Scaling Agricultural Inputs in Mali

How to cite

Dillon, Andrew; Tomaselli, Nicolo. 2023. "Scaling Agricultural Inputs in Mali." AEA RCT Registry. AEARCTR-0010995. https://doi.org/10.1257/rct.10995-2.0

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

Andrew Dillon* , Nicolo Tomaselli

* Primary investigator

Registry ID
AEARCTR-0010995
DOI
https://doi.org/10.1257/rct.10995-2.0
First registered
Last updated
Status
On going
Trial dates
2023-01-10 – 2025-08-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/10995

Abstract

In Mali, rural agricultural input markets are largely missing, leading to productivity and welfare losses. Ag-dealers are constrained in opening new input markets as transportation costs and demand uncertainty are both high. Furthermore, liquidity in rural markets is low as banks face high loan costs per client.

This project will evaluate the Village Input Fairs (VIF) model as it scales. The VIF provides a solution to creating agricultural input markets by easing multiple constraints on the demand and supply sides. On the demand side, farmers use early commitment contracts to order inputs in the post-harvest period for delivery in the planting season and have access to credit. On the supply side, ag-dealers select VIFs to supply which are coordinated by private VIF enterprises. Private VIF enterprises are the market-makers, providing ag-dealers access to VIF markets, coordinating with microfinance partners, and generating revenue from the license sales to cover their costs.

This research will answer questions about the effectiveness of private sector scaling strategies on input demand and supply side constraints of ag-dealers.

Intervention

This project will scale the Village Input Fair (VIF) model which integrates market organization principles of timing, liquidity, supply chain reliability, and interlinked transactions. VIFs are organized in the post-harvest period, in February or March. Ag-dealers participate in fairs to collect purchase orders from farmers for agricultural inputs, accompanied by a 10% deposit commitment, with the balance paid upon delivery. Microfinance loan officers, also present at the fair, play two roles. First, they collect farmers' deposits, hold them in blocked accounts, and release payment upon delivery of inputs. Second, they offer farmers the opportunity to apply for credit, with their loan formalized upon delivery of the inputs. Input delivery occurs in June, before the start of the rainy season. The selection of ag-dealers who will have the opportunity to sell their products at the fair will be selected through an auction-based licensing system. A market-making enterprise will be selected through a business competition. The enterprise coordinates amongst ag-dealers to organize the auction for licenses to participate in the fair and organizes the village input fair.

Experimental design

We select and randomize 150 villages into two market cohorts. In Year 1 of the experiment 75 village clusters will be assigned to treatment and 75 villages to control. In year 2, the 75 control villages from year 1 will also receive the treatment. In each treatment villages all households will have the opportunity to participate and buy agricultural inputs at the fair. Data for household and plot level outcomes will be collected through a household survey in 15 households in each village.

Update: The study has expanded to include a third market cohort where 40 additional villages will be selected to participate. Of the 40 villages, 25 villages will be assigned to treatment and 15 to control.

Design details

Randomization method
Randomization done in office on a computer using Stata.
Randomization unit
The treatment is randomized at the village level. The villages are stratified and divided according to observable characteristics such as population and distance to the paved roadway.
Clusters
150 village clusters Update: Additional 40 village clusters from a new region were added to the study
Observations
2250 households and 300 agro-dealers Update: The new region added to the study will survey 600 households (15 households per village)
Arms
In year 1 of the experiment 75 village clusters will be assigned to treatment and 75 villages to control. In year 2, the 75 control villages from year 1 will also receive the treatment. In treatment villages all households will have the opportunity to participate and buy agricultural inputs at the fair. Update: Additional 25 village clusters from the new region will be assigned to treatment.
Primary outcomes
The experiment focuses on outcomes at three levels: household level, ag-dealer level, and market level. The main treatment effects are at the household level, but information at the dealer and market levels is important to understand scalability. At the household and plot level we measure (a) farmers’ demand measured by participation during village input fairs and sales volume for inputs, (b) the extensive and intensive margin of farmers’ fertilizer usage during two subsequent agricultural seasons, and (c) farmers’ crop choices and agricultural yields. Input decisions and production will be measured at the plot level. At the market level we measure transactions at the input fairs, including buyers, sellers, products demanded, prices, and quantities. At the ag-dealer level we collect information on profits, costs, sales, revenues, employment, and the input supply chain.
Secondary outcomes
Welfare effects including potential income gains, food security, and poverty reduction.

Cited by

No citation data available for this registration yet.

Keywords