The General Equilibrium Effects of Large Universal Cash Transfers in Malawi

How to cite

Egger, Dennis; Walker, John; Mughogho, Winnie; Henning, David; Karlan, Dean. 2025. "The General Equilibrium Effects of Large Universal Cash Transfers in Malawi." AEA RCT Registry. AEARCTR-0015292. https://doi.org/10.1257/rct.15292-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Dennis Egger* , John Walker , Winnie Mughogho , David Henning , Dean Karlan

* Primary investigator

Registry ID
AEARCTR-0015292
DOI
https://doi.org/10.1257/rct.15292-2.0
First registered
Last updated
Status
On going
Trial dates
2025-02-01 – 2035-12-31
Countries
Malawi (Chiradzulu District)
Registry record
http://www.socialscienceregistry.org/trials/15292

Abstract

This is the AEA Registry for the General Equilibrium Effects of Universal Cash Transfer program in Malawi. It is a three-stage cluster-randomized controlled trial of the NGO GiveDirectly's unconditional cash transfer program at scale. The transfers are universal (every adult in a village is eligible), unconditional, one-time, and lump-sum, amounting to approximately 550 USD or 70-100% of baseline annual consumption expenditure of recipients. Randomization is at the village, Group-Village-Headman (GVH), and Traditional Authority (TA) level, creating exogenous variation in transfer receipt, and geographic variation in the extent of transfers going into local economies. Together with original panel data collected on households, enterprises, local leaders, markets, and prices, and administrative data, this will enable us to study the direct, spillover, and general equilibrium effects of the cash across the study area. Phase I of the project covers all of Chiradzulu district (approximately 900 villages and a population of 400,000), where transfers will amount to approximately 70 - 100% of GDP in treated villages, and 40% of district-level GDP. The project and study may expand to additional districts in the future.

Intervention

The intervention are one-time lump-sum cash transfers. They are unconditional, universal (all adults are eligible), and delivered to individuals via mobile money in two tranches of 50 USD and 500 USD approximately one, and three months after enrolment. They amount to approximately 550 USD per individual, though this will vary somewhat with inflation and the exchange rate.

In addition to cash transfers, the project will test several complementary supply-side interventions, including information provision about upcoming transfers and demand increases to local businesses, as well as business grants to local enterprises.

Experimental design

Randomization of cash transfers will be across 3 different geographic units (villages, GVHs, and TAs) in order to create variation in exposure to treatment not only at the village level, but also at higher geographic orders (including market catchment areas), allowing us to estimate treatment effect spillovers across larger geographies.

1. TA Saturation: Variation in the intensity of treatment across TAs generates variation in the geographic exposure to treatment of local economies. 80% of GVHs in high-saturation TAs will be treated, 40% in low-saturation TAs.

2. GVH Saturation: Within treated GVHs, approximately ⅔ of villages will be treated. Retaining control villages in treated GVHs allows us to estimate within GVH spillovers. In control GVHs, no villages will be treated, acting as pure controls.

3. Village-level treatment: Within treated villages, treatment is universal and at the individual level, i.e. every adult will be eligible to receive treatment.

This three-stage randomisation design generates substantial spatial variation in treatment exposure at various different geographies, ranging from regions with almost no eligibles being treated to regions where almost everyone is treated. We will use this spatial variation to study the impacts of cash inflows on non-recipients, and local economies and market catchment areas in the aggregate.

Design details

Randomization method
by computer
Randomization unit
As noted above, there are 3 levels of randomization for cash transfers: - TAs: high (80%) vs. low (40%) saturation - GVHs: treat (2/3 of villages treated) vs. control (no villages treated) - villages: treat (all eligible individuals treated) vs. control (no-one treated)
Clusters
- approx. 900 villages in Chiradzulu - approx. 110 GVHs in Chiradzulu - 10 TAs in Chiradzulu
Observations
Approximately 130,000 individuals will receive transfers. Approximately 1,000 businesses will receive information only, and approximately 2,000 businesses will receive information + business grants.
Arms
Households: We will sample approximately 12 households per village, or 11,000 households across the district. Household will be surveyed at baseline, midline (0 - 9 months after transfers), Endline 1 (9 - 24 months after transfers), and beyond. Enterprises: a) at markets: We will sample approximately 7,000 enterprises operating at all 77 weekly market centers within the district, including approximately 1,000 receiving information provision only, and approximately 2,250 receiving grant + information. These businesses will be surveyed at least 3-4 times within the first year, and approximately every 6 months thereafter. b) in villages: We will survey homestead-based enterprises as part of the household surveys, expecting approximately 2 - 3,000 enterprises. Additionally, we will survey 2 - 4 enterprises in villages, but outside homesteads in each village, for a total of approximately 3,000 enterprises.
Primary outcomes
This project will study the direct, spillover, and general equilibrium effects of cash transfers on households, enterprises, local markets, prices. It will also test the impact of ex-ante information and business grants on business growth and price setting, and how this in turn affects the impacts of cash at the local economy level. We will also study the impacts of the program on local public goods, politics, and electoral outcomes in the 2025 election. Analyses will be pre-specified in more detail in a series of pre-analyses plans to be filed on this page.

Cited by

No citation data available for this registration yet.

Keywords