PENERAPAN MODEL GEOGRAPHICALLY WEIGHTED PANEL REGRESSION PADA TINGKAT KEMISKINAN DI PROVINSI KALIMANTAN SELATAN

Abstract

South Kalimantan Province is one of the provinces in Indonesia which has the lowest poverty rate or percentage of poor people on the island of Kalimantan, even in Indonesia. The percentage of poor people in South Kalimantan Province in March 2022 was 4.49% or in the 2nd lowest poverty position in Indonesia, below the Bangka Belitung Islands Province and above the Bali Province. Geographically Weighted Panel Regression (GWPR) is a local regression model with repeated data at location points for each observation at different times. This study aims to estimate the GWPR model parameters and test the significance of the GWPR model parameters to determine the factors that influence poverty in South Kalimantan Province. The independent variables used affect the dependent variable in the form of the Percentage of Poor Population, namely Life Expectancy, Open Unemployment Rate, Economic Growth, Average Years of Schooling and Number of Crimes. The analysis in this study is descriptive analysis using thematic maps, panel data regression analysis to determine the global model and GWPR by combining the panel data model with the GWR model. The results of this study show that the fixed effect model is a global model and the fixed bisquare weighting function is the best weighting function for estimating the GWPR model. Based on the GWPR model formed, there are 7 model groups based on significant independent variables. Hulu Sungai Utara and Hulu Sungai Tengah districts are districts where poverty in these areas is influenced by many variables compared to other regions in South Kalimantan Province.