ESTIMASI PARAMETER RANDOM EFFECT MODEL PADA REGRESI PANEL MENGGUNAKAN METODE GENERALIZED LEAST SQUARE (STUDI KASUS: KEMISKINAN DI PROVINSI KALIMANTAN SELATAN)

Abstract

Poverty is a condition that concerns the inability to meet the most minimum demands of life, especially from the aspects of consumption, income, education, and health. The problem of poverty is very complex and multidimensional in nature, as it relates to social, economic, cultural and other aspects. This study focuses on observation areas in South Kalimantan Province, with the PPM value in 2021 reaching (4.83%) still above the target goal of the Regional Medium-Term Development Plan (RPJMD) of South Kalimantan Province (3.96- 4.01%), so that further interventions are still needed to be able to reduce PPM in poverty cases. This study aims to estimate the parameters of the panel regression model used to analyze factors that are suspected to affect poverty cases in South Kalimantan Province in 2016-2020. The Random Effect Model (REM) is the best model used in this study, assuming that there are differences in slopes and interceptions caused by residual due to differences between individual units and between time periods. The process of estimating parameters on REM is determined through the Generalized Least Squares (GLS) Estimator method . From the results of the data processing, it was obtained that the model is influenced by economic growth, life expectancy, open unemployment rate, and labor force participation rate. From the results of the analysis of 2 (two) models, it was tested significantly and affected poverty in South Kalimantan Province in 2016-2020. Keywords: Poverty, Data Regression Panel, Generalized Least Square Method (GLS).