Probabilistic Weather Forecasts and Agricultural Decision Making in Rural India

Last registered on August 10, 2023

View Trial History

Pre-Trial

Trial Information

General Information

Title

Probabilistic Weather Forecasts and Agricultural Decision Making in Rural India

RCT ID

AEARCTR-0011526

Initial registration date

June 09, 2023

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

June 15, 2023, 4:43 PM EDT

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

Last updated

August 10, 2023, 6:29 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

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

Request Information

Primary Investigator

Name

Vaishnavi Surendra

Affiliation

Precision Development

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Shawn Cole

PI Affiliation

Harvard University

Contact Investigator

PI Name

Tomoko Harigaya

PI Affiliation

Precision Development

Contact Investigator

Additional Trial Information

Status

On going

Start date

2023-08-05

End date

2024-12-31

Keywords

Agriculture, Behavior, Environment & Energy, Lab

Additional Keywords

Risk, Forecasts, Climate Change

JEL code(s)

C91, D81, O12, O13, Q54

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Weather uncertainty is a significant source of agricultural production risks, and is increasingly salient in the context of climate change. Short- to medium-range weather forecasts could help farmers optimize the timing of agricultural practices, increasing returns to agricultural investment. However, smallholder farmers in most parts of the developing world do not have access to highly localized, accurate forecasts, nor are forecasts customized to convey weather conditions relevant for agricultural decision-making. In this study, we investigate whether accurate, relevant-for-context, probabilistic, medium-range weather forecasts can aid farmer decision-making through a lab-in-the-field experiment. We will analyze whether climate change salience, probability training, prior luck impact choices in the experiment and willingness-to-pay for weather forecasts.

External Link(s)

Registration Citation

Citation

Cole, Shawn, Tomoko Harigaya and Vaishnavi Surendra. 2023. "Probabilistic Weather Forecasts and Agricultural Decision Making in Rural India." AEA RCT Registry. August 10. https://doi.org/10.1257/rct.11526-2.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Farmers are randomly assigned to one of the following groups: (1) climate change salience information treatment; (2) probability training and climate change salience information treatments; (3) control.

1. The climate change salience information treatment involves farmers watching a video describing changing weather patterns globally and locally, and the impact it has on coffee farming.
2. The probability training information treatment involves farmers watching a probability training video.
3. In round 1 of the experiment, farmers will receive a few hypothetical scenarios that involve choosing a location with specific weather conditions out of two locations, for which they see weather forecasts. Once a location is chosen, farmers will indicate their confidence in their choice ('points-at-stake'), and then in-experiment-round weather will be realized. Farmers’ payoffs will depend on the realized weather.
4. In round 2 of the experiment, farmers will receive a few hypothetical cultivation practice-related scenarios that involve choosing an action following a weather forecast. Once an action is taken, in-experiment-round weather will be realized. Farmers’ payoffs will depend on the realized weather.
5. Finally, we will elicit farmers’ willingness-to-pay for accurate, customized, probabilistic weather forecasts.

Intervention Start Date

2023-08-05

Intervention End Date

2023-12-31

Primary Outcomes

Primary Outcomes (end points)

Willingness to pay for probabilistic medium-range weather forecasts; choosing the correct forecast; choosing the action with the higher expected payoff.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

We will recruit small- and medium-holder coffee farmers in Karnataka, India. Farmers will be randomly assigned to (1) climate change salience information treatment, (2) probability training and climate change salience information treatment, (3) control.

Each farmer will answer survey questions, watch their assigned video(s), and play a hypothetical decision-making game where decisions depend on weather realizations and decisions are made after receiving a weather forecast. The game consists of two rounds. In the first round, farmers will receive hypothetical scenarios in which they choose between two locations with different weather forecasts, picking the location where a weather event is more (less) likely to occur. After their choice, a weather realization will occur in-game. Farmer payoffs depend on realized weather. In the second round, farmers will receive hypothetical agricultural scenarios, and choose whether to take an action or not, following a weather forecast. After their choice of action, a weather realization will occur in-game. Farmer payoffs depend on realized weather.

After the end of the game, we will elicit farmers’ willingness-to-pay for accurate, medium-range probabilistic weather forecasts.

Experimental Design Details

Not available

Randomization Method

Randomization is conducted using a random number generator in Stata, and within the survey software.

Randomization Unit

Individual-level, and within individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

At least 1200 farmers (actual number depends on number of participants enrolled).

Sample size: planned number of observations

At least1200 individuals (actual number depends on number of participants enrolled).

Sample size (or number of clusters) by treatment arms

504 farmers receiving the climate change salience information treatment; 348 farmers receiving the climate change salience and probability training information treatment; 348 farmers in the control group (actual numbers depends on number of participants enrolled).

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials

IRB