The Encouraging and Distortionary Effects of Conditional Cash Transfers

• Development Workshop

Date and Location

Tuesday, April 9, 2019, 12:10 PM - 1:30 PM
ARE Library Conference Room, 4101 Social Sciences and Humanities

Abstract

This paper assesses the effects of conditionality in cash transfer programs using a randomized experiment in Eastern Indonesia. Subsistence farmer households in West-Timor were randomized into being offered either a cash transfer conditional on seasonal migration coupled with information about work opportunities (CCT); or an unconditional cash transfer (UCT). The CCT treatment was further split into a low (‘CCT low’; IDR 150,000, USD 10.50) or a twice as high (‘CCT high) subsidy group. The CCT raised migration rates and household income during the migration season compared to the unconditional cash transfer, but strikingly, impact is lower for those assigned to the ‘CCT high’. The finding of lower treatment effects for the ‘CCT high’ group compared to the ‘CCT low’ group are consistent with two opposing impact channels of tying conditions to cash transfers. First, there is the effect of encouraging targeted behaviors that require inducement and/or framing for individuals to experiment with: cash alone seems not sufficient to induce individuals to exploit the profitable seasonal migration investment opportunity. Second, conditionality increase distortion, by changing the composition of types (in this case, lower returns to migration) selecting into the targeted behavior. If the transfer is too large (as in our ‘CCT high’ treatment), these distortionary effects outweigh the encouragement effect, implying that there is an optimal transfer size in welfare terms. Besides these general implications, the results also point to liquidity constraints to seasonal migration in Eastern Indonesia, and the dominance there of this type of CCT over UCT in benefit-cost terms.