Table(7-1) shows that, in this empirical study, MPK is not a significant determinant of the Japanese FDI in North America from 1984 to 1993. If the hypothesis presented in Section 4 holds in the late 1980s, the values of MPKs are likely to decrease, but such trends of MPK is not observed in Table(7-1). Likewise, if the hypothesis holds from 1989 to 1993, there is a strong possibility for the values of MPKs in that periods to increase, but such movements of MPKs is not observed in Table(7-1).
Although there are many to be considered to evaluate the empirical result, in this empirical study the marginal productivity of capital invested for domestic production(MPK) has not played significant roles in determining the FDI movements from 1984 to 1993.
But one must bear in mind the fact that the home firms' output(Y) and the home firms' capital stock(K) used in estimating MPK are not exactly equivalent to the sum of outputs and capital stocks of all home firms. This means that in this empirical study, outputs and capital stocks of non-multinational firms are included in the Y and K in the equation(5-1-2). Theoretically and conceptually, the variables estimated in the following way should have been used:
where 
is the home firm's output of the MNC 
,
is the capital stock of the home firm of the MNC ,
and 
is the number of the MNCs concerned.
Since it is hard to use this method statistically, this method is
not used in this paper.
With this in mind, one could say that there are some possibilities that
one can obtain the hypothetically
consistent MPK's values if one use the data of output and capital stock
estimated by (7-1-1).
One of ways to use (7-1-1) would be to use the firm-specific data
of the Japanese top 20-30 MNCs in manufacturing sector and
aggregate them
.