The effect of timely cash transfers during an extreme flood shock: Evidence from the 2022 Pakistan Floods

How to cite

Zizzamia, Rocco; Rafique, Nihan; Pople, Ashley; Khalid, Muhammad Bin. 2023. "The effect of timely cash transfers during an extreme flood shock: Evidence from the 2022 Pakistan Floods." AEA RCT Registry. AEARCTR-0012147. https://doi.org/10.1257/rct.12147-1.0

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

Rocco Zizzamia* , Nihan Rafique , Ashley Pople , Muhammad Bin Khalid

* Primary investigator

Registry ID
AEARCTR-0012147
DOI
https://doi.org/10.1257/rct.12147-1.0
First registered
Last updated
Status
In development
Trial dates
2023-10-02 – 2025-09-14
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12147

Abstract

Climate change driven natural disasters are pushing an estimated 26 million people into poverty every year and disproportionately impacting the world’s poorest countries. Typically, government and humanitarian responses to climate disasters are mobilized after the shock occurs when much of the damage has already been done. There is reason to believe that acting early by providing cash payments – before or during a climate disaster – may help households prevent some of the damage. Improvements in early action systems provide an opportunity for governments and humanitarian actors to mitigate the worst humanitarian impacts of climate disasters.
During the 2022 flooding in Pakistan, the World Bank was able to mobilise $150 million to provide a one-off cash transfer ($88 per recipient) to 1.3 million households from September to November. Given that a state of emergency was only declared on August 25 and flooding lasted until at least October 2022, the speed of this social protection response was unusually rapid in the context of humanitarian assistance during floods. This project aims to utilise a quasi-experimental approach – a fuzzy regression discontinuity design to understand the impacts of timely assistance on household resilience to extreme climate shocks. While there is a large body literature on the role of cash transfers in cushioning the negative income effects of shocks, experimental evidence on the effectiveness of cash transfers in response to extreme weather events, especially one-off cash transfers, remains scarce. Most research on disaster response focuses largely on the role of cash transfers in facilitating recovery from extreme weather events. This research explores the potential for “shock-responsive” cash transfers to limit the welfare cost of the shock through adaptive actions during a crisis.

Intervention

The objective of this study is to evaluate the medium-run impact of providing timely cash payments during an extreme flooding event on household welfare and behavioural choices, such as migration.

Pakistan is ranked 8th among countries most affected by climate change, according to the Global Climate Risk Index 2021. In 2022, flooding in Pakistan resulted in a significant loss of life and widespread devastation, with economic damages being estimated at a minimum of $30 billion, equivalent to approximately 10% of Pakistan's GDP.

To mitigate the adverse impact of the flood, a one-off cash payment was made per eligible household through the Benazir Income Support Program (BISP) to enable beneficiaries to meet their urgent and essential needs including support for food, water, shelter etc. in addition to regular social protection transfers implemented through BISP. Recipients were eligible to receive grants if they scored 32 or lower on a proxy means test (PMT). This initiative was funded by the Pakistan Hydromet and Climate Services Project (PHCSP), a World Bank-supported operation with the objective of restoring ecosystems in selected areas and providing an immediate and effective response to an Eligible Crisis or Emergency. Using this threshold, approximately the bottom 40% of households in affected areas were eligible for support. Those who scored just below the PMT threshold of 32, and were therefore eligible to receive cash support, are likely to be very similar to those who scored just above the threshold, and who were therefore not eligible to receive cash assistance. By exploiting the large variation in support provided to households straddling this threshold, this study will be able to measure the impact of timely cash support during a flood on a range of welfare and other outcomes.

To do so, the study aims to administer a short survey to households whose PMT scores in 2022 rendered them (just) eligible to receive cash support as well as to households whose PMT scores rendered them (just) ineligible to receive cash support. Surveys will be administered prior to the 2023 monsoon season. This allows the project to measure the extent of household recovery over the longer term (approximately 8-9 months after cash was received) and prior to further potential disruption caused by flooding in the 2023 monsoon season.

Experimental design

For this project, we exploit a sharp discontinuity in the eligibility rule for receiving cash transfers during the 2022 Pakistan floods, funded through a World Bank programme. Specifically, the Benazir Income Support Programme (BISP) used a proxy means test (PMT) scoring system to determine eligibility for receiving emergency cash transfers during the 2022 flooding event, where those with a PMT score below 32 were eligible to receive the cash transfer and those with a PMT score at or above 32 were ineligible to receive the cash transfer. By comparing outcomes of households immediately on either side of this eligibility threshold, this experiment uses a fuzzy regression discontinuity design to estimate the welfare impacts of receiving cash on those just below the cut-off compared to those just above it, measured one year after the cash transfers were made. Household survey data will be collected from a sample of households just above and just below the eligibility cut-off. The sample will consist of 4,000 households from 200 selected villages in Sindh province (approximately 10 per village with a PMT score of 31 and 10 per village with a PMT score of 32).

Design details

Randomization method
Randomization using computer assignment.
Randomization unit
Household
Clusters
200 villages
Observations
4000 households
Arms
2000 households in the control unit and 2000 households in the treatment unit.
Primary outcomes
Household consumption, household asset holdings, migration, adaptation
Secondary outcomes
Income, labour supply, health, education, savings, access to credit, access to government and non-government social services

Cited by

No citation data available for this registration yet.

Keywords