Insuring Against Extreme Heat: Experiential Learning and the Demand for Heat Insurance

How to cite

Zou, Weijian; Simpson, Pol; Jalal, Amen; Pople, Ashley. 2026. "Insuring Against Extreme Heat: Experiential Learning and the Demand for Heat Insurance." AEA RCT Registry. AEARCTR-0017810. https://doi.org/10.1257/rct.17810-1.0

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

Weijian Zou* , Pol Simpson , Amen Jalal , Ashley Pople

* Primary investigator

Registry ID
AEARCTR-0017810
DOI
https://doi.org/10.1257/rct.17810-1.0
First registered
Last updated
Status
In development
Trial dates
2026-02-02 – 2026-10-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/17810

Abstract

This randomised study evaluates demand for parametric heat insurance among informal women workers in India. The insurance product, designed by the Self Employed Women's Association (SEWA), provides automatic payments to informal workers when district-level temperatures exceed 43°C for two consecutive days.
We test whether experiential learning (via an "insurance game") increases willingness to pay for this novel product. Participants are randomly assigned at the individual level to either a control group, who receive standardised verbal information about the product, or a treatment group that additionally plays an interactive game simulating two years of weather during the month of May (typically the hottest month of the year). In the game, participants draw random weather cards and work with the enumerator to calculate payouts under these randomly drawn scenarios. We elicit willingness to pay using an open-ended question followed by a bidding ladder. We consider the measure incentive compatible because respondents are told they will be contacted by SEWA administration to collect the amount they are willing to pay. In addition to collecting willingness-to-pay in the lab during February 2026, we will also observe real sign-ups in SEWA's administrative data during the 2026 hot season.
We also seek to understand which factors most influence demand. To test the role of product characteristics, we implement a discrete choice module where participants choose between alternative insurance designs varying in price and product attributes. To examine other demand factors - risk tolerance, climate beliefs, financial experience and heat exposure - and estimate heterogeneous treatment effects, we collect extensive baseline data.

Intervention

Individual-level randomised experiment with approximately 1,000 SEWA members in Anand district, Gujarat, stratified by prior heat insurance purchase.
Participants are randomly assigned to either: (1) Control group: receives standardised verbal information about SEWA's heat insurance product; or (2) Treatment group: receives the same information plus plays an interactive insurance game simulating two years of weather outcomes.
All participants complete willingness-to-pay elicitation and a discrete choice module. We also observe real-world sign-ups in SEWA's administrative data.

Experimental design

Individual-level randomised experiment with approximately 1,000 SEWA members in Anand district, Gujarat, stratified by prior heat insurance purchase.
Participants are randomly assigned to either: (1) Control group: receives standardised verbal information about SEWA's heat insurance product; or (2) Treatment group: receives the same information plus plays an interactive insurance game simulating two years of weather outcomes.
All participants complete willingness-to-pay elicitation and a discrete choice module. We also observe real-world sign-ups in SEWA's administrative data.

Design details

Randomization method
Randomisation of individuals to treatment and control group done on Survey CTO
Randomization unit
Individual
Clusters
N/A
Observations
1,000 SEWA members (survey sample) + all available administrative records.
Arms
500 individuals control, 500 individuals treatment
Primary outcomes
Our primary outcomes are: Real-world sign-up: Binary indicator for whether the SEWA member signed up for heat insurance in 2026. Stated maximum willingness to pay, elicited via open-ended question and bidding ladder, with participants given the opportunity to revise after being reminded of the stakes, i.e., that their response will be passed to SEWA's grassroots leader who will contact them to collect payment.
Secondary outcomes
1. Family 1: a. Initial stated willingness to pay (INR), before opportunity to revise b. Payout beliefs (i). Expected total payout if purchases insurance (ii). Subjective probability of receiving a payout c. Heat insurance product comprehension (number of correct responses to comprehension check questions) d. Concern about heat impacts on earnings this year (0-10) 2. Marginal willingness to pay for product attributes (from discrete choice experiment) We will construct an Anderson (2008) index of outcomes in Family 1. Additionally, following Benjamini, Krieger and Yekutieli (2006), we will use false discovery rate corrections to account for multiple hypothesis testing within Family 1.

Cited by

No citation data available for this registration yet.

Keywords