This study aims to analyze China’s regional disparity during 1997-2006 by adopting Keng’s (2004) general equilibrium analysis. China’s 31 provinces are divided into ten regions. The intra- and inter regional disparity as well as the Effective Regional Disparity (ERD) are all investigated in this study. Based on ample quantitative findings, in 2006, the largest intra-regional disparity existed in the Coastal North while the smallest intra-regional disparity existed in the Inner West. The largest inter-regional disparity existed between the Coastal East and the Inner South not between the richest region (the Coastal North) and the poorest region (the Inner West). This study also uses the Effective Regional Disparity (ERD) to understand the magnitude of an individual province’s contribution to the overall regional disparity. The top five on the ERD ranking of China’s 21 lower income provinces in 2006 were Henan, Sichuan, Anhui, Hebei, and Hunan. Hence, in order to reduce the regional disparity, the Chinese central government needs to focus on the heavily populated provinces in the Inner area rather than in the West.

區域發展差異現象一直以來為國家政策所關注，而近年來台灣地區失業率有逐漸上升的趨勢，各縣市之表現亦大相逕庭，顯示各地區存在失業差異現象。過去研究較少以空間觀點觀察失業相關議題，此外，關於區域差異因素對失業率之影響鮮少納入政府規劃因素。因此，本研究以空間自相關分析方法檢測失業是否具有空間相關性及聚集性，並應用長期追蹤資料（panel data）迴歸模型，以人口、產業、所得、都市化程度及政府規劃因素，分析台灣22縣市1988至2008近二十年來各區域差異因素對失業率之影響，藉由實證結果提出相關都市及產業政策之建議。實證結果發現，台灣失業分佈具有一定程度的空間相關性，且高低失業率在各縣市間亦有聚集現象。再者，依固定效果模型實證結果發現人口數、工業及服務業就業者百分比、都市化程度、工業區面積百分比與失業率間呈現顯著正向關係；經濟發展支出百分比與失業率呈現顯著負向關係；區域固定效果，即排除自變數影響下，各縣市本身區域特質對失業率之影響，結果顯示台北縣及桃園縣之係數為負向，南投縣、嘉義縣、台東縣與花蓮縣之係數為正向；時間固定效果方面，大部分年度皆具顯著性，且係數有由負轉正之趨勢，代表特定時間衝擊會對失業率造成影響。 / Differences in regional development have been a focus on national policies. Recently, there is a increasing trend in the unemployment rate in Taiwan, and it also differs from cities and counties, indicating there exists differences in regional unemployment. Previous research rarely combined unemployment issues with spatial perspective. In addition, the effect of regional discrepant factors on the unemployment rate rarely take government planning factors into account. Therefore, this study uses spatial autocorrelation analysis to detect whether unemployment has spatial correlation and aggregation, and applies panel data regression model with population, industry, income, the degree of urbanization, and government planning factors to analyze the effect of regional discrepant factors on the unemployment rate in Taiwan's 22 cities and counties from 1988 to 2008. According to the empirical results, we come up with some urban and industrial policy proposals. Empirical results indicate that the distribution of unemployment in Taiwan has a certain degree of spatial correlation, and high or low unemployment rate also has aggregation among cities and counties. Furthermore, according to the results of the fixed effects model, population, the percentage of industrial and service sector employment, the degree of urbanization, and the percentage of industrial area show a significant positive relationship with unemployment rate. The percentage of expenditures for economic development shows a significant negative relationship with unemployment rate. Region-specific fixed effect, which exclude the influence of independent variables, is the effect of regional characteristics of counties and cities on the unemployment rate. This result shows the coefficient of Taipei County and Taoyuan County is negative, and the coefficient of Nantou County, Chiayi County, Taitung County and Hualien County is positive. As for time-specific fixed effect, almost all years are significant, and the coefficient has the trend from negative to positive, indicating that a particular time impact will affect the unemployment rate.