Spatial autocorrelation analysis of multivariate rural insurance data

Abstract

The environment in which agricultural activities are developed presents high risk and great uncertainty. Several factors related to the agricultural sector can generate fluctuations in the income of producers. These fluctuations must be faced through risk management support policies such as, for example, the hiring of rural insurance. This type of insurance enables the recovery of the financial capacity of the producer in the occurrence of adverse events that cause economic damage. Considering the relevance of rural insurance in the agricultural sector, this study aims to evaluate the spatial distribution of the variables of this type of insurance in Brazilian municipalities from 2006 to 2019. The data used were obtained from rural insurance censuses compiled by the Ministry of Agriculture, Livestock, and Supply. Principal Component Analysis (PCA)was used to reduce data dimensionality and Exploratory Spatial Data Analysis (ESDA) using the scores of the first PC was used to investigate the presence of spatial distribution patterns of rural insurance. By using PC scores, it was found that the highest concentrations of rural insurance policies are located in the South and Midwest regions of Brazil, and there is a tendency for an increase in the spatial dependence of rural insurance throughout the analyzed period. The identification of these areas shows how rural insurance is heterogeneously distributed in Brazil. This result suggests that some strategies can be adopted by policy makers and insurers in order to serve areas that have demand and are not yet covered by rural insurance.