Bundling crop insurance and extension services for Maize farmers in Mali

How to cite

Ölkers, Tim; Mußhoff, Oliver. 2023. "Bundling crop insurance and extension services for Maize farmers in Mali." AEA RCT Registry. AEARCTR-0011980. https://doi.org/10.1257/rct.11980-1.1

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

Tim Ölkers* , Oliver Mußhoff

* Primary investigator

Registry ID
AEARCTR-0011980
DOI
https://doi.org/10.1257/rct.11980-1.1
First registered
Last updated
Status
In development
Trial dates
2023-11-12 – 2025-08-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/11980

Abstract

Smallholder farmers comprise a large group of Mali’s inhabitants and also economically play an important role. However, despite the great amount of natural resources, experts all over the world coincide in the perception that a large portion of the countries’ agricultural potential remains untapped. Mali is severely affected by the climate crisis. Therefore, it is of importance and of interest to understand and develop tools that may reduce the negative effects related to the climate crisis for the region in Western Africa.
For the present study we, the Chair of Farm Management, cooperate with OKO Mali SarL, in the following referred to as OKO and WeatherImpact. OKO distributes weather index-based insurance policies for cereal farmers against drought or flooding. Their insurance products are tailored towards the needs of specific crops. We focus on maize insurance. The insurance pays out based on an index that is mainly linked to precipitation levels. It further considers different stages of the crop cycle by defining different strike levels for the germination period, growth period and maturation phase. If the accumulated precipitation in one phase is lower than the strike level, farmers are likely to suffer yield losses. The exit level is the accumulated precipitation level below which the indemnity is exactly the insured value. WeatherImpact provides the weather forecast which are send via SMS to the participants of our study.
While two groups receive the insurance, only one group of respondents received the insurance with additional weather information and agronomic advice. According to a predefined schedule, farmers in the treatment group receive two to three text messages per week. The text messages either contain weather information or agronomic advice. There are 23 different messages with basic agronomic advice including five messages with information on pre-planting activities and four on post-harvest management. The message also suggests to contact the local extension officer for more detailed information if the text message contained specific key words. The agronomic advice is based on practices commonly known from conservation agriculture. We plan a Randomized Control Trial (RCT). Hence, all participants have to be interviewed twice. One time in November 2023 and a second time one year later (November 2024) to estimate the effect of the bundled product. Additional we sample one control group, consisting out of Malian maize Farmers who did no buy the insurance product. They survey will target 1000 to 1200 customers.

We are interested in the following research questions:
a) What is the additional utility of the bundled product?
b) What is the value of the weather forecast?
c) Is the agronomic advice improving the maize outcomes and farmer’s wellbeing?
d) What is the farmers' willingness to pay for the bundled product?
e) How satisfied are the farmers with the bundled product?

Intervention

We have three treatment arms (one control and two interventions), each consisting of approximately 500 farmers, resulting in a total sample size for the baseline of approx. 1500.
While two groups receive the insurance (approx. 1000), only one group of respondents received the insurance with additional weather information and extension services via SMS (approx. 500). According to a predefined schedule, farmers in the treatment group receive two to three text messages per week. The text messages either contain weather information or agronomic advice.

Experimental design

We have three treatment arms (one control and two interventions), each consisting of approximately 500 farmers, resulting in a total sample size for the baseline of approx. 1500.
While two groups receive the insurance (approx. 1000), only one group of respondents received the insurance with additional weather information and extension services via SMS (approx. 500). According to a predefined schedule, farmers in the treatment group receive two to three text messages per week. The text messages either contain weather information or agronomic advice.

The assignment of who will receive the bundled product (index-insurance PLUS farming advise and weather information via SMS) will be based on stratification, taking into account factors such as agricultural zone, land ownership, village characteristics, and baseline output.

Design details

Randomization method
The insurance broker organizes village events to introduce and offer the insurance to maize farmers. Additionally, the maize cooperative and village leaders extend invitations to the broker for introducing the insurance to their members. These efforts create a platform through which farmers can opt to insure their plots. We will interview the farmers if they ask for a quote for insurance. The assignment of who will receive the bundled product (index-insurance PLUS farming advise and weather information via SMS) will be based on stratification, taking into account factors such as agricultural zone, land ownership, village characteristics, and baseline output. The stratification will be conducted by the University of Göttingen after the baseline collection in January/February 2024.
Randomization unit
The insurance broker organizes village events to introduce and offer the insurance to maize farmers. Additionally, the maize cooperative and village leaders extend invitations to the broker for introducing the insurance to their members. The assignment of who will receive the bundled product (index-insurance PLUS farming advise and weather information via SMS) will be based on stratification, taking into account factors such as agricultural zone, land ownership, village characteristics, and baseline output. We will randomly assign the additional treatment on village level based on stratifcation.
Clusters
The total sample size for the baseline is approx. 1500. We except to interview 20 farmers per village, resulting in 75 cluster.
Observations
We have three treatment arms (one control and two interventions), each consisting of approximately 500 farmers, resulting in a total sample size for the baseline of approx. 1500.
Arms
500 farmers control, 500 treatment insurance, 500 treatment bundled product.
Primary outcomes
We are interested in the following research questions: a) What is the additional utility of the bundled product? b) What is the value of the weather forecast? c) Is the agronomic advice improving the maize outcomes and farmer’s wellbeing? d) What is the farmers' willingness to pay for the bundled product? e) How satisfied are the farmers with the bundled product? (see more information in the Pre-Analysis Plan)
Secondary outcomes
(see more information in the Pre-Analysis Plan)

Cited by

No citation data available for this registration yet.

Keywords