This paper tries to account for the Japanese FDI's movement in North America, by introducing the dynamic-optimization-based model which determines the optimal FDI. The empirical result reveals the following interesting fact.
The hypothetically consistent estimated values of marginal productivity of capital invested for foreign production in North America are obtained in this empirical study. It is rare that microeconomic-, optimization-based models can explain well the economic up-and-downs empirically. This may shed light on the debate on the characteristic of the FDI phenomena: Is FDI a microeconomic phenomenon or a macroeconomic phenomenon? The empirical result in this paper implies that there is a possibility that FDI is not so much a macroeconomic phenomenon as a microeconomic phenomenon.