Several alternative explanations of U.S. defense expenditure policy-making behavior have been heralded as theoretically sound and empirically supported representations of U.S. defense expenditure policy making. This has created an atmosphere in which almost as much effort is directed towards criticizing and/or defending existing explanations as is given to improving them through continued theoretical and empirical investigations. To overcome this impasse, I have developed in this article a working synthesis of three current approaches–Arms Race, Organizational Politics, Bureaucratic Politics–based on the insights and observations of Samuel Huntington (1961), Warner Schilling (1962), Colin Gray (1971), and James Rosenau (1971). The resulting formulation, which is labeled the Reactive Linkage Model, characterizes the pattern of U.S. defense spending in terms of an initial response by the armed services to the anticipated USSR expenditure level which is subsequently filtered through the president, Congress, and the Departmen t of Defense to determine the magnitude, scale, and timing of the reaction.
On the basis of the theoretical and empirical analyses here presented, I have reached the following conclusions. First, the Reactive Linkage Model successfully synthesizes the three existing explanations of U.S. defense expenditure policy making into a coherent and plausible whole. Second, the estimated version of the model provides an accurate representation of the pattern of U.S. defense expenditure policy making from 1954 to 1973. Specifically, the individual equations account for a large portion of the variance with parameter estimates that are both sharp and plausible and the model as a whole generates extremely accurate historical forecasts. Finally, I investigate the impact of war, presidential party, and negative public opinion on individual decisions, basing my inquiry on the assumption that the model accurately represents U.S. defense expenditure policy making. The results show that different events have an impact on only some of the steps in the policy-making process and the steps which are so affected vary from event to event. As a consequence of these results, I argue that the Reactive Linkage Model, and hence the idea of combining the key elements from competing explanations, represents a positive step forward in the study of U.S. defense expenditure policy making.