significant benefits not only for migrants themselves, but also for the family members they leave behind. These benefits flow through multiple channels, with remittances representing the most direct and well-documented mechanism. However, the benefits are not only due to remittances as a mechanism. For example, migrants also provide informal insurance for households back home via their savings held overseas as well as their substantial (unremitted) income streams, and this improved ability to cope with risk can encourage households to make investments, such as in small enterprises and education. Migrants can also benefit their families in the origin country through knowledge sharing, and through providing a network that can facilitate future migration of other family members.

Yang (2008b) provides evidence of these effects by exploiting a natural experiment arising from the 1997 Asian Financial Crisis. Migrants from the Philippines working in dozens of countries experienced sudden, heterogeneous changes in exchange rates that generated exogenous variation in the Philippine peso value of their foreign earnings. When migrants experienced more positive exchange rate shocks, there were substantial positive impacts on their origin households over the subsequent 15 months. A 25% improvement in the exchange rate facing migrants led to a 15% increase in remittances. Overall, this positive shock led mainly to higher household investments rather than higher consumption. Educational spending rose by 13% and school attendance increased for both boys and girls. Households increased their entrepreneurial activity, with hours worked in self-employment rising by 12%. Notably, overall labour supply remained stable, suggesting that remittances enabled productive labour reallocation rather than simply reducing work effort. In related work examining the impacts of these same exchange rate shocks, Yang and Martinez (2006) demonstrate that these positive effects on household welfare translate into measurable poverty reduction.

Evidence from randomised migration lotteries provides further compelling evidence of benefits to migrant families. Gibson et al. (2018) analyse long-term impacts from New Zealand’s Pacific Access Category and Samoan Quota schemes, which randomly allocate migration opportunities to Pacific Island residents. They find that households of lottery winners experience substantial and persistent improvements in living standards. Consumption per adult equivalently increases significantly, and there are notable improvements in housing quality, durable goods ownership, and access to basic services. The effects persist for years after the initial migration, suggesting that the benefits represent genuine welfare improvements rather than temporary adjustments.

Mobarak et al. (2023) provide evidence from the Bangladesh-to-Malaysia visa lottery, which randomly allocated temporary work permits to potential migrants. Winning the lottery led to dramatic improvements in household welfare that extended far beyond the direct income gains. Households of lottery winners experienced large increases in consumption and asset accumulation, improved housing conditions, and enhanced food security. Women gained decision-making power when males emigrated. But it also postponed family formation and childbearing, since male migrants were temporarily separated from their partners and families. And in the short-run migration reduced entrepreneurial activity in those Bangladeshi households, as the would-be entrepreneur left to work in Malaysia. The consumption gains were so large that they moved many households above the poverty line.

Additional evidence comes from studies exploiting policy discontinuities for exogenous variation. Clemens and Tiongson (2017) examine a natural experiment in the Philippines where a language test cutoff score quasi-randomly allocated workers to temporary overseas employment in Korea. They find that migration has important reduced-form effects on household financial behaviour, tripling expenditure on education and health, reducing borrowing, and raising savings. Migration causes families to send children to private schools and private clinics, but does not affect non-migrants’ labour supply.[1]1. A number of other studies have also found that migration of a household member leads to increased educational investments in origin households (Gibson et al. 2011, 2013, Theoharides 2018).

Beyond education, migration can facilitate entrepreneurial investments in origin areas. Yang (2006) demonstrates how Filipino migrant workers’ decisions to return from overseas and invest in household enterprises are closely linked. Drawing on the 1997 Asian Financial Crisis exchange rate shocks, the study reveals that a substantial fraction of migrants appear to be ‘target earners’ who accumulate savings overseas until reaching a threshold amount, then simultaneously return home and invest in small enterprises. This pattern suggests that temporary migration can serve as a mechanism for overcoming capital constraints that would otherwise hinder entrepreneurial activity in origin areas. Relatedly, Kırdar (2009) studies return migration from Germany, finding that migrants return to origin countries once they have achieved target savings levels. Bossavie et al. (2025) find that international labour migration promotes capital accumulation for self-employment activities in Bangladesh. Their data shows that 76% of self-employed return migrants used savings accumulated abroad as their primary source of business financing, with return migrants accounting for one-third of all new non-agricultural enterprises in Bangladesh.

The benefits extend to health investments as well. Bossavie et al. (2025) document improvements in household dietary diversity and healthcare utilisation among lottery winners’ families, suggesting that the benefits of migration encompass multiple dimensions of human welfare.

Whether these benefits materialise or not will depend in part on the local economies from which individuals are leaving. Expecting remittances to be invested in small businesses in areas with small populations and where people are migrating for better opportunities is likely unrealistic. Gibson and McKenzie (2014a) give the example of seasonal workers from Vanuatu and Tonga not being invested in new business formation due to the small size of these economies. Likewise, remittances may not be invested in education in cases where the returns to additional education are low, such as when migration opportunities are for less-educated work (McKenzie and Rapoport 2011). Information flows also matter for the flows of remittances. Batista and Narciso (2018) randomly provide free international calling credit to migrants and find that treatment leads to increases in remittances, while Ambler (2015) shows that information asymmetries can reduce the amount of money that migrants remit.

When households send a member away as an international migrant, they also enjoy benefits of improved informal insurance – a better ability to cope with risk. Yang and Choi (2007) demonstrate that migrant family members provide substantial protection against income volatility in origin areas. Among households in the Philippines with overseas migrant members, rainfall-shock-driven income changes led to offsetting changes in remittances: remittances fell when local income rose and increased when local income fell. Roughly 60% of exogenous declines in income were replaced by remittance inflows from overseas. This insurance function extends to larger-scale disasters as well. Yang (2008a) shows that hurricane damage in the poorest developing countries leads to large inflows of migrants’ remittances, amounting to 20% of experienced damages — a response roughly one-quarter as large as the response of foreign aid.

Beyond the insurance provided by existing migrants, migration itself serves as an important risk-coping mechanism: households can respond to negative shocks by sending additional family members abroad. Mahajan and Yang (2020) and Ibáñez et al. (2025b) show that weather shocks in migration-source countries (respectively, hurricanes worldwide and high temperatures in El Salvador) increase migration to the US, and the migration responses are facilitated by existing migrant networks. Murathanoglu (2023) shows that typhoons increase international labour migration from affected Philippine municipalities by 3.7–4.3% of the mean migration rate, with remittances rising by 6.7% in the short run following typhoon exposure. Migrants accept jobs in lower wage destination countries and occupations. The effectiveness of migration as a shock-coping mechanism depends critically on international labour demand conditions, with both migration and remittance responses being substantially larger during periods of high overseas labour demand.[2]2. Studies of internal (not international) migration provide complementary insights. Meghir et al. (2022) show that providing subsidies for internal migration to rural migrants in Bangladesh leads to a 40% increase in risk-sharing in their origin villages. Munshi and Rosenzweig (2016) demonstrate that well-functioning informal insurance networks actually reduce internal migration in India by providing consumption smoothing, reducing the informal-insurance motivation for migration. This suggests that migration and local risk-sharing mechanisms are to some degree substitutes. But this relationship may vary depending on other factors. Internal migration can increase when financial technology allowing for instant remittances (like mobile money) is first introduced. This technological advance increases the returns of migration as an insurance strategy, prompting further migration, remittances and consumption-smoothing in shock-hit villages, as shown by Batista and Vicente (2025) for internal migration in Mozambique. Other studies emphasising the risk-coping role of internal migration include Minale (2018) in China, Kleemans and Magruder (2019) and Kleemans (2023) in Indonesia, and Letta et al. (2024) in Nigeria.

These findings demonstrate that the benefits of international migration extend well beyond the migrants themselves to encompass their families left behind in the home country. Well-being improves for those who remain in origin countries on several dimensions: higher consumption; enhanced investment in education, health and small enterprises; and better insurance against shocks.

The logic that families can benefit when a migrant member gains access to a better labour market should extend to moves that do not cross international boundaries. Indeed, an experimental literature that facilitates ‘internal’ rural-urban migration reports an analogous set of positive effects on migrant families. The magnitude of gains from internal migration are not as large as gains from international migration, which makes sense because rural-urban wage differences within the same country are quite a bit smaller than international wage differences across countries. Bryan et al. (2014) finds a 30% improvement in household food and non-food expenditures during the ‘hungry’ season when a migration subsidy intervention subsidises a member’s travel to nearby cities during the lean, low-labour-demand period in the agricultural cycle in rural Bangladesh. This insight also replicates in rural Indonesia in a separate experiment (Bryan et al. 2023). A similar subsequent experiment in Bangladesh conducted at a larger scale shows that rural wages increase when many workers are induced to emigrate out (Akram et al. 2025). Informal insurance and village risk-sharing improve when some workers emigrate (Meghir et al. 2022), but there is also ‘unobserved’ disutility from family separation that is quantitatively significant (Lagakos et al. 2023). Potential migrants also often underestimate the returns to rural-urban migration, partially due to income hiding by migrants (Baseler 2023).

In between short-run, circular internal migration and longer-term international migration is the ‘hybrid’ phenomenon of short-run circular migration crossing international borders to neighbouring countries for contract work, such as Mexican and Haitian seasonal farm workers in the US and Canada (Clemens et al. 2018), or Nepali seasonal agricultural workers in India (Mobarak et al. forthcoming). An intervention improving those Nepali migrant workers’ ability to share their remittance income with their families back home in rural Nepal significantly improves the family’s consumption, food security, and subjective measures of their well-being (Mobarak et al. forthcoming).

While most studies on the impacts of international migration on origin households find positive impacts, an exception is Gibson et al. (2011), who study Tongan families whose members migrated to New Zealand via lotteries. They find that households experience significant resource constraints in the initial post-migration period. Total household income per capita fell, as increased remittances were insufficient to offset the loss of migrants’ domestic labour earnings. Remaining family members had fewer durable assets and shifted towards cheaper foods like rice and root crops while consuming fewer fruits and vegetables. The only notable positive effects were reduced body mass index and waist-to-hip ratios for adults (resulting from decreased food intake in a population with high obesity rates). There were minimal impacts on children’s education and health. These results highlight that the initial period of adjustment to a member’s migration can impose welfare costs on those left behind.

While the literature in general has found positive impacts of migration on families, there are also costs of having household members migrate that are sometimes harder to quantify. Impacts of having an absent parent, spouse, or child, the absence of prime-age adults who would otherwise be contributing to household public goods, and the increased costs of communication can all offset some of the gains. Measuring these costs more clearly, and testing interventions to reduce these costs are useful directions for future research.

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