ESTIMATOR GENERALIZED LEAST SQUARE UNTUK SEEMINGLY UNRELATED REGRESSION PADA KASUS INDEKS PEMBANGUNAN MANUSIA DI SUMATERA

Abstract

Sumatera is one of the largest islands in Indonesia, and it consists of 10 provinces. Based on data from Badan Pusat Statistik the Human Development Index (HDI) in Indonesia, all provinces on the island of Sumatera have high HDI rates. This study aims to analyze and determine the factors that influence the HDI rate in Province of Sumatera island. Factors thought to be influential are the percentage of poor population (PPM), the level of open unemployment (TPT), literacy rate (AMH) and Gini Ratio. Based on the results of the analysis by comparing the results of Seemingly Unrelated Regression (SUR) estimation of Generalized Least Square (GLS) and also Ordinary Least Square (OLS), it’s concluded that the SUR estimation results with GLS estimate produce a more efficient equation with smaller error. The results of GLS estimation produce a greater coefficient of determination, which is 99.99%, while OLS estimation is 98.86%.