Research allocation
Vary village voucher access
Random village allocation
- 70 low-intensity villages
- Invite 13 farmers per village.
- 70 high-intensity villages
- Invite 34 farmers per village.
- 60 comparison villages
- No equipment vouchers or cash grants.
Illustrative summary of the source-described allocation across 190 villages. Survey all groups on machinery rental and labour reallocation.
See source →How
When to use
Program evaluators want to measure the impact of agricultural input subsidies at different community saturation levels. Spillovers or rental market responses might depend on how many community members receive subsidy access. Varying intervention density across villages reveals whether high or low local saturation alters program adoption.
Randomize intervention intensity across community clusters to evaluate technology adoption at different saturation levels.
- Select target villages across agricultural districts where smallholders farm land holdings under four acres.
- Assign villages randomly into low-intensity access, high-intensity access, and comparison groups.
- Invite thirteen farmers per village in low-intensity villages to receive agricultural equipment rental vouchers.
- Invite thirty-four farmers per village in high-intensity villages to receive agricultural equipment rental vouchers.
- Maintain comparison villages where farmers receive neither equipment vouchers nor cash grants.
- Conduct surveys across all groups to measure machinery rental usage and labor reallocation.
Why it works
Comparing low and high intensity against control communities isolates saturation effects from general market trends.
Example
Researchers assigned 190 villages with 7,100 farmers in Karnataka, India, into 70 low-intensity villages (13 farmers invited), 70 high-intensity villages (34 farmers invited), and 60 comparison villages. Farmers offered vouchers increased rental adoption from 11 percent to 41 percent.
About the study
Participants
7,100 small-scale farmers across 190 villages in eight districts of Karnataka, India, with 5,500 completing the final phone survey.
Study Design
Randomized controlled trial assigning 190 villages to low-intensity access (70 villages), high-intensity access (70 villages), or a comparison group (60 villages).
Results
Equipment rental vouchers increased the likelihood that farmers rented equipment by 30 percentage points relative to 11 percent of comparison households using machines.