INTERNATIONAL MIGRATION, REMITTANCES, AND INEQUALITY

Globalization in its most general terms is the greater integration of global economic activ­ity. This is manifested in larger trade and in freer movement of factors of production.

In 2010, the total stock of international migrants in the world (developed and devel­oping countries) was 214 million people, up from 191 million in 2005.^[350] This compares to an estimated 749 million for internal migrants. International migration is a significant and growing phenomenon. This is especially true of migration from developing to developed countries. The stock of immigrants in high-income countries increased at about 3% per year from 1980 to 2000. As a share of high-income country population, migrants increased from around 4% to above 8% over this 20-year period. ^[351]

24

How might the much higher rate of international migration affect the distribution of income in developing countries in theory? The answer depends on who migrates and what they do with their income after they migrate in terms of remittances to their fam­ily. If migration and remittance was representative of the domestic income distribution, then the distribution would not be affected, except for a translation to the right as remittances flowed back. Thus, poverty would decline as a result of international migration.

What if migration was not representative but selective on individual characteristics? Would the poverty results still hold? The impact effect of migration as the result of better income earning opportunities must surely be to reduce poverty at the origin. However, in the next round there is the possibility of externalities kicking in if the migrants are the most highly skilled, with knock on effects on the rest of the economy. This is the famous “brain drain” hypothesis that was popular in the 1970s and 1980s.^[352] In recent years, this has been countered by the “brain gain” hypothesis, which is based on the simple idea that the probability of having access to international migration depends on the education level of the prospective migrant. In order to improve this probability, prospective migrants invest in education. Only some of these will be selected for migration, but those left behind will serve to increase the stock of human capital compared to what it would have been without the prospect of migration.^[353]

There is some empirical support for the brain gain hypothesis, although others argue that its magnitude is greatly exaggerated.^[354] Furthermore, there is considerable evidence for the proposition that international migration reduces poverty in the origin country. In perhaps the most comprehensive such exercise, Adams and Page (2005) asked the ques­tion on the impact ofinternational migration on poverty using data from 71 developing countries:

The results show that both international migration and remittances significantly reduce the level, depth, and severity of poverty in the developing world. After instrumenting for the possible endo­geneity ofinternational migration, and controlling for various factors, results suggest that, on aver­age, a 10% increase in the share of international migrants in a country's population will lead to a 2.1% decline in the share of people living on less than $1.00 per person per day.

Adams and Page (2005, p. 1645)

These results are confirmed by a range of country specific studies on international migra­tion, remittances, and poverty—examples include Acosta et al. (2006) for Latin America, Lokshin et al. (2007) for Nepal, and Adams(2006) for Ghana.

So much for poverty, where theory and evidence is relatively clear cut. What about inequality? It should be clear that selectivity of migration and remittances makes this an intricate question theoretically and empirically. And the question ofidentifying such selec­tivity is an important one in the international migration literature. In particular, there is some debate about whether migrants are selected according to education level. Using data from Docquier and Abdeslam (2006), Hanson (2010) compares the share of emigrants with tertiary education to the share of total population with tertiary education. He finds that in the vast majority of the countries, the former exceeds the latter, indicating positive selec­tion into migration by higher levels of education. Mexico and Puerto Rico appear to be exceptions to this almost universal phenomenon, but research on migration from those origins to the United States seems to have had significant weight in the discourse. Hanson (2010) argues that a larger literature now seems to support selection on education.

What about migration selection based on unobserved variables? McKenzie et al. (2006) conduct an ingenious exercise using the results of a lottery for emigration from Tonga to New Zealand. They compare losers in the lottery with nonapplicants, both groups of course still being in Tonga. They find that the applicants have higher earnings after controlling for observables; and they conclude therefore, that those desiring to migrate are selected in terms of higher income earning potential.

If international migrants are selected from households that already have high earnings, and their migration raises income earning and, through remittances, adds to the income of the household in the origin area, it should be clear that such migration would tend to increase inequality in the sending country. However, to the extent that the selection goes the other way, inequality in the sending country will be mitigated by international migration. There is now a considerable literature on assessing directly the impact of inter­national migration on inequality, and we now turn to an overview of those studies.

The empirical results on international migration and inequality are inconclusive as a whole. Barham and Boucher (1998) compare the actual distribution post-migration including remittances for Nicaragua, with a Counterfactual of what the distribution would have been if the migrants had not left and earned their original income. They found that the Gini coefficient is higher by 12%. Adams (2006) finds a much smaller increase in the Gini coefficient for Ghana—of 3%. The difference made by the counter- factual approach is illustrated by comparing the findings of De and Ratha (2005) and Karunaratne (2008) for Sri Lanka. Using the 2003—2004 Socioecononmic Survey for Sri Lanka, Karunaratne (2008) shows that “income receivers belonging to lowest 10 per­cent receive 1.3 percent of their income as remittances with the top 10 percent of the income receivers getting 4.6 percent of their income from remittances” (p. 58). He uses this to argue that remittances increase inequality. However, De and Ratha (2005) con­duct counterfactual analysis and show that remittance income exceeds the counterfactual loss in income from migrating in the bottom two deciles, while the opposite is true for the top two deciles. Thus, they argue, remittances are equalizing.

A major issue in the empirical literature is the difference between short-term and long-term effects of international migration on inequality.

Overall, then, the final effect of globalization on inequality in developing countries through the channel of international migration is ambiguous in theory, and this is reflected in the conflicting empirical findings. Of course, the migrants are themselves better off—it is the consequences for those they leave behind that are uncertain. These results pick up on a theme of this chapter as a whole, namely that the consequences for distribution depend on the context and, in particular, on preexisting structural inequal­ities. When these inequalities are high and interact with the opportunities presented by globalization in such a way as to benefit those who already well off, inequality will increase. The next section turns to the policy implications of these findings.^[357]

20.8.