A Study to Determine Optimum Critical Yield Index of Crop Insurance Policy in Java

Abstract

Zone based area yield scheme in crop insurance is proposed to be an alternative policy to improve crop insurance in Indonesia, MPCI. Based on historical data, we have found that the use of province based for the zone area yield scheme is not ideal. The new cluster-based, that is lower disaggregate of province, results the smaller ratio Sw/Sb than province-based. Using Ward-like hierarchical method, it does not only utilize the average of paddy productivity which becomes a new potential to obtain the critical yield index of the zone, but also the Queen contiguity matrix by regency/municipality to cluster similar characteristics of farmers. The areas formed are three clusters on each West Java, Central Java, and East Java, two clusters on Banten, and DI Yogyakarta as a whole group. In addition, to facilitate the overview of the characteristics of clusters, we also classify the 12 clusters formed in Java into three levels of productivity which are reviewed nationally. It tends to have low (8.33%), middle (50%), and high (41.67%) productivity. The volatility rates of each clusters are also compared to the control limit of risk for its classified level, as the consideration of using over ten years paddy productivity data. The 12 clusters in Java are assumed to be controlled, because all clusters have volatility within their control cluster limit. The controlled risk of the clusters implies that the change in the critical yield index, which has been predicted in this paper based on regencies/municipalities productivity paddy from 2007 to 2018, in the next few years should not be extreme. This affects the premium rate making of conducting a cluster-based area yield scheme. As we have found, cluster-based area yield scheme in crop insurance is proposed to be an alternative policy to improve crop insurance in Indonesia. Developing clusters in order to group similar farmers or farmer groups can minimize heterogeneity yield as the basis of risk the and increase risk-management effectiveness. Using Bootstrap generated data of surveyed area in Java, the standard risk measure of extreme event (tail-value-at-risk) in the crop loss also shows that insurance company should retain smaller amount of indemnity as cluster scheme is applied. On top of that, lower bound of Two Sigma could be considered as an alternative of critical yield index. This technique provides lowest pure premium and standard deviation, which are obtained from its indemnity, among other methods.