Pemodelan Angka Kemiskinan dengan Geographically Weighted Panel Regression (GWPR) di Provinsi Aceh

Abstract

Aceh Province is one of the provinces in Indonesia that still faces poverty problems. Aceh prov-ince occupies first position with the highest percentage of poor people for the island of Suma-tra. This indicates that there are still weaknesses in the programs and policies implemented by the Aceh provincial government in alleviating poverty. Observations related to this are needed to find out the factors that influence the problem of poverty along with repeated observations so that we can observe changes in the percentage of poor people that occur from year to year. One regression analysis that can be used to model this problem is geographically weighted panel regression (GWPR). This research aims to model the poverty rate and determine the var-iables that significantly influence the poverty rate in districts/cities in Aceh Province. Predictor variables that are thought to influence poverty in this study include life expectancy, per capita expenditure, gini ratio, open unemployment rate, and average years of schooling. The selection of the best model is shown through the coefficient of determination value. Based on the analysis results, the largest coefficient value was obtained in the GWPR model with 78.6% compared to the FEM panel data regression model with 73.4%.