Responsible Researcher: Eduarda Miller de Figueiredo
Paper Title: Informal Risk Sharing, Index Insurance, and Risk Taking in Developing Countries
Authors: Ahmed Mushfiq Mobarak and Mark R. Rosenzweig
Location of the Intervention: India
Sample Size: 4.667 families.
Sector: Agriculture
Primary Variable of Interest: Risk-taking
Type of Intervention: Weather insurance offer
Methodology: Experimental Design
The aim of this study was to understand the complex interactions between informal risk sharing, formal insurance, and risk-taking. Formal insurance markets have low penetration in agricultural areas of developing countries because families in these areas have a habit of sharing risks. By using an indexed insurance model with a risk-based cooperative risk-sharing model, the authors observed that farmers increased their risk acceptance by acquiring rain insurance.
Policy Problem
In agricultural areas of developing countries, the penetration of formal insurance markets is surprisingly low, despite the fact that agriculture is highly susceptible to climate fluctuations. What happens is that most rural families participate in the informal sharing of risk; thus, according to the authors, farmers seem to sacrifice profitability to reduce risk.
Implementation and Evaluation Context
The aim of this study was to understand the complex interactions between informal risk sharing, formal insurance, and risk-taking. To this end, the authors used random offers of rain insurance contracts to a set of families living in indigenous villages, which possess readily identifiable and exogenous data characterizing the rich nature and extent of informal risk sharing in networks: the subcaste, or teak.
Jatis These are risk-sharing networks spanning villages and districts in India, where data has shown that most loans and transfers to households are from caste members, and most informal loans and transfers to households and other caste members originate outside the village. Therefore, Jatis They have the potential to indemnify the aggregate rainfall risk at the village level, and thus, such risk sharing can directly replace formal insurance.
Policy/Program Details
The 63 REDS villages in the sample contained 118 Jatis with at least 50 families. Of these, 42 REDS villages were randomly selected to receive insurance marketing.
To ensure a pure control group of families not receiving any insurance treatment, the sample was first stratified by caste, so members of 25 castes were randomly selected to receive no insurance offer. Next, stratification was performed by occupation, with half of the insurance offers going to farmers and the other half to families exclusively dedicated to agricultural work. In the end, approximately 4.667 families received the treatment, with about 40% of all families acquiring some form of insurance.
The product offered provided a cash payment to buyers if the rains were delayed beyond the expected start date of the monsoon season, as determined by... Agricultural Insurance Company of India Lombard (AICI). The price of insurance ranged from $1,6 to $4.
The target of climate insurance programs is farmers, and the provision of climate insurance induces them to take on more risk, which can increase the wage risk borne by landless people who depend on wage labor in agriculture. Mobarak and Rosenzweig (2013) study the overall equilibrium effects of providing rain insurance to farming families and landless workers. Along these lines, Jayachandran (2006) examined how the provision of financial services to landless families that allowed for income smoothing affected the supply of labor. Thus, such families would work more when rainfall is low, having more leisure time when rainfall is abundant, with wages varying between these two periods. This situation, according to the research, would increase the volatility of wage laborers' income, decreasing the volatility of farmers' profits. This study, discussed here, will also examine the labor market spillover effects of providing insurance to landless people on farmers' income.
Method
The data from was used NCAER Rural Economic Development Survey (REDS) 2007/2008, which allowed randomization between and within caste-based risk-sharing groups.
Mobarak and Rosenzweig (2012) place an indexed insurance model with base risk within the cooperative risk-sharing model of Arnott and Stiglitz (1991), demonstrating that:
- a) The underlying risk, or the imperfect correlation between losses and insurance payouts due in part to the remote location of rain gauges, reduces uptake of index insurance;
- b) When the contract does not carry underlying risk, the demand for indicator-based insurance is independent of the extent of informal loss coverage (specific to the family);
- c) Risk sharing and indicator-based insurance can be complementary when there is underlying risk, because the Jatis network will cover the idiosyncratic losses of families precisely when the index-based contract fails.
To test these predictions, the authors use REDS survey data on inter-household transfers in response to village-level rainfall shocks and household-specific adverse shocks to construct informal risk-sharing indices that measure how well each caste in the sample compensates against idiosyncratic losses and aggregate countershocks.
Main results
To assess whether and how variations in informal compensation for family losses empirically affect risk-taking, they explored the idea that among farmers who take more risks, crop production, input use, and profits should be more sensitive to rainfall. Using REDS data, the authors therefore find that in Jatis With higher compensation for individual losses, profits per acre are less sensitive to rainfall, whereas in Jatis In cases where compensation is more heavily dependent on climate shocks, profits were more sensitive to rainfall variations.
Using a random variation in the supply of weather insurance, when evaluating the effects of rain insurance and informal loss compensation on risk-taking, the authors found exactly the same relationships as in the REDS data. Agricultural production increases dramatically with rainfall for farmers with random offers of weather insurance, thus demonstrating that although caste groups in India are evidently successful in mitigating risk, this comes at a substantial cost, i.e., more risk-averse production with lower average returns. Therefore, acquiring rain insurance allows farmers to increase their risk acceptance.
When assessing the labor market spillover effects of offering insurance to landless people on farmers' income, the results suggest that if landless families are aware of the impact of insurance on risk and wage levels, these families will be more receptive to climate insurance when farmers also purchase insurance.
Lessons in Public Policy
By examining the interrelationships between informal insurance agreements, the demand for formal weather insurance, and risk-taking among farmers and landless families in an environment of widespread risk sharing, the authors observed that farmers increased their risk acceptance by acquiring formal insurance.
References
MOBARAK, Ahmed Mushfiq; ROSENZWEIG, Mark R. Informal risk sharing, index insurance, and risk taking in developing countries. American Economic Review, v. 103, no. 3, p. 375-80, 2013.
