Japan's inflation under the Bretton Woods system: How large was the Balassa–Samuelson effect?

Abstract

Japan's real exchange rate appreciation during the post-WWII manufacturing-led growth period has been regarded as a classical example of the Balassa–Samuelson effect. We choose the most conspicuous sub-period—1956–1970—to confirm the effect. Japan was in a rapid growth period under the U.S. dollar peg (real GDP growth, 9.7% per annum). The nominal anchor was weak as Japan's inflation rate (GDP deflator-based, 5.4%) was markedly higher than the U.S. rate (2.6%) during the 15-year period. The decomposition of the annual 2.7% (geometric) Japan–U.S. inflation rate gap (real exchange rate appreciation of the Japanese yen) reveals that the Balassa–Samuelson effect accounted for 0.7%; most of the real exchange rate appreciation (1.7%) was attributed to greater price increases in Japan's tradables. Although Japan's tradable sector achieved high TFP growth, the joint effect of the tradable–nontradable TFP growth difference between the two economies was too small to generate a sizable Balassa–Samuelson effect. Japan's example may suggest that even in rapidly growing economies, the magnitude of the effect in long-run real exchange rate appreciation is generally modest.