Enhancing access to index-based weather agricultural insurance: a new marketing approach in Burkina Faso

Large segments of the population in developing countries, especially in rural areas, have a high level of vulnerability to weather-related shocks but have limited means to insure themselves against them. In recent years, microfinance institutions have experimented with insurance products, in particular rainfall index insurance, to address this need in different parts of the world. But the uptake of these products has generally been very low because of liquidity constraints, lack of trust in insurance provider and unfamiliarity with formal financial products. We study how existing ties between urban migrants and rural farmers can be used to provide the latter improved access to formal insurance. The study is motivated by well-established evidence regarding the use of rural-urban migration as a risk-coping and risk-management strategy and that rural households in developing countries often rely upon assistance from close relatives among urban migrants to cope with adverse weather-related shocks.

We randomize at the village level in the rural area, and at the individual level in the city. In villages randomized to the treatment group, the insurance provider will offer its standard product, i.e. market rainfall index insurance to rural farmers. Migrants from both treatment and control villages who live in Ouagadougou will be randomly assigned to two groups. The first migrant group will be offered to purchase the insurance product to cover their relatives in the villages, while the insurance product will not be marketed to the second migrant group (the control). To simplify the exposition, we denote by V0 and V1 the control and the treated villages, and by U0 and U1, the control and treated urban migrants. Combining rural households and urban migrants produces four experimental groups that can be described as follows.
1. V0U0: The pure control group: the insurance product is not marketed to rural households or to their migrant relatives.
2. V0U1: The insurance is marketed to the urban migrants, but not to their rural relatives.
3. V1U0: The insurance is marketed to rural households, but not to their urban relatives.
4. V1U1: The insurance is marketed both to rural households and to their urban relatives.

2019-06-03

2022-03-31

 Take-up of the rainfall index insurance which is measured as 1 if a paired rural household-urban migrant link purchases insurance and 0 otherwise.
 The value of the capital or acreage insured, measured in CFA insured for the capital, and in hectares for the acreage.

i. Risk coping and management mechanisms
a. Land area allocated to food crops and cash crops; and to crops with different levels of sensitivity to variations in rainfall;
b. Seasonal migration from rural households and participation in non-farm economic activities;
c. Off-farm activities;
d. Education: measured as enrollment of children between 6 and 15 years old;
e. Livestock (herd size and sales);
ii. Investments in agriculture
a. Cultivated area: measured as hectares of land cultivated;
b. Fertilizer use, measured in kilograms of fertilizers per hectare of cultivated land;
c. Labour input, measured in labour per unit time per hectare of cultivated land, at different stages of the farming cycle;
d. Improved seeds, measured as acreage planted with improved varieties;
e. Agricultural productivity, measured as the value of harvest (net of marketed input costs) per hectare.
iii. Consumption smoothing and income growth
a. Consumption smoothing by rural farmers, measured as changes in consumption. Insurance should shelter consumption from income shocks;
b. Income growth: measured as the change in household total income between the baseline and the endline.

The evaluation exploits the random assignment of rural households and their migrant relatives to treatment and control groups. First, we assign villages into a treatment group (where the standard insurance product will be marketed) and into a control group. Urban migrants from both treatment and control villages will be randomly assigned to two groups. The first migrant group will receive offers to purchase the insurance product to cover their rural relatives, but the second migrant group will not (the control). We denote by V0 and V1 the control and the treated villages, and by U0 and U1, the control and treated migrants. Thus, considering pairs of rural farmers and urban migrants, we obtain four experimental groups: V0U0 (the insurance product is not marketed either to rural households or their migrant relatives); V0U1 (the insurance is marketed to the urban migrants, but not to their rural relatives); V1U0 (the insurance is marketed to rural households, but not to their urban relatives); V1U1 (the insurance is marketed both to rural households and their urban relatives). Randomization will occur at the village-level in the rural areas, and at the migrant-level in urban areas.

We randomize at the village level in the rural area, and at the individual level in the city. In villages randomized to the treatment group, the insurance provider will offer its standard product, i.e. market rainfall index insurance to rural farmers. Migrants from both treatment and control villages who live in Ouagadougou will be randomly assigned to two groups. The first migrant group will be offered to purchase the insurance product to cover their relatives in the villages, while the insurance product will not be marketed to the second migrant group (the control). To simplify the exposition, we denote by V0 and V1 the control and the treated villages, and by U0 and U1, the control and treated urban migrants. Combining rural households and urban migrants produces four experimental groups that can be described as follows.
1. V0U0: The pure control group: the insurance product is not marketed to rural households or to their migrant relatives.
2. V0U1: The insurance is marketed to the urban migrants, but not to their rural relatives.
3. V1U0: The insurance is marketed to rural households, but not to their urban relatives.
4. V1U1: The insurance is marketed both to rural households and to their urban relatives.

Randomization was done in the office by a computer.

The randomization is the village in the rural area, and the individual migrant for the urban migrants sub-sample.

Yes

88 villages

1770 rural households and 970 urban migrants

For the rural survey: 44 villages in the control group, and 44 villages in the treatment group (with 20 households per village).
For the urban sample, 466 urban migrants in the control group and 534 in the treatment group.

We start with 80 villages, and randomly sample 20 households in each village. We will block villages at the commune level with 8 villages per commune. We then proceed to estimate the minimum detectable effect (MDE) for the economic outcomes of interest: consumption, fertilizers, cultivated area, and education, using data from the Burkina Faso 2010 DHS (for education) and from the Ministry of Agriculture of Burkina Faso for the remaining outcomes (the construction of the latter variables are described in Kazianga and Wahhaj (2017). The data are summarized in Table 02 attached. We calculate the MDE under two scenarios. In the first case, we treat any rural household in a treated link as treated. We also account for the proportion of variance explained by blocking. We assume 20% take-up rate. Under these assumptions and with the figures shown, we can detect a minimum change equivalent to roughly 40% of the standard deviation. This MDE corresponds to CFA 198,802 increase in consumption, 17.6 KG increase in fertilizers, 1.7 ha increase in cultivated area, and 19% increase in current enrollment for education. In the second case, we consider comparing V0U1 with V0U0, and V1U1 with V1U0. In this case, we can again ignore the intra-cluster correlation (as when comparing take-up) and block at the village level. Assuming again a take-up rate of 20%, the MDE is, in this case, about 0.30 for the outcomes we consider.