Forecasting the rain: Digital dissemination of rainfall quantity forecasts

How to cite

Burlig, Fiona; Jina, Amir; Kelley, Erin; Lane, Gregory. 2024. "Forecasting the rain: Digital dissemination of rainfall quantity forecasts." AEA RCT Registry. AEARCTR-0013642. https://doi.org/10.1257/rct.13642-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Fiona Burlig* , Amir Jina , Erin Kelley , Gregory Lane

* Primary investigator

Registry ID
AEARCTR-0013642
DOI
https://doi.org/10.1257/rct.13642-2.0
First registered
Last updated
Status
In development
Trial dates
2024-05-18 – 2025-09-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/13642

Abstract

Small-holder farmers in low-income countries face substantial weather risk. To reduce these risks, governments provide long-range forecasts about the amount of rain that will occur during the agricultural season, which can be disseminated in advance of the start of the season. In this project, we partner with the Indian Ministry of Agriculture and Farmers' Welfare (MoA&FW) to randomize SMS-based dissemination of the Indian Meterological Department's India-wide total monsoonal rainfall forecast among 128,000 farmers living in 22 subdistricts where this forecast has been shown to be particularly effective. We will measure the effects of these forecasts on farmer beliefs and agricultural activity.

Intervention

Farmers randomized into treatment receive an SMS with information about the all-India quantity of monsoon rain expected to arrive in summer 2024.

Experimental design

In 22 Indian subdistricts, we will randomize 25% of farmers enrolled on the PM-Kisan platform to receive forecast SMS messages (treatment) and 75% of PM-Kisan farmers not to receive these messages (control), with approximately 97,000 farmers in the control group and 32,000 farmers in the treatment group across all subdistricts. The randomization is performed at the individual level, and is stratified by subdistrict.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Individual
Clusters
128,000
Observations
2,000 surveys
Arms
1,000 surveys control; 1,000 surveys treatment
Primary outcomes
Beliefs about the monsoon, agricultural outcomes

Cited by

No citation data available for this registration yet.

Keywords