Digital Innovations in Crop Insurance Product Design

How to cite

Porter, Maria; Kramer, Berber; Abate, Gashaw T.; Wassie, Solomon. 2022. "Digital Innovations in Crop Insurance Product Design." AEA RCT Registry. AEARCTR-0010694. https://doi.org/10.1257/rct.10694-2.1

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

Maria Porter* , Berber Kramer , Gashaw T. Abate , Solomon Wassie

* Primary investigator

Registry ID
AEARCTR-0010694
DOI
https://doi.org/10.1257/rct.10694-2.1
First registered
Last updated
Status
Completed
Trial dates
2022-08-15 – 2024-12-31
Countries
Ethiopia (Amhara)
JEL codes
Q12, Q14, G52, D13
Registry record
http://www.socialscienceregistry.org/trials/10694

Abstract

Index-based insurance schemes have been implemented to protect smallholder farmers against agricultural losses, but basis risk and delays in insurance payouts suppress demand for such products, fostering mistrust in insurance when insured farmers experience crop losses and do not receive a (timely) payout.

The proposed project will partner with the World Food Programme’s R4 Rural Resilience Initiative in Ethiopia and researchers from the Ethiopian Institute for Agricultural Research, Bahir Dar University, the University of Arizona, the International Food Policy Research Institute, and the University of Manchester.

The project mainly aims to lower basis risk and premiums of R4 products and improve timing of payouts. While (drought tolerant) improved seeds are promoted through trial packs to reduce production risks, picture-based insurance (PBI) is introduced to reduce basis risk. Mobile banking will be adopted to facilitate more timely payouts to farmers, so that farmers can use insurance payouts to mitigate further crop losses. Biophysical crop simulations of the impacts that drought-tolerant varieties promoted through the project can have on agricultural risk management will help optimize index design and lower premiums for farmers adopting risk-reducing technologies and practices.

We will evaluate the impact of these innovations on insurance demand, productivity, profitability, risk mitigation, consumption smoothing and resilience, by randomizing the types of products/interventions being offered across villages.

Intervention

The experiment consists of the following four key interventions (including the fundamental intervention provided to all study participants).

(1) Index-based insurance. The provision of index-based insurance is the underlying intervention provided by R4 Rural Resilience Initiative (R4) and it estimates crop losses using rainfall and vegetation indices measured through satellite remote sensing.

(2) Picture Based Insurance (PBI). The PBI is a complementary (or adaptive) insurance intervention that aims to overcome information asymmetry and trust issues on insurance. With PBI, images of crop damage will be used as an instrument to lower basis risk and provide fail-safe insurance contracts: when the index-based insurance does not trigger payouts, but crop images provide visible evidence of damage from perils covered by the policy, then these images will be used to settle claims.

(3) Seed promotion. The promotion of recently released (drought tolerant) improved varieties through trial packs is the third key intervention, which aims at improving varietal turnovers and thereby reducing production risks.

(4) Mobile banking. The mobile banking intervention involves encouraging insurance enrollees to sign-up for digital payment and it aims to speed up delivery of insurance payouts and ensure recipients control over insurance payouts.

The project will also quantify the effects of adopting risk-reducing practices (e.g., adoption of drought tolerant seed) on risk exposure through biophysical crop simulations, to inform the R4 program of how it may adjust its triggers and premiums for farmers adopting improved risk management. Smartphone images from previous seasons will be included in the portal to provide index designers with information on the crop growth stage at a given point in time. This information helps determine the periods during which the crop is vulnerable and customize, for instance, the starting and ending date of a weather index designed for the germination or flowering phase. Moreover, the images and other data collected at the time of sending in the images will provide information on watershed management and other risk management practices in a community. When combined with the results from the crop simulations, this information can help index designers adjust coverage windows, insurance triggers, and other index parameters, to adequately reflect farmers’ management practices and allow those investing in risk reduction to benefit from lower insurance premiums.

Experimental design

The experiment employs a three-arm stratified cluster randomized design. Stratification occurs at the level of the district/woreda, and clusters are defined at the kebele level, which is the lowest administrative unit in Ethiopia. Within each district, kebeles were randomly allocated to one of the following three groups.

(T0) – R4 only group (control group) in which the standard index-based insurance is offered (40 kebeles);
(T1) – R4 + seed promotion group in which trial packs of recently released (drought tolerant seeds) are provided in addition to the index-based insurance (40 kebeles); and
(T2) – R4 + seed promotion + PBI group in which the Picture Based Insurance is provided as a complementary or adaptive insurance product (40 kebeles).

Finally, within each of these 120 villages, 8 households will be randomly assigned to receive mobile banking as an option for receiving insurance payouts, while another 8 households will be randomly assigned to our study but will only be offered R4's standard insurance payment methods.

Design details

Randomization method
randomization done in office by a computer
Randomization unit
Kebele/village level randomization for assignment to one of three treatments: (1) R4 insurance product, no adapted insurance index, no seed promotion; (2) R4 insurance product, no adapted insurance index, seed promotion; (3) R4 insurance product, adapted insurance index, seed promotion. Household level randomization to one of two treatments: (1) No offer of mobile banking for insurance payouts; (2) Households encouraged to use mobile banking for receiving insurance payouts.
Clusters
120 kebeles/villages
Observations
1,920 households
Arms
40 villages: R4 insurance, no adapted index, no seed promotion, 40 villages: R4 insurance, no adapted index, but with seed promotion, 40 villages: R4 insurance, adapted index, and seed promotion. In each village: 8 households per village without mobile banking encouragement, 8 households per village with mobile banking encouragement
Primary outcomes
Household-level outcomes: insurance take-up rates, whether farmer grows an improved variety, area cultivated with improved varieties, amount spent on seeds, farmer-estimated crop damage Among farmers reporting crop damage: receipt of insurance payouts (i.e., downside basis risk), period of time between damage and receipt of insurance payment, self-reported satisfaction with insurance product.
Secondary outcomes
Household-level outcomes: total area cultivated, adoption of improved technologies and practices, total cost of production (both excl. and incl. family labor), yield per hectare cultivated, agricultural net income per hectare, financial inclusion index, food consumption score, women's bargaining power.

Cited by

No citation data available for this registration yet.

Keywords