Impact of participation in a farming contract on rice farmers’ livelihood in Benin: A Randomized Control Trial approach

How to cite

Michler, Jeffrey; Arouna, Aminou. 2017. "Impact of participation in a farming contract on rice farmers’ livelihood in Benin: A Randomized Control Trial approach." AEA RCT Registry. AEARCTR-0002619. https://doi.org/10.1257/rct.2619-4.0

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

Jeffrey Michler* , Aminou Arouna

* Primary investigator

Registry ID
AEARCTR-0002619
DOI
https://doi.org/10.1257/rct.2619-4.0
First registered
Last updated
Status
Completed
Trial dates
2016-06-02 – 2018-06-30
Countries
Benin (Savalou, Bante)
JEL codes
C93, D13, L14, O13, Q12
Registry record
http://www.socialscienceregistry.org/trials/2619

Abstract

In the last five years, contract farming has become an increasingly common way for small producers in developing countries to ensure the supply of inputs, obtain new technologies, and gain access to markets. While these three elements are important to increasing productivity in developing country agricultural, little knowledge exists regarding 1) which element is most vital to the agent (farmer), 2) which is most vital to the principal, 3) and what are the relevant welfare impacts on both sides. In this project, we will work with a local NGO to develop and offer three different types of contracts designed to provide answers to these questions.

Rice farming is relatively new to Benin and productivity there lags behind productivity levels in Asia. Government, policy analysts, NGOs, development agencies, and research institutes have invested heavily in rice production and all have proposed numerous reasons why output remains low. Three key potential reasons emerge from a review of this literature. First is that output markets are thin, making farmers uncertain of their ability to sell their harvest at a profitable price. This uncertainty limits investment in inputs and depresses productivity. Second is that, since rice cultivation is relatively new to Benin, farmers lack the skills and knowledge to cultivate rice in a highly productive way. Without skills training or technical backstopping, farmers remain below the global production possibilities frontier. Third is that farmers are credit constrained and unable to purchase improved inputs when they are needed. Working with a local NGO that is already engaged in contract rice farming, we have developed three types of contracts designed to uncovered what are the binding constraints to rice production in Benin.

Intervention

The main aspect of the treatment to be randomized is what benefits the contract offers to the farmers. The control contract will offer to purchase a given amount of the rice harvest at a given price. The two treatment arms offer technical training (T1) or inputs (T2), in combination with a guarantee to purchase a given amount of the rice harvest at a given price, less the costs of the training and/or inputs.

Experimental design

Working with a local NGO that is already engaged in contract rice farming, we have developed three types of contracts designed to uncovered what are the binding constraints to rice production in Benin. In the first contract, the control, the NGO offers a guarantee to purchase a given amount of the rice harvest of a farmer at a given price. Comparing levels of rice production under this contract with rice production data collected in the baseline, we can determine the impact on investment and production of providing a guaranteed output market. In the second contract, treatment 1, the NGO offers to provide skills training and technical backstopping in exchange for purchasing a given amount of the rice harvest of a farmer at a given price. Comparing levels of rice production under this contract with baseline production and production in the control group, we can determine the impact on investment and production of moving farmers closer to the production possibilities frontier. In the third contract, treatment 2, the NGO offers to provide inputs, on loan, in exchange for purchasing a given amount of the rice harvest of a farmer at a given price. Comparing levels of rice production under this contract with baseline production and production in the control group, we can determine the impact on investment and production of easing farmers' credit constraints.

Design details

Randomization method
From a census of all rice farmers in Savalou district, we will contact each farmer in order to obtain their voluntary consent to participate in the experiment. Within each village, the subset of willing farmers will be randomly assigned into producer groups of no more than 8 farmers. At this point, we will gather baseline data on all willing farmers. These producer groups will be entered into a spreadsheet and randomly assigned to treatment. In a follow-up visit, we will be accompanied by officers from the NGO with contracts. Based on the previously assigned random status, the NGO officers will offer contracts to the producer groups.
Randomization unit
The unit of randomization is the individual farmer with treatment clustered at the producer group level.
Clusters
96 producer groups
Observations
768 farmers
Arms
22 producer groups control, 37 producer groups in T1, 37 producer groups in T2
Primary outcomes
Our primary analysis will test the difference in rice production under the three randomized states. We will also look at heterogeneity of take-up by each of the following covariates. - Rate of inter-temporal substitution of income and time inconsistency of this rate. - Wealth (land ownership, number of livestock, quality of housing stock) - Index of experience and satisfaction with contract farming prior to the experiment - Historical production levels - Household characteristics (gender of head of household, education level, experience with rice farming, household size) If any of these variables presents a particularly skewed distribution, we will report the heterogeneity by the trimmed variable or by deciles. In addition, if the production levels in T1 and T2 are similar, we will bundle these two treatments to maximize power on the heterogeneity analysis.
Secondary outcomes
Our secondary analysis will test the differences in household welfare under the three randomized states. We will measure household welfare as income (farm, non-farm, expenditure, and consumption) and an index of food security measured using a subset of the USDA's standard questionnaire. We will also look at heterogeneity of take-up by each of the covariates previously listed.

Materials

Papers from this trial

  1. Arouna, A., Michler, J.D., and Lokossou, J.C. 2021. "Contract Farming and Rural Transformation: Evidence from a Field Experiment in Benin." Journal of Development Economics 151: 102626. https://doi.org/10.1016/j.jdeveco.2021.102626

Cited by

No citation data available for this registration yet.

Keywords