What is the value of early response? Multi-country RCT of shock-responsive cash transfers in the Sahel

How to cite

Pople, Ashley; Premand, Patrick; Dercon, Stefan; Bossuroy, Thomas; Brunelin, Stéphanie; Vinez, Margaux. 2022. "What is the value of early response? Multi-country RCT of shock-responsive cash transfers in the Sahel." AEA RCT Registry. AEARCTR-0010097. https://doi.org/10.1257/rct.10097-1.1

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

Ashley Pople* , Patrick Premand , Stefan Dercon , Thomas Bossuroy , Stéphanie Brunelin , Margaux Vinez

* Primary investigator

Registry ID
AEARCTR-0010097
DOI
https://doi.org/10.1257/rct.10097-1.1
First registered
Last updated
Status
On going
Trial dates
2022-01-15 – 2026-12-31
Countries
Mauritania (), Niger (), Senegal ()
JEL codes
I38: Government Policy – Provision and Effects of Welfare Programs
Registry record
http://www.socialscienceregistry.org/trials/10097

Abstract

Shocks such as extreme weather events are becoming more intense and more frequent with climate change, with well-documented long-lasting implications for welfare and poverty. However, ex-post responses to such shocks are typically provided at the height of food insecurity during the agricultural lean season and often reach households late. More recently, several governments and humanitarian actors are experimenting with early response interventions based on rapidly available satellite-based triggers to strengthen households’ ability to manage shocks, such as through shock-responsive cash transfers. Such approaches are at the core of efforts to build adaptive social protection systems.

This study coordinates a multi-country randomized control trial (RCT) in the Sahel to assess the value of early response and evaluate the impact of the timing of cash transfers in mitigating the negative effects of extreme drought on household welfare. The study is integrated in the Sahel Adaptive Social Protection Program managed by the World Bank and implemented in close coordination with government agencies across the Sahel. We test the relative effectiveness of early responses to drought (providing temporary cash transfers after the weather shock but before its peak effect on household food security and welfare) versus the traditional response to drought (providing temporary cash transfers at the peak of the food insecurity crisis).

Intervention

The study assesses the value of early response by evaluating modalities of shock-responsive social protection programmes that deliver cash transfers either early or later during the lean season. These programmes are being implemented by governments in the countries of interest. They are targeted to geographical areas hit by failures in the rainy season identified through satellite-based triggers. Social protection is expanded to these areas to mitigate the worst impacts of failed rains on welfare insecurity the following year, and in particular during the next lean season.

We test the impact of timing across three cash transfer interventions:

T1. “Early cash transfers” entail cash transfers delivered in advance of the lean season.

T2. “Late cash transfers” entail cash transfers delivered during the lean season.

T3. “Early and late cash transfers” entail cash transfers delivered both before and during the lean season.

In total, the cash transfers amount to roughly 10 to 15% of annual household consumption. The total value of the cash transfers is the same across the three modalities. T1 and T2 entail relatively larger cash transfers, but fewer payments, compared to T3, which entails relatively smaller cash transfers, but more payments.

Households are targeted based on their baseline vulnerability ahead of all three interventions, either through a PMT score, community-based targeting, or a combination of both. Households receive cash transfers, but no additional messaging about drought response.

Experimental design

We aim to measure the impact of the timing of cash transfers on the welfare impact of an extreme drought. In each country, communes/districts will be selected based on their exposure to an extreme weather event. Selected villages within communes/districts will be randomized into three groups:

T1. “Early cash transfers” entail cash transfers delivered in advance of the lean season.

T2. “Late cash transfers” entail cash transfers delivered during the lean season.

T3. “Early and late cash transfers” entail cash transfers delivered both before and during the lean season.

Households will receive the cash transfers and will be targeted based on their baseline vulnerability. Eligible households for all three treatments will be pre-identified at the same time before the start of the first treatment arm. The total value of the cash transfers will be kept constant across the three treatment arms.

Design details

Randomization method
Randomization will be carried out in an office with a computer using STATA.
Randomization unit
Smallest possible admin level (e.g. village or community or neighbourhood)
Clusters
160-180 villages
Observations
4,000-4,500 households
Arms
Treatment 1 (early cash transfers): 1330 to 1500 Treatment 2 (late cash transfers): 1330 to 1500 Treatment 3 (early and late cash transfers): 1330 to 1500
Primary outcomes
1. Household food security, as proxied by the food consumption score 2. Subjective well-being, as proxied by Cantril’s ladder of life satisfaction 3. Household food consumption
Secondary outcomes
Please see the pre-analysis plan for more details on secondary outcomes.
Sponsors
Name: World Bank Adaptive Social Protection Program
Partners
Name: DIME, World Bank Partner_type: none Name: University of Oxford Partner_type: none

Cited by

No citation data available for this registration yet.

Keywords