AEA RC T Registr y

The American Economic Association's registry for randomized controlled trials

2010-02-01

2011-12-31

Knowledge Retention by Communicators (Acquiring and Sending A Signal in the Model)

- Dependent Variable: Communicators' Knowledge scores


Communicator Effort

- Dependent variable: Designated communicator held at least one activity


Knowledge After One Season Among Recipient Farmers

- Dependent Variable: Knowledge scores in household survey


Adoption After Two Seasons

- Dependent Variable 1: Used on at least one household plot in 2010/11

- Dependent Variable 2: Directly observed usage on at least one plot in 2010/11

- Dependent Variable 3: Plan to use next year

- Dependent Variable 4: Household produced at least compost heap


Communicators per HH

- Dependent Variable: Household adopted target technology in 2010/11 season


Types of Target Farmers Persuaded by PFs with and without Incentives

- Dependent Variable: Household adopted target technology in 2010/11 season


Communicator Adoption

- Dependent Variable 1: Comm used tech

- Dependent Variable 2: Non-comm HH used tech

- Dependent Variable 3: Comm used tech (share of PFs)

- Dependent Variable 4: Non-comm HH used tech

This project uses a randomized controlled trial to explore strategies of promoting new technologies using the power of social influence. The randomized controlled trial varied the dissemination method for two new agricultural technologies across villages in Malawi. The role of main communicator about the new technology was assigned to government-employed extension workers, 'lead farmers' who are educated and able to sustain experimentation costs, or 'peer farmers,' who are more representative of the general population and whose experiences may be more applicable to the average recipient farmer's own conditions. Random subsets of these communicators were offered performance-based incentives in the experimental design.

Randomization was done on a computer using Excel and Stata.

Randomization of treatment groups occurred at the village level

Yes

168 villages

Number of observations varies depending on the outcome variable

Extension worker

- Incentive: 13 villages

- No Incentive: 12 villages


Lead Farmer

- Incentive: 25 villages

- No Incentive: 25 villages


Peer Farmer

- Incentive: 23 villages

- No Incentive: 22 villages


Control

- 48 villages

Low adoption of agricultural technologies holds large productivity consequences for developing countries.
Agricultural extension services counter information failures by deploying external agents to communicate
with farmers. However, social networks are recognized as the most credible source of information
about new technologies. We incorporate social learning in extension policy using a large-scale field
experiment in which we communicate to farmers using different members of social networks. We
show that communicator effort is susceptible to small performance incentives, and the social identity
of the communicator influences learning and adoption. Farmers find communicators who face agricultural
conditions and constraints most comparable to themselves to be the most persuasive. Incorporating
communication dynamics can take the influential social learning literature in a more policy-relevant
direction.

BenYishay, Ariel, and A. Mushfiq Mobarak. "Social Learning and Communication." Working Paper, June 2015.