Regional Price Differences In Urban China: Estimation And Application

Price plays a remarkable role in the comparison of living standards among residents in different regions.The corresponding conclusions may not reflect the real situation if we ignore the fact that there may be regional price differentials in such comparison and employ nominal indicators directly,influencing the formulation and effectiveness of relevant policies.Therefore,it is of practical significance to estimate the spatial price indexes within China,Previously,most of the studies on this theme adopt similar methods as ICP does.On the one hand,these methods have several limitations in comparison-resistant areas such as education and health.On the other hand,these methods have a high requirement of price data.At present,price data satisfying such requirement are very limited on regional level,affecting the accuracy of empirical results to certain degree.In addition,another weakness of the Basket Cost Method is that it assumes that individuals of different regions have the same basket and does not take into account the discrepancy of environment and preference in different regions.In view of literatures on spatial price indexes,this paper adopts the Engel curve approach proposed by Dipankor Coondoo et al(2011)to estimate the Spatial Price Index(SPI)in urban China.Based on certain demand system,the procedure uses consumption expenditures of urban households to estimate the interprovincial SPIs within China in 2003,2005 and 2012.Besides,the author discusses the reasonableness of the results from three aspects:the Penn Effect within a country,the relationship between SPI and CPI and the comparison with other studies,and then applies the empirical results to the analysis of the income inequality of urban residents in different provinces.There are several findings according to this study.First,there are significant regional price differentials in urban China,in other words,urban people of different regions have different purchasing power parities:the ratios of the province with the highest price level and the one with the lowest price level are respectively 2.7 and 2.3 in 2003 and 2012.Second,the price level of the eastern area is generally higher than that of the middle and western area.Third,Penn effect within the country is significant.The Pearson correlation coefficients between the price level index and the relative nominal income index of urban residents are separately 0.93?0.91and 0.91 with all the p-values much smaller than 0.001,which means the correlation relationship is very significant.Forth,the variation of our estimated SPIs from 2003 to 2012 agrees with the change of corresponding CPIs.During this time,the CPIs of the eastern area increase less than that of the middle and western areas,which can explain the phenomenon that the ranges of SPIs in this period become smaller.Fifth,the empirical results of this paper agrees more with the results from Cathy Honge Gong and Xin Meng,who also employ an economic index approach than the results from Fangdong Yu,who adopts a statistic method.Sixth,the regional price differential is larger based on the Engel curve method compared with those obtained from the Basket Cost method.Finally,the real income inequality adjusted by the SPIs is significantly smaller than the nominal one when we apply the estimated results to this issue,taking into account the range,the coefficient of variation,the Gini coefficient,the Theil coefficient,and the Atkinson index.The innovations of this paper include two aspects.First,this paper tries to estimate the interprovincial SPIs in China from the Engel curve approach which is based on consumer behavior theories from microeconomics,providing a supplement to the measurement of regional price differentials within China.Besides,another benefit of the method adopted is that the whole estimation process does not need price data,which means it does not have to face the common data constrain such as the reconcile of comparability and representativeness and the inadequacy of price data sample,shared by ICP related methods when they were employed to estimate the regional price differences within China.Second,the author mainly employs the consumption expenditure data grouped by income level instead of the price data or unit-value data,which does not need to face the challenges from ICP related methods such as the reconcile of comparability and representativeness,the inadequacy of price data sample and problems from comparison-resistant areas such as health and education.Third,this paper provides a better understanding of real income inequality of different areas though applying the estimated SPIs to the analysis of this issue with the adoption of five different indexes,namely the range,the coefficient of variation,the Lorenz curve and the Gini coefficient,the Theil coefficient,and the Atkinson index.The main shortcoming of this paper is that it does not cover all the provinces in China because several provinces do not publish consumption expenditure data grouped by income level.