Spatial autoregressive quantile regression modeling of gross regional domestic product data in Java Island

Abstract

This study analyzes the Gross Regional Domestic Product (GRDP) data of Java Island to understand the region's economic conditions and the key factors influencing GRDP. Java Island contributes the largest share to the national GRDP; however, the data show the presence of spatial dependence, spatial heterogeneity, and outliers. To address these issues, this study employs the Spatial Autoregressive Quantile Regression (SARQR) model. The findings show that the number of workers significantly influences GRDP across all quantiles, whereas Local Own-Source Revenue (LOSR), Regency/City Minimum Wage (RCMW), and the Human Development Index (HDI) only impact specific quantiles. The SARQR model demonstrates the best performance at the 0.6 quantile, with a Quantile Verification Skill Score (QVSS) of 0.9052, indicating strong explanatory power. • The model was estimated using the Instrumental Variable Quantile Regression (IVQR) method. • Variable significance at each quantile was tested using the Wald test. • The best model was selected based on the QVSS value.

Access to Document

Other files and links

• Link to publication in Scopus
• Open Access Version Available

Fingerprint