A viable and cost-effective weather index insurance for rice in Indonesia

The potentially adverse effects of droughts on agricultural output are obvious. Indonesian rice farmers have no financial protection from climate risk via catastrophic weather risk transfer tools. Done well, a weather index insurance (WII) program can not only provide resources that enable recovery, but can also facilitate the adoption of prevention and adaptation measures and incentivise risk reduction. Here, we quantify the applicability, viability, and likely cost of introducing a WII for droughts for rice production in Indonesia. To reduce basis risk, we construct district specific indices that are based on the estimation of Panel Geographically Weighted Regressions models. With these spatial tools, and detailed district level data on past agricultural productivity and weather conditions, we present an algorithm that generates an effective and actuarially sound WII, and measure its effectiveness in reducing income volatility for farmers. We use data on annual paddy production in 428 Indonesian districts, reported over the period 1990-2013, and climate data from 1950-2015. We use the monthly Palmer Drought Severity Index and identify district-specific trigger and exit points for the insurance plan. We quantify the impact of this hypothetical insurance product using past production data to calculate an actuarially-robust and welfare-enhancing price for this scheme.