Purchasing power parity in Korean city panels with disaggregate price indices

Abstract

We examine whether the purchasing power parity holds between Korean major cities. We use the disaggregated consumer price index data for Korean cities including 13 tradable or nontradable goods. We apply panel unit root tests of Maddala and Wu (1999), Choi (2001) and Im et al. (2003). From the empirical result, we are able to reject the nonstationary null hypothesis for six of eight tradable goods and for two of five nontradable goods at the 10% level. Thus we could conclude that the theory holds between Korean major cities. Also, it holds more for tradable goods than nontradable goods.

Notes

¹ Ulsan, the seventh metropolitan city of Korea, is not considered because it has relatively short time series data by having become a metropolitan city only relatively recently in 1997.

² Esaka (2003) used 13 disaggregated CPI of seven cities including the benchmark city Tokyo, over a period of 1960–1998. He made use of the IPS and Maddala and Wu (1999) test.

³ We take the results of the original data from Table 2 in Esaka (2003).

Table 4.  Results of Korea and Japan data