PENDEKATAN REGRESI SPASIAL DALAM PEMODELAN TINGKAT PENGANGGURAN TERBUKA (Spatial Regression Approach in Modeling of Open Unemployment Rate)
Abstract
Solution of unemployment rate becomes one of the focuses of Indonesia development. The highest open unemployment rate is in Java. So it is interesting to study about factors that influence open unemployment rate in Java. The connection between area (spatial effect) are need to be considered. In this case, there is a spatial effect that can be solve with regression of area approach. Spatial regression with area approach are Spatial Autoregressive Model (SAR), Spatial Error Model (SEM), and General Spatial Model (GSM). The model selection criteria are the coefficient of determination (R2 ), value of variance estimate and the value of AIC (Akaike’s Information Criterion). The results show that GSM and SAR is better regression model than OLS and SEM and the factors that affect open unemployment rate are the percentage of people who did not complete primary school (SD), the gross regional domestic product constant prices for regency/city, percentage of people working in agriculture and minimum salary for regency/city. Keywords : Spatial Regression, GSM, SAR, SEM.