The main purpose of this paper is to investigate the effects of bureaucracy on macroeconomy in a two-country model of endogenous growth with overlapping generations. In county 1, it is assumed that budgetary policies are determined by bureaucrats. However, no political factor in country 2 is considered. By the way, in Japan, bureaucrats have more influence on the determination of budgetary policies than politicians, and vice versa in the U.S. Therefore, if country 1 is regarded as Japan and country 2 as the U.S., the twocountry model in this paper is very useful in analyzing how Japanese bureaucracy affects Japan-U.S. economic relations.
Considering the characteristics of Japanesebureaucracy, the budget decision-making process in country 1, with country l's public capital to GDP ratio raised, is constructed as follows There are two types of bureaucrats. One type aims to increase country 1's government consumption to GDP ratio, and the other type to rise country l's growth rate. The former-type bureaucrat is consistent with the budgetmaximizing bureaucrat analyzed in Niskanen [1971] . The latter-type bureaucrat, which plays an important role in Japan, hasn't been investigated in the economic theory of bureaucracy. The two economic variables in country 1 can be obtained as the asymmetric Nash bargaining solution between the two types of bureaucrats.
In spite of the rise in country 1's public capital to GDP ratio, which can improve economic welfare, an inefficiency caused by bureaucracy becomes the greater the stronger the former-type bureaucrat is, in that the increase in the overall utility of each generation in both countries can be reduced in case of the rise in country 1's government consumption to GDP ratio. However, the former-type bureaucrat contributes to the reduction of foreign assets in country 1. In view of Japan-U.S. political relations, it is desirable that the foreign assets in Japan should be decumulated. Thus, the former-type bureaucrat is appreciated as regards the political relations, but is not in terms of the economic relations. In conclusion, if we consider Japan-U.S. relations in the long-run, it is required that the bargaining power of the latter-type bureaucrat should be intensified in the budget decision-making process.

View full abstract

Milton Freedman (1955) proposed voucher system in education as one means for improving public system of compulsory education. It introduces a market principle into public educational system. And since then there have appeared many studies on the voucher system in the U. S. (Jencks (1971), Coons (1978) ), the U. K. (Wiseman (1959), Peacock (1966) ) and Japan (Ichikawa (1982), Kurosaki (1989) ) . But they have lacked for economic and financial analysis of the system. The purpose of this paper is to try to make good the defects.
The main results are the following:
1. We verify the superity of private to public system by comparing the educational outputs of the two system on the basis of an estimation of a production function of educational service. We found the difference between behavioral objects of schools is the most significant factor affecting the educational output. From this result and a hypothesis that teachers are economically rational, introduction of market mechanism into the public system can be justified.
2. We found, by regression analysis, that main determinants of educational expenditure per pupil are class member and teacher's salary.
3. The financial possibility of voucher system in Japan depends on how to finance land. Perfect self-supporting system can work even within existing educational budget, if land price is less than \190, 000/m².
4. We must break through political barriers in order to introduce voucher system. We must admit that we face a well-known difficulty, as Olsonian theory says.

View full abstract