Law Of One Price Across Regions Studied

From the perspective of Law of One Price(LOP), this paper studies the price relationship among 36 major Chinese cities by using disaggregated absolute price data. Firstly, Balassa-Samuelson effect is found across cities:average price across goods of each city is positively correlated with its wage per capita. Moreover, wage per capita is also positively correlated with average price across tradable goods of each city and with that across non-tradable goods, respectively. Wage per capita has more significant impact on the average price across non-tradable goods. This is because there is non-traded inputs in price of all kinds of goods and service, and on average, the share of non-traded inputs in tradable goods is lower than that in nontradable goods. Secondly, difference in LOP deviations across goods is observed. We find that good-by-good measures of price dispersion are negatively correlated with the tradeability of the good and positively correlated with the share of non-traded inputs in each good or service. This result expands the commonly held view that tradeability is the most important reason which determines price dispersion among Chinese cities. Finally, we find that there is neither trend of convergence nor divergence in the goods market during 1990-2005, and the relative prices of almost all goods are stationary. We argue that the stationarity of relative prices and the observation that there is neither trend of convergence nor divergence is consistent with each other. This view challenges the related literatures which studied the integration of China's goods market by using stationarity test. Moreover, we find that the half-lives estimated are much shorter than the findings in the PPP research within other countries. We guess this is due to the small sample bias and the short time span of our data.