Weather Index Insurance and Women's Decision to Insure in Ghana: An Experimental Game Approach

Last registered on October 31, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

Weather Index Insurance and Women's Decision to Insure in Ghana: An Experimental Game Approach

RCT ID

AEARCTR-0017097

Initial registration date

October 28, 2025

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

October 31, 2025, 8:29 AM EDT

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

Locations

Country

Ghana

Region

Northern Ghana

Primary Investigator

Name

Isaac Doku

Affiliation

University of Education, Winnebah-Ghana

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Benedicta Akrono

PI Affiliation

University of Education, Winneba-Ghana

Contact Investigator

Additional Trial Information

Status

In development

Start date

2025-11-15

End date

2026-02-28

Keywords

Agriculture, Behavior, Environment & Energy, Finance & Microfinance, Gender

Additional Keywords

JEL code(s)

Q14, Q54

Secondary IDs

SGS-EC-72-25

Prior work

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

Abstract

Climate change is adversely impacting the life of all humans on earth today, both men and women. However, women are disproportionately affected as compared to men, especially women farmers in Sub-Saharan Africa who are thought to be more vulnerable to climate change and its impacts. To deal with the exacerbating effect of climate change on farmers led to the introduction of weather index insurance in Ghana by the Ghana Agriculture Insurance Pool (GASIP). Sadly, most farmers in Ghana do not buy index insurance due to discouragement because they hardly qualify for indemnity payment. This characterizes index insurance in developing countries with moral hazard and adverse selection problems. As a result, GASIP has tried refining index insurance by bundling it with agricultural inputs. We surmised that lack of information about index insurance, and index insurance products are targeted to men maybe reasons for low demand. We want to use an experimental game to sensitize householders to find out whether it will increase demand for index insurance. Furthermore, we would find out women demand for insurance when index insurance is bundled with seedlings against bundled with fertilizer using the experimental game. This study would use a tablet-based experimental game to simulate the life of a smallholder farmer, allowing participants to make real world decisions about investment and insurance with probabilities and payoffs. Data for this experiment will be collected by randomly selecting 30 communities by 10 households per community, and each assigned to 1 of two treatment groups or a control group. In this study, 100 couples will be expected to play the game of weather index insurance bundled with fertilizer (T1). Further, another 100 couples will be expected to play the game of weather index insurance bundled with seedlings (T2). Those in the control group (100 couples) will be introduced to riskiness in farming, but not be sensitized like those in the treatment group and be presented with the status quo-only weather index insurance. Data collection instruments will be embedded in the game, so as individuals and couples play the game and make decisions their information will be gathered.

External Link(s)

Funders description of the project

Registration Citation

Citation

Akrono, Benedicta and Isaac Doku. 2025. "Weather Index Insurance and Women's Decision to Insure in Ghana: An Experimental Game Approach." AEA RCT Registry. October 31. https://doi.org/10.1257/rct.17097-1.0

Sponsors & Partners

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

Request Information

Experimental Details

Interventions

Intervention(s)

Demand for index insurance among farmers is very low, as a result most studies focus on subsidized index insurance and male farmers in patriarchal societies. This study seeks to find out whether bundling index insurance with seedlings or fertilizer will increase demand, especially for female members in a smallholder farming household. This will be done by employing experimental games following the study by Hobbs (2022). This study will use a tablet-based experimental game to simulate the life of a smallholder farmer, allowing participants to make real world decisions about investment and insurance with probabilities and payoffs. Data for this experiment will be collected in 30 sessions among 300 couples.

Intervention Start Date

2025-11-30

Intervention End Date

2026-01-31

Primary Outcomes

Primary Outcomes (end points)

1. Women’s insurance demand
2. Women’s insurance demand when decision made jointly (couple stage)

Primary Outcomes (explanation)

1. Women’s insurance demand: For each female player
ℎ in round 𝑚, the proportion of her game-budget spent on index insurance.

2. Women’s insurance demand when decision made jointly (couple stage): Same proportion measure but from the couple-play stage (the share of household/couple-budget spent on insurance when the couple makes the decision together)

Secondary Outcomes

Secondary Outcomes (end points)

1. Comparison between bundling types (seedling vs fertilizer)
2. Men’s insurance demand (individual-play)
3. Household purchase/uptake (binary)
4. Allocation trade-offs across budget categories
5. Knowledge/learning score
6. Women empowerment heterogeneity

Secondary Outcomes (explanation)

1.Comparison between bundling types (seedling vs fertilizer): Difference in women’s
insurance demand between seedling-bundled (S=1) and fertilizer-bundled (S=0) arms
2. Men’s insurance demand (individual-play): Measured analogously to women’s IN; estimated via OLS specification. Use this to compute gender gap in demand and test whether treatments close the gap.
3. Household purchase/uptake (binary); Binary indicator equal to 1 if the household chooses to purchase any insurance in a given round or overall.
4. Allocation trade-offs across budget categories: Proportion of budget allocated to education; to other investments; to inputs (seedlings/fertilizer).
5. Knowledge/learning score: Sum of correct answers to 5 true/false knowledge questions (range 0–5). Also analyze each question separately.
6. Women empowerment heterogeneity: Empowerment index from baseline decision-making Likert scale (1–5 per item; compute average)

Experimental Design

Participants are randomly assigned to one of three experimental arms. Each household consists of a couple (husband and wife) who participate first individually and then jointly.
Group Treatment Description Sample (couples) Expected Mechanism
T1 – Seedling Bundle Weather index insurance bundled with climate-resilient seedlings 100 Bundling with seedlings increases salience and female engagement
T2 – Fertilizer Bundle Weather index insurance bundled with fertilizer 100 Bundling with fertilizer highlights productivity potential, neutral gender appeal
Control Weather index insurance only (no bundling), minimal sensitization 100 Baseline comparison for insurance uptake and learning
Total = 300 couples (600 individuals), across 30 communities.
Each game session lasts approximately two hours, divided into 8 rounds (years). In each round, participants allocate a virtual household budget across:
1. Insurance,
2. Education,
3. Agricultural investment, and
4. Consumption.
Random weather shocks occur with probabilities reflecting real agroclimatic risks (sourced from GAIP data). Income and insurance payouts are computed automatically, mirroring realistic farm outcomes.

Experimental Design Details

Not available

Randomization Method

• Unit of Randomization: Household (couple).
• Stratification/Blocking: Communities (30) serve as randomization blocks to ensure balanced representation across regions and to minimize information spillover.
• Randomization Method: Within each community block, random assignment of households to one of the three treatment arms (1:1:1 ratio) using computer-generated random numbers in Stata (set seed, sample).
• Randomization Timing: Conducted before workshops, with sealed assignment lists shared only with enumerators on the day of the experiment.
Thus, randomization occurs at the household level within community blocks, yielding balanced treatment groups.

Randomization Unit

Household and or couple level

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

300 couples

Sample size: planned number of observations

300 couples

Sample size (or number of clusters) by treatment arms

100 couples for Insurance bundled with seedlings, 100 couples for insurance bundled with fertilizer and 100 couples for control arm.

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

With 100 observations per group and the hypothesized proportions of 0.1 0.5 0.6, I have approximately 77.55% power to detect a statistically significant difference between these groups at the 0.05 significance level

Supporting Documents and Materials

IRB

Institutional Review Boards (IRBs)

IRB Name

Institutional Ethical Review Board, University of Education, Winneba

IRB Approval Date

2025-09-24

IRB Approval Number

SGS-EC-72-25

Analysis Plan

Analysis Plan Documents

Experimental Game

MD5: d3c76bc58460b449490459a3e7372c74

SHA1: 0027589b4ae399dbcbd4eb00921336539727d62f

Uploaded At: October 23, 2025