Published on Development Impact

The Impact of Low-Skilled Migrants on High-Skilled Women’s Work

This page in:

Much of the debate about the effects of immigration on native workers focuses on possible negative consequences for wages or employment. However, a series of recent papers highlights a big positive effect – having immigrants as cleaners, nannies, and home-care assistants allows high-skilled women to work more.

Evidence for the U.S. comes from Patricia Cortés and José Tessada in the latest issue of the American Economic Journal: Applied Economics. They estimate the impact of variation in changes in the concentration of low-skilled immigrants in different U.S. cities in 1980, 1990 and 2000 on female labor supply, hours worked, houses spent doing household chores, and expenditure on household care. Identification comes from instrumenting new immigration waves with the 1970 immigrant shares in different cities, based on the idea that migrant networks are self-reinforcing. They find the low-skilled immigration wave of the 1980s and 1990s increased by 20 minutes a week the time that women at the top quartile of the wage distribution spent working in the market, with this effect only at the intensive margin (how much skilled women work, not whether they work) and mostly coming through increases in the probability of high-skilled women working more than 50 and 60 hours per week - and this same group of high-skilled women also reduce time devoted to housework and increase their expenditure on household services.

Despite having written papers that use historic (80 year ago) migration networks as instruments for current migration in Mexico, and the prevalence of this approach in both the immigration and emigration literatures, I am not a big fan of the use of relatively recent migration networks as an instrument – indeed I receive many papers each to referee that use say migrant networks from 5 or 10 years ago as an instrument for current migration, and am fairly critical of such work. The problem is that, unlike in the Mexican case, there is usually not a story of why migrant networks should be exogenously different across locations. Cortes and Tessada recognize this issue, and do include a good number of additional controls including the use of city fixed effects, region*decade time trends, and interactions of 1970 labor force participation rates of college educated women and other key 1970 variables with time trends to at least ease some of these concerns. However, a further issue which the paper doesn’t seem to address is whether high-skilled women migrate to work where there are low-skilled immigrants – instead only discussing the possibility of migration responses by low-skilled natives.

Guglielmo Barone and Sauro Mocetti in a recent working paper look at the same issue for Italian women, who on average have low labor force participation rates by European standards. Their instrument is similar in spirit, but with a slight twist – they exploit the tendency of newly arriving female immigrants from countries whose workers tend to work in domestic services to settle in places where males of the same country already live. Like the U.S. case, they find no effect on the extensive margin, but increases in hours worked for high-skilled women. The magnitudes are larger than in the U.S. case – the almost 6 percentage point increase in immigration into Italy over the 1990s is estimated to have increased hours worked by skilled natives by about 3 hours per week. Interesting heterogeneity is that the results are stronger for women with children aged under 3, and for women in municipalities with weaker social and family policies – i.e. immigration is serving as a substitute to the childcare services sometimes offered by the state.

A new working paper by Tiago Freire argues that the same phenomenon occurs with rural-urban migration in Brazil. Using rainfall shocks in rural areas to explain why people leave rural areas, and distances to different cities to predict where the move to when they receive a shock, he finds migration of low skilled women to cities lowers the wages of maids in those cities, increasing the labor supply of high-skilled women.

Facilitating more migration of low-skilled women, both internally and internationally, therefore offers the possibility for a wide range of benefits. Indeed Michael Kremer and Stanley Watts calculate that the gains from a program like Singapore or Hong Kong’s could increase native welfare by 1.2 percent (which far exceeds the magnitudes calculated in traditional estimates of the benefits of migration which ignore impacts on high-skilled women’s work) – while at the same time potentially lowering wage inequality (as an increased supply of high skilled workers lowers their wages) and reducing gender disparities among high-skilled natives. The migrant workers gain from the additional income, although more research is still needed on what the long-term costs of separation from their families are.


David McKenzie

Lead Economist, Development Research Group, World Bank

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000