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

Area of Science:

• Economics
• Regional Science
• Spatial Econometrics

Background:

• Java Island significantly contributes to national Gross Regional Domestic Product (GRDP).
• Analysis revealed spatial dependence, heterogeneity, and outliers in Java Island's GRDP data.
• Understanding factors influencing regional economic performance is crucial for development.

Purpose of the Study:

• To analyze the economic conditions of Java Island by examining Gross Regional Domestic Product (GRDP).
• To identify key factors influencing GRDP across different economic levels.
• To apply advanced spatial econometric models for robust analysis.

Main Methods:

• Employed the Spatial Autoregressive Quantile Regression (SARQR) model to handle spatial dependencies and heterogeneity.
• Utilized Instrumental Variable Quantile Regression (IVQR) for model estimation.
• Assessed variable significance using the Wald test and model performance via the Quantile Verification Skill Score (QVSS).

Main Results:

• The number of workers significantly impacts GRDP across all quantiles.
• Local Own-Source Revenue (LOSR), Regency/City Minimum Wage (RCMW), and Human Development Index (HDI) showed quantile-specific effects.
• The SARQR model achieved its best performance at the 0.6 quantile (QVSS = 0.9052).

Conclusions:

• The number of workers is a consistent driver of GRDP on Java Island.
• Economic factor impacts vary depending on the GRDP level, necessitating quantile-specific analysis.
• The SARQR model provides a robust framework for analyzing regional economic data with spatial complexities.