Learning vs. Costly Information Processing in Technology Adoption: Evidence from a Cluster Randomized Controlled Trial in India

Intervention 1 comprises a one-day training workshop focused on increasing farmers’ knowledge about the general features and benefits of using Early Sown Wheat Varieties (ESW) and Zero-Tillage Equipment (ZT), with the goal of facilitating uptake and usage of these two technologies.

Intervention 2 consists in providing handholding support to participants designed to assist individual famers in staying attentive to changes in environmental conditions and react in appropriate ways to any challenges arising on their wheat plots, thereby reducing uncertainty about optimal usage practices and timing of agricultural tasks related to ESW and ZT.

2024-05-01

2025-04-30

- Farmers’ adoption decisions of ESW
- Farmers’ adoption decisions of ZT
- Farmers’ perceptions towards ESW and ZT, including: (a) farmers’ general awareness of ESW and ZT, (b) farmers’ beliefs about whether ESW and ZT offer advantages over conventional wheat varieties and methods of soil preparation, and (c) farmers’ stated intention to use ESW or ZT in the future.

1) Wheat yields and occurrence of crop failure
2) Agricultural income from wheat cultivation (gross and net of input costs)
3) Use of agricultural inputs (manure, chemical fertilizers, herbicide, pesticide) on wheat plots
4) Irrigation practices and amount of water used on wheat plots
5) Labor input and management decisions for different tasks in wheat cultivation (including land preparation, sowing, weeding, irrigating, applying fertilizer and other inputs, harvesting, threshing, grain cleaning)
6) Agriculture waste disposal practices (incl. instance of crop residue burning)
7) Farmers’ perceptions towards the practice of residue burning and air pollution

We will conduct a cluster RCT with multiple treatment arms designed to test several hypotheses motivated by our theoretical framework. Randomization of the participating farmers into treatment arms will be conducted both between clusters (i.e., villages) and within clusters. Specifically, there will be three types of villages: Control villages, Treatment(1) villages, and Treatment(2) villages. The Control villages constitute a pure control group in which the participating farmers will be asked to respond to our baseline and endline surveys, but will not benefit from any intervention. The participants within each Treatment(1) village will be randomly assigned to one of two groups. The first group (T1) will receive Intervention 1, whereas the second group (C1) will receive no information treatment regarding ESW or ZT (to facilitate implementation and compliance, the farmers assigned to C1 will receive a placebo information treatment). Similarly, the participants within each Treatment(2) village will be randomly assigned to either the T2 or the C2 group. Participants in T2 will receive both Intervention 1 and Intervention 2, whereas the participants in C2 will only receive Intervention 1.

Not available

Randomization done in office by a computer

Step 1: randomly select 60 clusters (villages) from the village sampling frame.
Step 2: randomly assign 20 villages to Control, 20 villages to Treatment(1), and 20 villages to Treatment(2) groups, respectively.
Step 3: Within each Treatment(1) village, randomly assign half of the participants to T1 group and half to C1 group. Within each Treatment(2) village, randomly assign half of the participants to T2 group and half to C2 group.

Yes

60 villages

1200 farm households

20 villages Control, 20 villages Treatment(1), 20 villages Treatment(2)

For β=0.7 (statistical power) and K=0.1 (coefficient of variation of true proportions between clusters within each group), we will be able to detect a relative increase in the adoption rate of ESW of 6.6%-points (e.g., from a baseline of 14.4% to 21.3%) under Intervention 1, and an increase in the adoption rate of ZT of 8.2%-points (e.g., from a baseline of 24.5% to 32.7%), both at a statistical significance level of 5%. For β=0.8 and K=0.2, we will still be able to detect an increase of 8.1%-points for ESW and an increase of 9.8%-points for ZT, at a statistical significance level of 5%.