The Relative City Price Convergence in Pakistan: Empirical Evidence from Spatial GLS

It is evident from general experience that price of same good may differ considerably among countries, regions, cities in same country and even adjacent shopping malls and outlets. It is also common knowledge that stronger competitive forces and information about market price tend to ensure convergence of prices. In the presence of these forces price differentials cannot be persistent and are hence short lived. The recent literature on price convergence has focused on country studies using regional commodity prices and Consumer Price Index (CPI) data.1 The analysis of relative prices or real exchange rates between regions or cities in a country has certain advantages in estimating Purchasing Power Parity (PPP) puzzle. There are no trade barriers and non tradable goods in a single country. Krugman and Obstfeld (2007) consider transportation costs, trade barriers and goods market segmentations as obstacles to hold international PPP. Furthermore they mention that countries have different endowments, baskets of goods and consumption weights in their inflation index. So PPP may not hold even if there are no non tradable goods and barriers. The PPP theory is related to the law of one price through arbitrage of international goods. The estimation of real exchange rates among countries shows that the convergence towards PPP is very slow.2 This study attempts to use overall Consumer Price Index (CPI) data on 35 Pakistani cities from July 2001 to June 2008 to estimate relative city price convergence with Karachi and Lahore, two numeraire cities. The case of Pakistan is interesting primarily due to the following reasons.