Back to History Current Version

Impacts of localization of school meals procurement: Experimental evidence from a home grown school feeding program in Burundi

Last registered on December 23, 2024
View Trial History

Pre-Trial

Trial Information

General Information

Title
Impacts of localization of school meals procurement: Experimental evidence from a home grown school feeding program in Burundi
RCT ID
AEARCTR-0014620
Initial registration date
December 20, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 23, 2024, 1:30 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
DAHYEON JEONG
Affiliation
WORLD BANK
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Dahyeon Jeong
PI Affiliation
World Bank
Contact Investigator
PI Name
Hannah Uckat Uckat
PI Affiliation
World Bank
Contact Investigator
PI Name
Jonas Heirman
PI Affiliation
World Food Programme
Contact Investigator
PI Name
Florence Kondylis
PI Affiliation
World Bank
Contact Investigator
PI Name
Simone Lombardini
PI Affiliation
World Food Programme
Contact Investigator

Additional Trial Information

Status
On going
Start date
2023-09-01
End date
2026-12-31
Keywords
Agriculture, Education, Other
Additional Keywords
School feeding, child nutrition, market access
JEL code(s)
H57: Procurement; Q13: Agricultural Markets and Marketing
Secondary IDs
Prior work
This trial is based on or builds upon one or more prior RCTs.
Abstract
This study evaluates the impacts of decentralized school meal procurement from agricultural cooperatives on children and the local economy in Burundi using a two-dimensional impact evaluation. The research examines how this procurement model affects the local economy—including cooperatives and farmers—alongside children's nutrition, health, and education outcomes.

The intervention replaces the centralized delivery of school meal crops by the World Food Programme (WFP) with a system where schools procure these crops using Commodity Vouchers (CVs). Impacts on schools and children are assessed through a randomized controlled trial (RCT) involving 243 schools across 200 school clusters, randomly assigned to either the decentralized procurement model or the status quo. Simultaneously, impacts on the local economy are evaluated by randomizing contract awards during competitive tendering at the district level. Eligible cooperatives compete for contracts to supply up to three key crops: maize, rice, and beans.

The study leverages extensive survey and secondary data, including digitized meal and attendance records from schools, anthropometric and cognitive assessments of children, seasonal production and revenue data from cooperatives and farmer households, and commodity price data from local market vendors. This research builds on a pilot evaluation conducted in 2022–23 (AEARCTR-0011995).
External Link(s)

Registration Citation

Citation
Heirman, Jonas et al. 2024. "Impacts of localization of school meals procurement: Experimental evidence from a home grown school feeding program in Burundi." AEA RCT Registry. December 23. https://doi.org/10.1257/rct.14620-1.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2024-02-01
Intervention End Date
2026-06-30

Primary Outcomes

Primary Outcomes (end points)
Primary outcomes are measured across five domains:
1. Schools
* School feeding days (and share of meal days) per month
* School meal attendance and enrolment
* Nutritional composition: Global Diet Quality Score (GDQS)-Meal index

2. Children
* Child absences
* Nutrition: food consumption, dietary diversity and food security
* Standardized test scores (EGRA, EGMA, Stroop, Ravens CPM, Digit Span)
* Anthropometrics: Measures of height and weight constructed standardizing for age, referenced with growth charts.

3. Farmer households
* Seasonal production and sales per crop (price and quantity)
* Food consumption expenditure

4. Cooperatives
* Seasonal aggregate production and revenue per buyer
* Share of revenue from school feeding
* Distribution of surplus at member level

5. Markets
* Commodity prices of school feeding crops and household staples
Primary Outcomes (explanation)
1. School meal diversity will be constructed using the Global Diet Quality Meal Score (GDQS-Meal) as per Intake.org guidelines.
2. Child anthropometrics will be referenced against WHO growth charts to determine height-for-age, weight-for-age, and BMI-for-age measures.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The experiment employs a two-dimensional randomization design:

1. School-Level Assignment
* Randomization unit: Clustered randomization at the school compound level stratified by province. A compound is a cluster of schools, often sharing a kitchen.
* Sampling frame: 243 schools (200 school compounds) across 4 provinces in Burundi
* Treatment arm: Decentralized procurement through Commodity Vouchers (CV)
* Control arm: Centralized procurement through WFP

2. Cooperative-Level Assignment
* Random assignment of school feeding contracts following a competitive tendering process open to 29 eligible cooperatives
* Random allocation of qualified bids to win up to three school feeding contracts
* Separate randomization for three staple crops: beans, maize, and rice stratified by province
* Treatment: Cooperative bids that were randomly awarded a contract for supplying crops directly to treatment schools
* Control arm: Cooperatives that did not win a school feeding contract
* Design maintains experimental validity under supply-demand matching and geographical constraints
Experimental Design Details
Not available
Randomization Method
The sampling and randomization were conducted using Stata 18.
Randomization Unit
Schools are randomized at the school compound level.
Cooperative randomization is assigned stratified by crop.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
243 schools
34 cooperative bids
Sample size: planned number of observations
243 schools 2430 children 60 cooperatives 300 farmers 50 agricultural markets
Sample size (or number of clusters) by treatment arms
120 control schools (98 compounds)
123 treatment schools (102 compounds)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
HML IRB
IRB Approval Date
2024-01-12
IRB Approval Number
2475
Analysis Plan