MODELING FACTORS AFFECTING EDUCATED UNEMPLOYMENT ON JAVA ISLAND USING GEOGRAPHICALLY WEIGHTED POISSON REGRESSION MODEL

Abstract

The eighth goal of the SDGs, which aim to promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all, addresses the problem of unemployment. Indonesia, the fourth-largest country in the world, keeps on dealing with unemployment and its negative consequences. Three provinces on the island of Java have higher unemployment rates for educated people than any other provinces. The purpose of this study is to examine the variables affecting educated unemployment in Java. This study uses cross-sectional data published from BPS-Statistics Indonesia website and the Indonesia Investment Coordinating Board (BKPM) for 119 regencies/cities across six provinces on Java Island in 2021. The predictor variables are Gross Regional Domestic Product (GRDP), investment, labor force participation rate, mean years of schooling, regency/city minimum wage, and inflation. The number of working-age population is used as an exposure measure. Four models were used to analyze the factors affecting educated unemployment on Java Island: Poisson regression model and Geographically Weighted Poisson Regression (GWPR) model, both with and without an exposure. Based on the smallest AIC/AICc, the best model is the GWPR model with an exposure. This model creates 11 groups of locations based on the same predictor variables that significantly affect educated unemployment