Penerapan Metode Small Area Estimation dengan Pendekatan Hierarchical Bayesian (Studi Kasus: Tingkat Pengangguran Terbuka Di Provinsi Kepulauan Riau)

Abstrak

The Open Unemployment Rate is an important indicator in describing the labor market conditions of a region. The Riau Islands Province recorded the highest open unemployment rate in Sumatra in 2024. However, direct estimation based on the survey is less accurate for small areas due to limited sample sizes. This study aims to estimate the OUR using the Small Area Estimation (SAE) approach based on a Hierarchical Bayesian (HB) Log-Normal–Normal model. The analysis results show that among several predictor variables used, only the Gross Regional Domestic Product (GRDP) has a significant effect on the OUR, with a regression coefficient of −1.2600. This indicates that an increase in GRDP is correlated with a decrease in the unemployment rate. The best model obtained is log(θi) = 1,5875 − 1,2600 x1i + vi with a Mean Squared Error of 0.309. For further development, it is recommended to conduct estimation at smaller area levels (sub-districts or villages) and to consider adding more predictor variables to improve model precision.