Social Incentives of Non-Elected Leaders in Village Economies

How to cite

Kwak, Minji. 2024. "Social Incentives of Non-Elected Leaders in Village Economies." AEA RCT Registry. AEARCTR-0014292. https://doi.org/10.1257/rct.14292-2.1

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

Minji Kwak*

* Primary investigator

Registry ID
AEARCTR-0014292
DOI
https://doi.org/10.1257/rct.14292-2.1
First registered
Last updated
Status
Completed
Trial dates
2024-09-09 – 2024-11-01
Countries
Malawi (Lilongwe)
JEL codes
D73, I38, O12, O22, O43
Registry record
http://www.socialscienceregistry.org/trials/14292

Abstract

This document outlines a lab-in-the-field randomized controlled trial (RCT) in rural Malawi, which explores the disparity between the beliefs of villagers and traditional chiefs regarding social subsidy allocation, where hereditary leadership lacks monetary incentives and electoral accountability. The trial also examines the social incentives of traditional chiefs concerning subsidy allocation and the impact of involving villagers and higher-level chiefs as community monitors on policy outcomes (fairness and efficiency) and citizen satisfaction. Specifically, this experiment includes two distinct parts. First, villagers and village chiefs are asked to rank 15 randomly chosen villager profiles in terms of who should receive the social subsidy the most and the least. This project compares the beliefs of village chiefs with those of villagers, as well as with the Proxy Means Test (PMT) score. Following this ranking exercise, the first intervention takes place in the second part: the chiefs are asked to rank another set of 15 randomly chosen villager profiles, with the knowledge that their final rankings will be publicly announced to the villagers through public postings. The second intervention involves informing the chiefs that their final allocation decisions will be shared with their superiors (higher-ranked village chiefs). By comparing the rankings of village chiefs from the first and second parts, this project explores ways to improve the accountability of traditional chiefs.

Intervention

Household head profiles have been randomly assigned to one control group and two treatment groups. Randomly selected villagers in each village, along with the GVH, rank household profiles in each treatment group based on their own thoughts about who should receive the social subsidy. Meanwhile, the VH also ranks the household head profiles, but with different instructions. For the profiles in the control group, the VH ranks them based on their own thoughts, just like the villagers and the GVH. However, for the household head profiles in treatment group 1, there is an additional instruction that the VH's rankings will be shared with the villagers through public postings. Finally, the profiles in treatment group 2 come with the instruction that the VH's rankings will be shared with their superiors, the GVH.

Experimental design

This experiment includes two distinct parts. First, villagers and village chiefs are asked to rank randomly chosen villager profiles in terms of who should receive the social subsidy the most and the least.
Following this ranking exercise, the first intervention takes place in the second part: the chiefs are asked to rank another set of randomly chosen villager profiles, with the knowledge that their final rankings will be publicly announced to the villagers through public postings. The second intervention involves informing the chiefs that their final allocation decisions will be shared with their superiors (higher-ranked village chiefs).

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Profiles are randomized in an individual level.
Clusters
20 groups villages, each containing 10 villages, with a total of 1,000 villagers, 200 Village Headman (VH), and 20 Group Village Headman (GVH).
Observations
1220 individuals, 3660 sets, and 54900 profiles
Arms
15 profiles in control group, 15 profiles in treatment 1 group (sharing with villagers), and 15 profiles in treatment 2 group (sharing with superiors)
Primary outcomes
1. Correlation between the rankings from villagers and traditional chiefs with the ranking calculated using the PMT. 2. Rank of profile in each treatment group 3. Correlation between the VH's rank and the PMT-calculated rank, the correlation between the VH's rank and the villagers' rank, and the correlation between the VH's rank and the GVH's rank.

Cited by

No citation data available for this registration yet.

Keywords