CLUB CONVERGENCE & REGIONAL SPILLOVERS IN EAST JAVA

This study try to identify the â-convergence process among regions in East Java using panel data of 37 regencies & municipalities between 1983-2002, taking into account the presence of spatial heterogeneity and spillover effects. Detection of spatial regimes using G-I* statistics (Getis & Ord, 1995) on regional per capita GDP values in 1983 found cluster of high income regions (group of “rich”) in central & eastern part of East Java, and cluster of low income regions (group of “poor”) in western part. The result of OLS & GLS regression on absolute convergence model found the existence of â-divergence process of East Java in overall period (1983-2003), consistent with the ó convergence which showing upward trend (divergence). Meanwhile, the same divergence process is also found in absolute convergence equation estimated for each club, even though in slower rate than East Java divergence rate. Using the methodology proposed by Burn, Combes, & Renard (2002) this study founds the existence of negative spillover effects between regions in “rich” clubs and from “rich” clubs to the “poor” one, where the magnitude is greater in the latter case. The club of “poor” regions is diverging faster than the “rich”. This finding is robust in every convergence equation (with or without the spillover effects). The lack of diversity on East Java’s manufacturing industries (Santosa & Michael, 2005 and Landiyanto, 2005) seems contribute to its divergence process by engaging a competitive mode between regions.