Climate-indexed insurance as a climate service to drought-prone farmers: Evidence from a discrete choice experiment in Sri Lanka

Abstract

The droughts of 2016–2017 caused nearly 25% reduction in harvest and nearly 30% reduction in the area cultivated in Sri Lanka. As a result, renewed attention is being paid to farmer income smoothing and adaptation strategies, among which insurance contends as one of the more viable solutions. This study assesses the willingness to pay for a climate-indexed agricultural insurance package through a discrete choice experiment. We use the ‘stated preference’ approach by offering farmers choice scenarios constructed as a fractional-factorial assignment. The attributes include the coverage area for weather index calculation (village, divisional-secretariat area, or district), the managing authority for the insurance scheme (government, a commercial bank, or an agribusiness company), the method of calculation of compensation (fixed rate, based on cost of inputs or based on the value of output/revenue) and the premium per term (at three levels). We analyzed a total of 2583 choice scenarios evaluated by 287 individuals using a conditional-logit model and estimated the ‘marginal willingness to pay’ (MWTP) for each attribute. The results of the choice experiment reveal the following. The smaller-sized administrative area level is preferred by respondents as a reference area for weather-index calculation while the government is preferred as the managing authority. The revenue-based compensation approach is preferred as the method of calculating compensation. Negative MWTP was observed for larger area indices and for insurance administered by an agribusiness company. The MWTP for revenue-based compensation is LKR 326 while the MWTP for fixed rate compensation schemes is LKR -420.