The dramatic yen appreciation after the Plaza Accord in September in 1985
brought about the several kinds of impact on the Japanese economy.
Among them are the increase of the Japanese foreign direct investment
in the United States and East Asia(notably, ASEAN economies) and
the shift from domestic production to foreign production.
Horaguchi(1992) shows the strong negative correlations
between the the Japanese FDI flows and the exchange rate(yen/dollar)
from 1971 to 1989.
This increasing movement of FDI, however, stopped in the year around 1990, when the so-called ``bubble-economy'' collapsed, according to the Annual Report of the International Finance Bureau,(Ministry of Finance).
In theoretical literatures, recent years have witnessed the breakthroughs of theoretical study of FDIs, most of which take the game-theoretic, contract-theoretic approaches, like Yanagawa(1994)'s work. However, our theoretical understanding about FDI is still limited.
In empirical literatures, there are many empirical studies on FDI. The previous empirical studies of FDI can be classified into the following three categories. The first is an effort to explain FDI by many factors(to regress FDI on many explanatory variables). Among them are Dunning(1988)'s popular eclectic FDI model and Sazanami(1992)'s work on the determinants of the Japanese FDI in Europe. This effort can be called industrial-organization approaches.
The second is an effort to explain the determinants of FDIs by conducting lots of questionnaires to lots of multinational firms. This effort can be said to be managerial, sociological approaches. Although this explanation has the advantage of being easy to understand intuitively, it tends to be ad-hoc to a great extent.
The third is an effort to explain FDI by some macroeconomic variables, such as Kojima(1985)'s work which takes into accounts the nation's comparative advantages. This effort can be called macroeconomic approaches.
These days most FDI works belong to the former two approaches:
industrial-organization approaches and managerial, sociological approaches.
This might be due to the fact
that there seems to be a consensus that FDI is not a macroeconomic
phenomenon but a microeconomic, managerial and industrial-organization
one.
There are three major, different-level goals of this paper.
The first goal is to analyze empirically the working factors
behind the increasing movement of the Japanese FDI in the late 1980s and
its decreasing movement in the early 1990s, by introducing more
general, not ad-hoc, framework.
Although well-disaggregated empirical analysis is needed to
analyze empirically the Japanese FDI, this paper will focus
on the Japanese FDI in the manufacturing sector in North America(
the United States, Canada and Mexico combined) from 1984 through
1993.
The second goal of this paper is to show the empirical method which is based on the microeconomic theoretical model. As I see it, there are few empirical works on FDI which is based on the rigorous microeconomic theory(model). Most research on FDI is highly managerial, including the excellent works, like Caves(1982), Helleiner(1989) and so on. There are, however, few empirical work within the framework of economics. This paper is expected to serve to present the macroeconomic analytical framework based on microeconomic model.
The third goal is to verify the applicability of the
neo-classical model, which springs from the optimization behavior
of economic agents. In most empirical work on macroeconomics and
microeconomics, it is reported that the neo-classical model
does not explain the real economic phenomena well.
Some macroeconomists
argue that highly microeconomic-based
macroeconomic approach will fail to explain the real
macroeconomic phenomena. If this argument is true and
the FDI is a macroeconomic phenomenon, the following
microeconomic-based model will be likely to fail to explain
the FDI movement. In other words, if this argument is
true and the FDI is a microeconomic phenomenon,
the following FDI model based on optimization of economic agents
will be likely to succeed in accounting for the FDI.
The structure of this paper is as follows. In section 2, the movement of the Japanese foreign direct investment in North America from the late 1970s and the early 1990s is observed by the available statistics. In section 3, the theoretical FDI model based on the dynamic optimizations of the Japanese ``representative'' multinational firms is built. In section 4, the hypothesis on the determinants of the Japanese FDI in North America in the late 1980s and the early 1990s are presented. In section 5, the empirical methodology based on the theoretical model described in section 3 is offered. In section 6, the empirical result is shown. In section 7, both the empirical result and the acceptability of some assumptions in the theoretical and empirical model are discussed. In section 8, some theoretical and policy implications are presented. Finally in section 9, some concluding remarks are offered.