NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
The Power of Agency and Agricultural Transformation: Evidence from India
Last registered on July 15, 2019


Trial Information
General Information
The Power of Agency and Agricultural Transformation: Evidence from India
Initial registration date
July 13, 2019
Last updated
July 15, 2019 9:19 AM EDT

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
University of Chicago
Other Primary Investigator(s)
PI Affiliation
University of California at Berkeley
PI Affiliation
The World Bank
PI Affiliation
University of California at Berkeley
Additional Trial Information
In development
Start date
End date
Secondary IDs
Community-driven development (CDD) emphasizes a “bottom-up” approach focusing on community control over planning and implementation decisions to improve development outcomes. We propose to assess the value of community participation in choosing the location and implementation of local infrastructure projects by comparing a participatory “bottom-up” approach to the standard “top-down” planning and implementation where the community only plays a limited role. The context for the study is the construction of minor irrigation channels in the command area of small irrigation tanks in Telangana. The government has recently launched a program to rehabilitate the storage capacity of the tanks but has left the distribution of water from the tank to plots in the command area to the farmers. The research project thus varies the extent of community involvement over the choice and implementation of field channels and assesses its impact on irrigation resource allocation, agricultural outcomes, and continued maintenance.
External Link(s)
Registration Citation
Gine, Xavier et al. 2019. "The Power of Agency and Agricultural Transformation: Evidence from India." AEA RCT Registry. July 15. https://doi.org/10.1257/rct.4336-1.0.
Former Citation
Gine, Xavier et al. 2019. "The Power of Agency and Agricultural Transformation: Evidence from India." AEA RCT Registry. July 15. http://www.socialscienceregistry.org/trials/4336/history/50019.
Experimental Details
The intervention varies how the location of channels is chosen (choice) crossed with how they are constructed (implementation). All treatments are randomized at the tank-ayacut level and since we only include one tank in the study per village, treatments are also randomized at the village level. We describe each of the intervention and control groups in detail below.

1A. Top-down choice: Channel location is chosen by a third party (the research team presenting itself as an NGO to the village). In practice, the chosen channel is the one with highest aggregate valuation from the private elicitation module collected during the baseline. As mentioned, farmers in Top-down choice villages, should feel as if they had no say in the location decision of the channel.

1B. Bottom-up choice (Secret Ballot): The location of the channel is chosen by ayacut farmers via a secret ballot at a designated polling station. Farmers choose among the top three choices from the elicitation process. We expect that the channel that maximizes aggregate private valuations will be chosen by secret ballot and therefore both 1A and 1B should choose the same channel but under different levels of community involvement.

2A: Top-down implementation: Channel construction in these villages will be carried out by an agency/third party/ NGO identified by the research team. We will monitor the implementation closely and assess progress on the ground periodically.

2B: Bottom-up implementation: The ayacut leaders (VRA + key farmers) will receive and manage the funds for the construction of the channel and are responsible for its implementation. We will not intervene in the process of implementation in this arm.

In addition to the 2x2 matrix implied by the two-part randomization above, the design also includes a pure control group. Farmers in this control group will receive the cash equivalent of the per capita cost of the total project. Across experimental arms, we will collect data on various metrics at various points in time.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
After the intervention, the following outcomes will be collected through multiple rounds of survey: (a) whether the channel gets built, (b) labor/management/monitoring contribution by the community to channel construction, (c) quality and timeliness of channel construction, (d) attitudes and satisfaction - both with respect to the choice of the channel as well as its implementation, and (e) trust of other ayacut farmers by playing a version of the trust game.

The following longer run outcomes will be measured at endline: (a) the condition and maintenance of the channel, (b) quality of water flow at various locations, (c) changes to water sharing rules/allocations, (d) cultivated area, and (e) agricultural outcomes including yield and total sales and revenue.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Based on our extensive fieldwork and initial piloting in 20 villages, we propose an RCT where we take the costs of participation seriously. In particular, we vary (a) whether farmers in the ayacut choose the location of the channel and, once its location has been decided, (b) the extent of local control over the construction of the channel.

The experimental protocol involves the following steps. First, we obtain necessary permissions for the study from the village representative, the village official in-charge of revenue and water management (Village Revenue Assistant or the VRA), and other key members of the farmer community, who are typically involved in village irrigation management. In this stage, we sketch out a schematic map of the tank and ayacut and identify plausible locations for different field channels connecting the mouth of the tank to different locations downstream in the ayacut.

Next, we conduct a detailed baseline survey of farms/plots within the ayacut by interviewing the landowner. This survey includes basic demographic background details of the farmer, plot level characteristics such as soil type, whether there is a functioning well, main irrigation sources, quantity of water available for irrigation by season, value per acre, and agricultural investment and yield. We also ask about water conflicts with other farmers in the ayacut and elicit the farmers’ valuation of the different locations for the construction of a field channel identified in Step 1, using an incentive compatible elicitation method.

In the third step, we announce whether the choice of field channel to be constructed is “bottom-up”, that is, made by every farmer with a plot in the ayacut via a secret ballot process (one farmer one vote) or whether it is made “top-down” by a third party (researchers/NGO/centralized agency) without any community involvement. During the pilot the elicitation method worked because the field channel selected using the secret ballot always coincided with the highest valuation channel. As a result, in top-down villages we also plan to select the channel with the highest valuation in the baseline elicitation exercise to make the location of field channels comparable in both groups of villages.

In the fourth and final step, we vary whether the field channel selected is constructed by the community (“bottom-up” approach) or by a local contractor or NGO directly hired by us (“top-down” approach). In the bottom up approach, the funds are given to the VRA and since farmers have full control of the construction, they may decide to extend the length of the channel, add additional feeder channels, or make other modifications as they deem fit. In contrast, funds are given directly to the local contractor or NGO under the top-down approach.
Experimental Design Details
Not available
Randomization Method
The village/ayacut is our unit of randomization and will be stratified at the level of river basin (hydrological strata).
Randomization Unit
All treatments are randomized at the tank-ayacut level and since we only include one tank in the study per village, treatments are also randomized at the village level.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
500 ayacuts/villages
Sample size: planned number of observations
Our sample size includes 500 villages/ayacuts, covering approximately 30 farmers per village/ayacut, so we will survey a total of 15,000 farmers.
Sample size (or number of clusters) by treatment arms
Our sample size includes 100 villages/ayacuts per treatment arm. The breakdown is:
1. Top-down choice: 100 villages/ayacuts
2. Bottom-up choice (secret ballot): 100 villages/ayacuts
3. Top-down implementation: 100 villages/ayacuts
4. Bottom-up implementation: 100 villages/ayacuts
5. Control group: 100 villages/ayacuts
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
University of Chicago
IRB Approval Date
IRB Approval Number