India Grain Legume Cluster Development: Experiment on Salience of Evaluation

Last registered on July 02, 2019

Pre-Trial

Trial Information

General Information

Title
India Grain Legume Cluster Development: Experiment on Salience of Evaluation
RCT ID
AEARCTR-0004405
Initial registration date
July 01, 2019

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
July 02, 2019, 5:04 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
UC Davis

Other Primary Investigator(s)

PI Affiliation
University of California, Davis
PI Affiliation
University of California, Davis
PI Affiliation
University of California, Davis

Additional Trial Information

Status
On going
Start date
2017-04-24
End date
2021-06-30
Secondary IDs
Abstract
This study follows a randomized evaluation of a program to promote the production of pulses in Bihar, India. Farmers in treatment villages were given the opportunity to buy pulse seeds and other inputs at discounted rates, coupled with training and extension services to promote better cropping practices. All implementation was conducted by local development NGOs that have been operating in the area for several years. A key measure of the sustainability of the program is the demand for purchasing and planting pulse seeds once subsidies are no longer in place. We elicit this demand through conducting seed auctions in all treatment and control villages. During demand elicitation, we randomize the salience of the evaluation itself by emphasizing that we are collecting data to evaluate the local partner's performance in promoting pulse production in half the villages. By using this two-level randomization, we seek to assess how the salience of an evaluation affects the measured treatment effects of a development program.
External Link(s)

Registration Citation

Citation
Bourdier, Tomoé et al. 2019. "India Grain Legume Cluster Development: Experiment on Salience of Evaluation." AEA RCT Registry. July 02. https://doi.org/10.1257/rct.4405-1.0
Former Citation
Bourdier, Tomoé et al. 2019. "India Grain Legume Cluster Development: Experiment on Salience of Evaluation." AEA RCT Registry. July 02. https://www.socialscienceregistry.org/trials/4405/history/49178
Experimental Details

Interventions

Intervention(s)
The initial evaluation, starting in 2017, comprises 228 villages over five districts that were randomly assigned to either treatment or control status. Over the following two years, farmers in treatment villages were given the opportunity to buy pulse seeds and other inputs at discounted rates, coupled with training and extension services to promote better cropping practices. All implementation was conducted by local development NGOs that have been operating in the area for several years. In June 2019, we completed the training and subsidies and farmers will have the opportunity to continue cultivating pulses by paying the market rate for inputs.

A key measure of the sustainability of the program is the demand for purchasing and planting pulse seeds once subsidies are no longer in place. We elicit this demand through seed auctions in all treatment and control villages. During demand elicitation, we randomize the salience of the evaluation itself. In about half of the villages, designated high-salience, we explicitly emphasize that we are collecting data to evaluate the local partner's efforts to promote pulse production; in the other half, we say we are collecting data on the desirability of pulse cropping in the region. Since the project is ongoing, with further interventions related to output and marketing to come, it is plausible in the low-salience group that our activities are not directly tied to evaluation.

By using this two-level randomization, we seek to assess how the salience of an evaluation affects the measured treatment effects of a development program. Participants in a development program may be interested in seeing it evaluated successfully, either to promote more such programs in the future, as a form of gift-exchange in response to development aid, or for other potential reasons. We will investigate whether this interest translates into measurable outcomes by comparing the measured treatment effect of the main input intervention on future input demand across the high- and low-salience groups.
Intervention Start Date
2019-06-06
Intervention End Date
2019-06-18

Primary Outcomes

Primary Outcomes (end points)
Seed input demand
Primary Outcomes (explanation)
We will convert farmer responses in the auction into a series of implied reservation prices for different quantities of distinct varieties of black gram and pigeon pea seeds. We will construct a simple weighted average of the reservation prices for (a) each farmer's composite demand
for each type of seed and (b) each farmer's composite demand for all of the pulse seeds on offer at the auction in their village.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We randomly assign approximately half of the 228 villages to the high-salience group and half to the low-salience group, stratified by initial treatment status. We will analyze whether the salience of the evaluation increases demand for pulse seeds. Data on input demand comes from an incentive-compatible elicitation of willingness to purchase pulse seeds at various prices. In both treatment and control villages, we invite 6-10 farmers to an auction where farmers are given a list of five different prices in Rs. 20 increments from Rs. 60 to Rs. 140. For each price, farmers name their quantity demanded of seed. At the end of the auction, one of those prices is randomly drawn and farmers are given a coupon to purchase their desired amount at the selected price. The coupon must be redeemed in full; farmers cannot name a high quantity demanded and then purchase a smaller amount. The coupons and distributor are uniform across both treatment status as well as salience status to minimize differences caused by perception of the quality of product being sold.

Farmers are invited to the auction in two ways. First, a random sample of roughly seven farmers per village, which we selected previously and have surveyed in the past, are directly invited to participate, with scheduling set to accommodate as many of them as possible. Then, to fill as many remaining spots as possible up to ten, villagers are asked to invite others who may be interested in purchasing pulse seeds. Farmer selection is identical across salience status, but may not be identical across treatment status as past extension services may alter who in the village is interested in purchasing seeds.
Experimental Design Details
Randomization Method
Randomization done in office by a computer
Randomization Unit
Village
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
228 villages
Sample size: planned number of observations
1,824 seed auction participants
Sample size (or number of clusters) by treatment arms
106 villages low-salience, 122 villages high-salience
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
UC Davis IRB
IRB Approval Date
2017-11-02
IRB Approval Number
1142147
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials