On the Balassa-Samuelson Effect in Japan

The real effective exchange rate (RER) is inherently a general equilibrium variable and its fluctuations are influenced by various factors. In addition to supply factors such as productivity, demand factors, home bias, risk sharing, fiscal and monetary policies also affect the RER. In this context, the "Balassa-Samuelson effect" (B-S effect) focuses on the role of productivity differentials in the tradable sector in explaining the long-run trend of the RER. In this paper, we quantitatively examine the extent to which the B-S effect has been observed in Japan's RER since the 1970s by constructing and estimating a two-country (Japan and the United States), two sector (tradable and non-tradable), dynamic stochastic general equilibrium (DSGE) model. In addition, we also examine cases where the law of one price does not hold in tradables (dominant currency pricing and local currency pricing) and restrictions on labor mobility across sectors. Our results indicate that the long-run trend of the RER in Japan and the United States can be explained to a considerable extent by the B-S mechanism. In other words, according to the model analysis, the yen's appreciation trend in the RER from the 1970s to the mid-1990s can be explained by the rising relative productivity of Japan's tradable sector relative to the U.S., as pointed out in previous studies, and the effects of the Plaza Accord in 1985. In addition, the depreciation of the yen in real terms since the mid-1990s can be explained by a decline in the relative productivity of Japan's tradable sector relative to the United States; the "reverse B-S effect" from Japan's perspective.