This paper analyzes influence of the budgeting bureaucrats on the size of a nation's budget and welfare of consumer-citizens.
It models a nation's budget making process as a bargaining process between the bureaucrats in the line-ministry who wish to expand the size of their ministry's budget and those in the budgeting bureaucrat who wish to cut the size of their budget. It finds that while bureaucrats in the both sides motivated to maximize their own private benefits, their struggle for a larger share of the surplus arising from provision of government services generates unwittingly, as if they were guide by an invisible hand, benefits to the consumer-citizens. Our model predicts that the size of government expenditures would be restrained to a size smaller than those predicted by the models of Niskanen (1971) and Migue-Belanger (1974), and the consumer-citizens would receive a part of the surplus accruing to the nation, while Niskanen and Migué-Bélanger claimed complete bureaucratic absorption of the surplus.
The paper also, compares Japan statistically with the United States and infers to what extent the Japanese budgeting bureaucrats use revenue estimates as a strategic variable in directing the nation's budget to their target size so that it maximizes benefits of the budgeters.

View full abstract

It is often said that the SNTV that Japan applied for Lower House elections strongly induces LDP Legislators to provide porkbarrel legislation for their constituents in order to win. If this is a unique feature of the system, we should see this phenomenon only among the Legislators but not among the LDP Councilors elected through different systems.
This paper clarifies through game theoretic models the relationship between the electoral systems and the strategies that the candidates should take for winning. The models suggest that providing porkbarrel legislation is the strategy best suited to reach optimal equilibrium under the SNTV, while providing more public interest legislation is the winning strategy under the Upper House electoral systems, i. e., the FPTP and the Party List PR.
The data analysis comparing the involvement of the Legislators and Councilors in the PARC as proxies for their strategies indicates that the Legislators and the Councilors under the Party List PR behave as the models predict. However, the latter under the FPTP behaves differently from what the models lead us to predict: Councilors under the FPTP also provide porkbarrel legislation in order to win.
Although this odd result can possibly be explained within the models, there might be a bug in the logic. We need further development of the models and of the data analysis to grasp the relationship between the electoral systems and the strategies of the candidates.

View full abstract