Association for Public Policy Analysis & Management

We study the impact of three national unconditional cash transfer programs on household migration decisions. The programs are administered by the governments of Malawi, Zambia and Zimbabwe and represent the largest poverty-support programs in each country. The new economics of migration proposes that migration can be viewed as an investment decision or a risk diversification mechanism. Households may ‘invest’ in sending family members to seek employment elsewhere, with the expectation of remittances to offset the loss in labor, as well as serving as a way to diversify risk. Other forms of investment can include schooling or training, all of which will raise the future expected income of the household. In rural Africa, the setting for this comparative study, in the face of an adverse income shock, households may marry their daughters off at younger age to obtain the bride-price. Both these scenarios can be affected by an unconditional cash transfer (UCT). A UCT can finance out-migration for investment purposes, or reduce the need for out-migration and allow households to diversify their income portfolio locally. And in the face of a shock, a UCT can reduce the need for households to resort to bride-price to support their immediate consumption needs.

We use data from impact evaluations of three UCTS:

The Malawi Social Cash Transfer Programme: This three year longitudinal study with one baseline (2013) and two follow-up household surveys (2014, 2015), with a sample of 3500 households. Half of the sample were randomized out to a delayed-entry control status.

The Zambia Multiple Categorical Programme: A three year longitudinal study with one baseline (2011) and two follow-ups (2013, 2014), with a sample of 3200 households. Half the sample was randomized out to a delayed-entry control group.

The Zimbabwe Harmonized Social Cash Transfer Programme (HSCT): A four year longitudinal study with one baseline and two follow-ups (2014, 2017), with a sample of 3000 households. The comparison group consists of program eligible households from neighboring districts who were scheduled to enter the program in the next phase of expansion.

All surveys included detailed information on household membership, in- and out-migration, reasons for migration, date and destination/source, and remittances. We use this information to construct migration histories or portfolios, and then estimate the impact of the UCT on these portfolios to understand the full effect of the UCT on the economic diversification strategies of eligible households. A strength of the paper is that we can look across three different countries who have somewhat similar populations and levels of development, and roughly similar target populations (ultra-poor, labor constrained). We can exploit some heterogeneity of program design (a flat transfer in Zambia versus transfers linked to household size in Malawi and Zimbabwe) to understand how these affect migration portfolios and livelihood strategies. The results will deepen our understanding of the full effects of unconditional income support programs in sub-Saharan Africa on household well-being.