In this section, the hypothesis about the Japanese FDI's determinants is presented.
This paper hypothesizes that Japanese FDI flows in North America is the increasing function of the marginal productivity of capital of domestic production(denoted by MPK, later) minus the marginal productivity of capital of foreign production(denoted by MPK*, later). Under some conditions, the Japanese FDI in North America is the increasing functions of MPK*s and the decreasing functions of MPKs.
If we assume that the prices of home firms' and host firms' products are the same, that is:
the following regressable equation is obtained:
where ¦Á and ¦Â are parameters.
The equation(4-1-b) can be rewritten as follows:
where ¦Â is a positive parameter.
If this hypothesis, expressed by the equation(4-1-c), holds for the Japanese FDI movement since the late 1980s, the estimated value of the marginal productivity of capital invested for foreign production(MPK*) is likely to rise in the late 1980s and to decline in the early 1990s, under the condition that the marginal productivity of capital invested for domestic production(MPK) remains unchanged overtime. Likewise, the estimated value of the marginal productivity of capital invested for domestic production(MPK) is likely to decrease in the late 1980s and to increase in the early 1990s, under the condition that MPK* remains unchanged over time.
Although the equation(4-1-b) should be used in the following empirical study, it is not used explicitly due to the reasons described below. So, instead of examining the FDI flow and the differences between MPK* and MPK, the movements of estimated MPK and MPK* by themselves are examined.