Overcoming Gender-Specific Constraints to Agricultural Technology Adoption: New approaches to Extension and Seed Delivery for Female Kenyan Farmers

How to cite

Trachtman, Carly; Murigi, Michael; Nganga, Muthoni; Kivuva, Benjamin. 2024. "Overcoming Gender-Specific Constraints to Agricultural Technology Adoption: New approaches to Extension and Seed Delivery for Female Kenyan Farmers." AEA RCT Registry. AEARCTR-0015087. https://doi.org/10.1257/rct.15087-1.0

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

Carly Trachtman* , Michael Murigi , Muthoni Nganga , Benjamin Kivuva

* Primary investigator

Registry ID
AEARCTR-0015087
DOI
https://doi.org/10.1257/rct.15087-1.0
First registered
Last updated
Status
On going
Trial dates
2024-12-02 – 2026-05-31
Countries
Kenya (Lower Murang'a Country)
JEL codes
O13, O33, Q12, J16
Registry record
http://www.socialscienceregistry.org/trials/15087

Abstract

Women smallholder farmers in low-income countries often have lower adoption rates of profitable agricultural technologies, in part due to unequal access to information about these technologies. In this experiment, we test variations of a female-led peer extension model, which provides training and free seeds, to promote the adoption of improved cassava and bean varieties among female farmers. We cross-randomize two components of the extension model in a 2x2 factorial design: 1) Household agricultural extension visits conducted by female (rather than male) trainers, and 2) community election of these trainers (rather than selection by local authorities). We also compare the results with a pure control group (receiving no extension at all) and a group that receives the seeds, but without training. We explore the impact each treatment has on adoption of the promoted technologies in the subsequent season.

Intervention

We test the efficacy of a women-led peer extension model to promote the adoption of agricultural technologies among female farmers. We experimentally vary two key gender-sensitive components: 1) Household agricultural extension visits conducted by female trainers (as opposed to male trainers), and 2) community election of these trainers (as opposed to appointed by local leaders). The extension model will be used to promote either improved, drought tolerant cassava or high-iron beans (randomized at the cluster level).

Experimental design

The experiment is a cluster randomized controlled trial (cRCT), in which identification of treatment effects come from random assignment. There are 6 main treatment conditions, as follow:
1. C--Pure Control (50 villages): Receive no intervention.
2. T1--Seed Delivery Only (50 villages): Receive free seeds only, no training.
3. T2--Extension w/no gender reservation and no community selection (100 villages): In these villages, we will ask local leaders to nominate 2 individuals to be ``lead farmers'' that will provide peer extension (including free seeds) to women in their village.
4. T3--Extension w/gender reservation and no community selection (100 villages): In these villages, we will ask local leaders to nominate 2 individuals to be ``lead farmers'' that will provide peer extension (including free seeds) to women in their village, where at least one of the lead farmers nominated must be a woman.
5. T4--Extension w/no gender reservation and community selection} (100 villages): In these villages, we will invite community members to a meeting in which they will elect 2 individuals to be ``lead farmers'' that will provide peer extension (including free seeds) to women in their village.
6. T5--Extension w/ gender reservation and community selection (100 villages): In these villages, we will invite community members to a meeting in which they will elect 2 individuals to be ``lead farmers'' that will provide peer extension (including free seeds) to women in their village, where at least one of the lead farmers elected must be a woman.

Design details

Randomization method
Done in office by computer
Randomization unit
Village
Clusters
500 villages
Observations
10, 000 (500 clusters with 20 household each)
Arms
Control: 50 clusters, T1: 50 clusters, T2: 100 clusters, T3: 100 clusters, T4: 100 clusters, T5: 100 clusters
Primary outcomes
Adoption of improved variety promoted in cluster
Secondary outcomes
Alternate adoption indicators, agricultural outcomes, welfare outcomes, training delivery compliance, training quality perceptions, gender/empowerment indicators, norms indicators

Cited by

No citation data available for this registration yet.

Keywords