Revitalizing rural Malawi: Experimental evaluation of the economic impacts of capital injections to village savings and loans associations

How to cite

Hirvonen, Kalle; Siyame, Edwin; Weerdt, Joachim De; Mtemwa, Chikumbutso; Sibande, Lonester; Venable, Elizabeth. 2024. "Revitalizing rural Malawi: Experimental evaluation of the economic impacts of capital injections to village savings and loans associations." AEA RCT Registry. AEARCTR-0014464. https://doi.org/10.1257/rct.14464-2.0

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

Kalle Hirvonen* , Edwin Siyame , Joachim De Weerdt , Chikumbutso Mtemwa , Lonester Sibande , Elizabeth Venable

* Primary investigator

Registry ID
AEARCTR-0014464
DOI
https://doi.org/10.1257/rct.14464-2.0
First registered
Last updated
Status
In development
Trial dates
2024-09-25 – 2026-11-30
Countries
Malawi ()
JEL codes
D14, O12, O17, J46
Registry record
http://www.socialscienceregistry.org/trials/14464

Abstract

This trial tests an innovative intervention to boost the lending capacity of village savings and loans associations (VSLAs) in rural Malawi. Our pre-analysis plan describes a cluster randomized control trial (RCT) in which VSLAs in treated communities will receive between $350 and $590 USD capital injection to boost the groups’ lending capacity, depending on the group size. Outcomes are assessed in five domains: financial inclusion, livelihoods, welfare, food security, and gender. The trial is designed to document spillover effects in local labor markets.

Intervention

The capital injection model intervention is a component of a broader USAID Growth Poles initiative focused on specific anchor firms. These firms are expected to serve as catalysts for economic development, aiming to mobilize investments and result in positive impacts for nearby communities. The VSLA capital injections are intended to help households in nearby communities to tap into the new economic opportunities created by the Growth Pole initiative. In total, there are more than 15 growth poles (or anchor firms) in which USAID investments take place. The capital injections that are part of this trial take place in communities that are in the vicinity of four growth poles: Ilovo Dwangwa in Nkhotakota district located in the Central region of Malawi; Tropha in Rumphi and Mzimba districts in Northern region; and Pyxus in Dowa and Kasungu districts in the Central region; and African Parks in Nkhotakota district.

The VSLA capital injection intervention is implemented by Women's Legal Resources Centre (WOLREC), a Malawi-based NGO, in collaboration with Palladium International, a global consulting firm with an operational presence in Malawi. The intervention grants a sum of $79 PPP per member in the VSLA group to expand their loan sizes. To qualify for the capital injection, VSLAs must adhere to specific criteria, including a 4-month loan term and an interest rate of at most 10 percent (equivalent to approximately 33 percent on an annualized basis). This capital injection program complements the regular VSLA group loans, which the groups will continue to provide under their original terms, typically characterized by higher annualized interest rates than what is offered through the capital injection scheme. In addition, the groups will be equipped with local bank accounts to enhance security and establish a connection to formal financial services. Moreover, VSLA members will receive training in financial literacy, as well as in business and VSLA management. The training package also includes Gender Action Learning Systems (GALS) trainings that aim to empower women to contribute to planning and decision-making within their households.

The capital injections and training will be rolled out between December 2024, i.e., shortly after the baseline survey.

Experimental design

Identification strategy
We evaluate the economic impacts of the capital injections using a cluster RCT design, where the cluster is defined as the Group Village Headman (GVH), the second-lowest administrative unit in Malawi. The implementation partner will generate a list of 150 eligible GVHs. Using this list, 75 GVHs will be randomized into treatment and 75 GVHs to control before the intervention begins. The randomization will be conducted by the research team using Stata. The randomization will be stratified at the growth pole level so that each of the four growth pole areas have an equal amount of treatment and control GVHs.

The capital injection will then be provided into all eligible VSLAs in the GVHs selected into treatment. The eligibility is determined based on a VSLA mapping exercise at the onset of the project. A VSLA is eligible to receive a capital injection if it has been in operation for at least 12 months prior to the mapping visit. In each GVH, we expect there to be a significant number of households that do not belong to a VSLA. We call these ‘non-VSLA households’ and will sample these households to measure local spillover effects.

Study arms
The study will have two arms:
• T1 – VSLA households in treatment GVHs: Households belonging to a VSLA group that is eligible and receives the capital injection.
• C1 – VSLA households in non-treatment GVHs: Households belonging to an eligible VSLA group, but no capital injection takes place in the GVH.

To measure the direct impacts of the program, we will compare the outcomes between households belonging to eligible VSLAs in treated communities (T1) and households belonging to eligible VSLAs in non-treated communities (C1).

To measure the local spillover effects, we will compare the outcomes between households that are not part of eligible VSLAs in treated communities (T2) and households not part of eligible VSLAs in non-treated communities (C2):
• T2 – Non-VSLA households in treatment GVHs: Households residing in GVHs in which capital injections are taking place, but they do not belong to an eligible VSLA.
• C2 – Non-VSLA households in non-treatment GVHs: Households residing in GVHs in which capital injections are not taking place and they do not belong to an eligible VSLA.

Both control and treatment clusters follow the same VSLA – and non-VSLA – eligibility criteria, thus creating ex ante comparable groups of eligible and ineligible households.

Design details

Randomization method
Randomization will be done in office by a computer.
Randomization unit
Group Village Headman )GVH=
Clusters
150 GVHs
Observations
The target sample is 2,100 households in 150 GVHs for the primary sample and 750 households in the same 150 GVHs for the ‘spillover’ sample, amounting to 14 VSL households and five non-VSL households in each GVH. To account for attrition, we add one additional VSL households and one additional non-VSL in each GVHs, resulting in a total sample size of 3,150 households.
Arms
75 GVHs
Primary outcomes
Household total savings Net income from non-agricultural businesses Household per capita consumption-expenditures Food security (Reduced Coping Strategies Index) An indicator capturing women’s control over income (based on A-WEAI questionnaire)
Secondary outcomes
Took a loan in the past 12 month (0/1) Annual share-out in the VSLA Number of non-agricultural businesses operated Number of days of selling ganyu labor Number of days of hiring ganyu labor Expenditure on agricultural inputs Livestock assets (Tropical Livestock Units) Durable asset index Number of meals by adults
Sponsors
Name: USAID Location: USA Url: https://www.usaid.gov/

Cited by

No citation data available for this registration yet.

Keywords