Cash transfers, migration, and gender norms

Overall, the findings suggest that the migration patterns of women and men were likely to be affected quite differently by a cash transfer program. Men predominantly migrated for employment, whereas women migrated to rural areas for marriage and urban areas for employment.

• The probability of rural–rural migration among men in beneficiary households (households receiving the cash transfer) increased by 0.9 percentage points (an effect size of 100%) compared to households not receiving the cash transfer. Rural-rural migration is any form of migration to, from, or between rural areas.
• The probability of rural–urban migration among women decreased by 0.2 percentage points (an effect size of 50%) with the cash transfer.
• The cash transfers revealed no impacts on women’s rural–rural migration or men’s rural–urban migration, on average.
• Further analysis by baseline wealth revealed that rural-rural migration for men was concentrated in the bottom four of the five wealth quintiles, and rural-rural migration for women was concentrated in the top wealth quintile.
  • Women in less poor beneficiary households were more likely to engage in rural–rural migration as a result of the cash transfer, whereas women in poorer beneficiary households experienced no immediate migratory impacts.

This study suggests that cash transfers can potentially foster investment in profitable endeavors outside of subsistence agriculture for men, but may also affect the marital migration of women.

To evaluate the impact of Mali’s national cash transfer program, the authors relied on a cluster of randomized controlled trials in five southern regions of the country. Across 96 communes, 71 were randomly selected to start receiving the cash transfer in 2014-2015 (early treatment). 20 additional communes were assigned to start receiving the cash transfer in 2016-2017 (late treatment). Moreover, a subset of 19 early treatment communes deemed more nutritionally vulnerable were assigned to either receive or not receive preventative nutrition packages (provided by the World Food Program and consisting of flour for young children and pregnant women) for one year. The randomization of the beneficiary households allowed the authors to provide cause-and-effect estimates of the cash transfer receipts on migration through comparisons of outcomes across the early treatment (henceforth treatment) and late treatment (henceforth control) groups.

From the original sample, the authors analyzed 2,209 households across 65 early treatment (treatment) communes and 19 late treatment (control) communes from 2014 to 2016. This excluded some communes where security concerns prevented data collection, and restricted the sample to households with at least one child between the ages of 23 months and 6 years at baseline, as well as with members between the ages of 15 and 48 at endline who had per capita food consumption data available at baseline.