Evaluating Alternative Cash Transfer Designs in Kenya Using Behavioral Economics
How to cite
Niehaus, Paul; Mani, Anandi; Mullainathan, Sendhil; Shah, Anuj. 2014. "Evaluating Alternative Cash Transfer Designs in Kenya Using Behavioral Economics." AEA RCT Registry. AEARCTR-0000541. https://doi.org/10.1257/rct.541-2.5
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study aims to draw on insights from behavioral economics to conduct a rigorous impact evaluation of alternative unconditional cash transfer designs in Kenya. The study's implementing partner, GiveDirectly, which provides unconditional cash transfers to poor households in rural Kenya, will be experimentally varying several features of these transfers. One area of focus will be the timing of transfers, where GiveDirectly will vary the structure and schedule of transfer payments, as well as recipients' control over these timing aspects. A second area of focus will be the role of social information and norms, varying the amount and kind of information recipients receive about assets purchased by their peers. This study will exploit the variation across these different designs to quantify impacts on 1) household-level socio-economic outcomes (income, assets, etc.), 2) measures of well-being (food security, intra-household discord/conflict, various psychosocial scales), and 3) recipients' decision-making process and cognitive ability.
Intervention
All households in this study will receive an unconditional cash transfer totaling around $1000 USD from GiveDirectly. The structure and timing of these transfers, as well as the information provided to recipients, will be randomized at the household level.
Experimental design
The experimental design comprises two arms looking at 1) transfer structure and 2) information sharing.
On transfer structure, cash transfer recipients will be asked to choose the timing and tranching of payments that they would most prefer to receive their transfer. One randomly selected group of recipients will receive the transfer according to their preferred payment structure and schedule, while the other group will be assigned a random payment structure and schedule.
On information sharing, the study will take a list of uncommon investments made by past recipients and randomly vary, at the recipient-investment level, (a) whether the recipient is told about this investment, and (b) if so, the type of information each recipient receives.
Design details
Cited by
No citation data available for this registration yet.