Agro-Dealers as Information Agents: The Impact of Maize Seed Performance Data on Stocking and Sales

Last registered on March 18, 2025
View Trial History

Pre-Trial

Trial Information

General Information

Title
Agro-Dealers as Information Agents: The Impact of Maize Seed Performance Data on Stocking and Sales
RCT ID
AEARCTR-0015537
Initial registration date
March 10, 2025

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
March 18, 2025, 8:33 AM EDT

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

Locations

Country
Kenya
Region
Sub Saharan Africa

Primary Investigator

Name
Francisca Muteti
Affiliation
International Maize and Wheat Improvement Center (CIMMYT)
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Sarah Kariuki
PI Affiliation
International Maize and Wheat Improvement Center
Contact Investigator
PI Name
Michael Ndegwa
PI Affiliation
International Maize and Wheat Improvement Center
Contact Investigator
PI Name
Moti Jaleta
PI Affiliation
International Maize and Wheat Improvement Center
Contact Investigator

Additional Trial Information

Status
On going
Start date
2025-02-03
End date
2025-12-31
Keywords
Agriculture, Behavior
Additional Keywords
seed systems, value chain, agro-dealers
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Despite continuous advancements in maize breeding, the adoption of newer, higher-yielding hybrids in East Africa remains low, highlighting barriers to varietal turnover. Challenges across the seed value chain may be contributing to this slow adoption: farmers prefer familiar varieties, agro-dealers prioritize stocking seeds with established demand, and seed companies have little incentive to promote new hybrids. Objective seed performance data could help address these barriers by improving decision-making by the seed value chain actors, stimulating farmer demand for superior varieties, and encouraging greater investment in their production and marketing. This study examines whether agro-dealers, key intermediaries in the seed supply chain, can serve as effective information agents to improve farmers' access to information about and adoption of superior seed varieties. Using a randomized controlled trial (RCT) in three Kenyan counties, we test the impact of providing agro-dealers with maize seed performance data on their stocking decisions, sales, and information-sharing behavior. The findings will offer evidence on the role of private sector input sellers in bridging agricultural information gaps and accelerating the adoption of improved seed varieties.
External Link(s)

Registration Citation

Citation
Jaleta, Moti et al. 2025. "Agro-Dealers as Information Agents: The Impact of Maize Seed Performance Data on Stocking and Sales." AEA RCT Registry. March 18. https://doi.org/10.1257/rct.15537-1.0
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
2025-03-10
Intervention End Date
2025-03-25

Primary Outcomes

Primary Outcomes (end points)
i. Agro-dealer’s stock of the best-performing varieties
ii. Agro-dealer’s stock of the worst-performing varieties
iii. Agro-dealer’s sales of the best-performing varieties
iv. Agro-dealer’s sales of the worst-performing varieties
Primary Outcomes (explanation)
i. Agro-dealer’s stock of the best-performing varieties defined as i) a dummy =1 if an agro-dealer stocked any of the top two varieties and ii) the total amount of the best two performing varieties stocked iii) the share in the total stock of the best two-performing varieties according to the list in the information script.
ii. Agro-dealer’s stock of the worst-performing varieties defined as i) a dummy =1 if an agro-dealer stocked any of the bottom two varieties and ii) the total amount, and iii) the share in the total stock of the worst three-performing varieties according to the list in the information script.
iii. Agro-dealer’s sales of the best-performing varieties defined as i) a dummy =1 if an agro-dealer sold any of the top two varieties, and ii) the total amount of the best two products sold iii) the share in the total sales of the best three-performing varieties according to the list in the information script.
iv. Agro-dealer’s sales of the worst-performing varieties defined as i) a dummy =1 if an agro-dealer sold any of the bottom two varieties and ii) the share in the total sales of the worst three-performing varieties according to the list in the information script.

Secondary Outcomes

Secondary Outcomes (end points)
i. Agro-dealer’s efforts to promote the best-performing varieties to their customers
ii. Agro-dealer’s stocking intentions
Secondary Outcomes (explanation)
i. Agro-dealer’s efforts to promote the best-performing varieties to their customers defined as i) a dummy=1 if an agro-dealer shows the information script to a mystery shopper, ii) a dummy =1 if an agro-dealer mentions any of the top 2 varieties to a mystery shopper and iii) a dummy =1 if an agro-dealer mentions any of the bottom two varieties to a mystery shopper.
ii. Agro-dealer’s stocking intentions, which captures agro-dealers’ intended stock, defined as a dummy=1 if an agro-dealer wished to stock any of the best-performing varieties and ii) the amount of seed the agro-dealer intended to stock.

Experimental Design

Experimental Design
The study consists of two experimental groups: a control group (T0) and a treatment group, referred to as the varietal performance information group (T1). The agro-dealers assigned to the control group did not receive any form of intervention. The agro-dealers
assigned to the treatment group received information on the performance of various maize seed varieties cultivated by farmers during the March- July season in 2023. The information was presented to the agro-dealers in a tabular format, detailing the name of the variety, the
average yield reported by the farmers who grew that variety, the average yield achieved by the top 20% of farmers (serving as a proxy for the yield potential), and the variety’s maturity level. The information was presented on a laminated leaflet that the agro-dealer retained.
The yield data, collected from farmer reports through household surveys conducted from November to December 2023, was presented as county-level averages. Information on the maturity level of the variety was extracted from the Kenya Plant Health Inspectorate Service
(KEPHIS) National Crop Variety list, which is typically compiled using data collected during national performance trials conducted for licensing purposes.
The study was rolled out in February 2025. Agro-dealers were visited approximately one month before the beginning of the seed sales season - likely before they stock for the season - specifically in mid-February, and provided with the information intervention. The information intervention was implemented by trained enumerators.
Experimental Design Details
Not available
Randomization Method
Assignment to the two experimental groups was done at the market level. The aim was to minimize the possibility of within-market spillovers and to ensure that all the shops in a market have the same treatment status. The treatment assignment was stratified by county.
Randomization Unit
Market
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
160 markets
Sample size: planned number of observations
480 agro-dealers
Sample size (or number of clusters) by treatment arms
80 markets control, 80 markets treated with variety performance information
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We used the outcome on the share of the best-performing varieties in agro-dealers’ total stock for power calculations, drawing on data from a previous CIMMYT survey to calculate the average share of these varieties stocked by agro-dealers during the 2024 March season. Based on a mean of 0.12 (standard deviation 0.09) and an ICC of 0.270, we need 160 markets with an average of 3 shops each to achieve 80% power to detect a minimum effect of 0.03—a 25% increase relative to the control mean. 2. For outcome variables exhibiting extreme values, we will consider applying transformations, such as the inverse hyperbolic sine transformation and winsorization, to minimize the loss of power associated with these outliers.
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
CIMMYT Internal Research Ethics Committee
IRB Approval Date
2025-02-17
IRB Approval Number
IRB00012744
IRB Name
JKUAT Institutional Scientific and Ethical Review Committee
IRB Approval Date
2025-01-30
IRB Approval Number
JKU/ISERC/02316/1309
Analysis Plan

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

Request Information