Regional Economic Growth and Income Inequality in Indonesia and China

(1)基于区级GDP和人口数据，采用Theil指数估算1995-1998年中国区域收入不平等，并进行两阶段嵌套不平等分解分析，探讨区域收入不平等的影响因素。我们还进行了回归分析，以探讨省内收入不平等的可能决定因素。分解分析表明，1998年省内不平等因素占总体区域收入不平等的62%，而地区间不平等因素占27%。通过回归分析发现，累积人均FDI和指定内陆边境省份的虚拟变量对解释省际不平等具有显著意义。我们还发现，经济二元论（以农业劳动生产率与非农业部门劳动生产率之比较低为表征）是导致省内不平等的另一个重要因素。(2)利用印度尼西亚1993 - 1999年的省级数据，运用加权变异系数和Theil T指数测度了人均GDP和劳动生产率的区域不平等，并运用多种不平等分解技术探讨了区域收入不平等的影响因素。我们还比较了印尼1993-1999年的区域不平等模式与日本战后时期的区域不平等模式，并预测了印尼未来的区域不平等模式。(3)研究了1985—1995年日本制造业地理格局的变化，探讨了制造业地理集聚的影响因素。本文通过回归分析对新经济地理学早期模型的一些假设进行了检验。回归结果表明，日本制造业的地理集中度似乎是由内部规模经济、运输成本和要素强度的某种组合决定的。然而，产业间联系对地理集中度的影响不显著。根据NEG理论，内部规模经济较大、单位运输成本较小的日本制造业往往具有较高的地理集中度。日本制造业数据也支持Heckscher-Ohlin理论：劳动或资本密集型产业往往具有更高的地理集中度。(4)运用加权变异系数和Theil T指数对战后日本地区收入不平等进行测度，并运用多种不平等分解技术探讨影响地区收入不平等的因素。以加权变异系数衡量的人均国内生产总值的地区不平等首先扩大，并在1958年达到0.38的峰值。随后稳步下降，并在1979年跌至0.25的最低点。1979年后，它再次上升，并在1990年达到0.37的峰值。1990年以后有下降的趋势。人均GDP的地区不平等在很大程度上是由劳动生产率的地区不平等决定的。劳动参与率的地区差异不显著。第一部门的人均GDP不平等程度的迅速上升是由于其就业份额的不平等程度的上升。另一方面，直到20世纪70年代中期，第二产业人均GDP不平等程度的下降主要是由于其就业份额的下降，而第三产业人均GDP不平等程度的下降主要是由于其劳动生产率和就业份额的不平等程度的下降。少

(1)We estimated regional income inequality in China over the 1995-1998 period using a Theil index based upon district-level GDP and population data, and conducted a two-stage nested inequality decomposition analysis to explore factors determining regional income inequality. We also performed a regression analysis to explore the possible determinants of within-province income inequality. The decomposition analysis showed that the within-province inequality component accounted for 62 percent of overall regional income inequality in 1998, while the between-region component contributed 27 percent. According to the regression analysis, cumulative per capita FDI and a dummy variable designating inland border provinces are found to be significant in explaining within-province inequality. We also found that economic dualism, as denoted by a low ratio of agricultural labor productivity to labor productivity in non-agricultural sectors, was another significant factor contributing to within-provi … More nce inequality.(2)We employed provincial data to measure regional inequalities in per capita GDP and labor productivity in Indonesia from 1993 to 1999 by the weighted coefficient of variation and the Theil T index, and investigated factors determining regional income inequalities by using several inequality decomposition techniques. We also compared Indonesia's pattern of regional inequalities in 1993-1999 with Japan's pattern in the postwar period and predicted the future pattern of regional inequalities in Indonesia.(3)We investigated the changing geographical pattern of manufacturing industries in Japan between 1985 and 1995 and explored factors of their geographic concentration. A regression analysis was conducted to test some hypotheses that were derived directly from early models of the New Economic Geography (NEG). Regression results indicated that the geographic concentration of Japanese manufacturing industries seems to be determined by some combination of internal economies of scale, transportation costs, and factor intensity. However, inter-industry linkages were found to be an insignificant factor of geographic concentration. As posited by the NEG theories, Japanese manufacturing industries with larger internal economies of scale and smaller unit transportation costs tended to have a higher level of geographic concentration. Japanese manufacturing data also supported the Heckscher-Ohlin theory : labor- or capital-intensive industries tend to have a higher level of geographic concentration.(4)We measured regional income inequalities in Japan in the postwar period by using the weighted coefficient of variation and the Theil T index and explored factors determining regional income inequalities by using several inequality decomposition techniques. Regional inequality in per capita GDP, as measured by the weighted coefficient of variation, first increased and reached a peak in 1958 at 0.38. It then declined steadily and hit the bottom at 0.25 in 1979. After 1979, it rose again and reached a peak in 1990 at 0.37. There is a declining trend after 1990. To a considerable extent, regional inequality in per capita GDP was determined by regional inequality in labor productivity. Regional inequality in labor participation rate was not significant. A rapid rise in primary sector's inequality in per capita GDP was attributable to a rise in its inequality in employment share. On the other hand, a decrease in secondary sector's inequality in per capita GDP until the middle of the 1970s was due mainly to a fall in its inequality in employment share, while a decrease in tertiary sector's inequality in per capita GDP until the middle of 1960s was attributable to a fall in its inequality in both labor productivity and employment share. Less