軍拡競争の理論的考察-計量分析を中心にして-

現代の安全保障

抄録

This paper deals with arms race phenomenon from the view point of mathematical model building. In Japan very few scholar studies arms races quantitatively. By this reason I reviewed mathematical studies done abroad, before presenting my own model and empirical analysis. The outline of this paper is as follows: 1) Brief explanation of Richardson's models. 2) Asymmetry model of mine. 3) Optimal Control Theory. 4) Non-Linear models. 5) Reactive Linkage model. 6) Adding new variables.
I present my asymmetrical model which originated from my graduation thesis of Doctor Course of Hitostubashi University. Whole empirical analysis is presented at Hiroshima conference of Peace Science Society (International), August 18-19, 1978, titled “A Simple Model of Arms Races: Richardson's Model Revisted.” Very briefly stated, I found that original basic model of Richardson lacked explanatory power in USA-USSR arms race, but in Egypt-Israel arms race it had large R². Though basic model could not explain the variance of arms race between USA and USSR, if we introduce Δx/x and Δy/y as economic variable in each equation, we got surprizingly high R².
In the Richardson's model, arms race is conceived as statistically determined, however using optimal control theory, we can conceive arms race as a result of the particular policies of the decision-makers. Though optimal control theory gives us quite interesting interpretation of arms races, empirical analysis shows ambiguous finding: the parameter estimation is contradictory to our common sense of arms races. Much empirical analyses are needed, but I highly evaluate this approach promising.
Non-linear model of Saris, W. E. and C. P. Middendorp which used the number of missiles with saturation level is introduced. The studies of Wallace, M. D. were reviewed critically as a good example of scientific study. Lucier, C. E. advanced the studies of Moll and Dennis, made one variable Organizational Politics model, and explicitely made the parameter time-varying.
Ostrom, C. W., Jr. invented integrated model including arms race model, organizational politics, and bureaucratic politics, but for me this model gives nothing new perspective of arms races.
Two studies of adding new variables are reviewed. Hollist, W. L. introduced economic constraint and technology as new variables, and examined those equations by four pair of nations. Wallace, M. D. found that security problems rather than strategic competition affected the change of arms expenditures between two superpowers.
Lastly but not least important problem is to gather reliable arms expenditure data.