Stephen J. Majeski

Expectations and Arms Races

Recent empirical studies of arms races have produced two puzzling results. First, for several important arms races, particularly the U.S. -USSR case, there appears to be no interaction. Arms acquisition appears to be driven only by internal factors. Second, in several cases, nations appear to be responding to the current military expenditures of their opponent. This statistical result presents interpretation problems since nations do not know the current defense expenditure of their opponent at the time that they make their own military expenditure decision. A number of interpretations for these results have been provided. They include inappropriate model specification, inappropriate operationalization of military threat, and error-ridden military expenditure data. Nevertheless, the results persist and conflict with our more substantive and intuitive understanding of arms processes. A process based upon expectations of current arms expenditures of the opponent is developed to deal with this puzzle. In this paper several alternative expectation mechanisms are formulated and tested for a number of contemporary arms races.

Mathematical Models of the U.S. Military Expenditure Decision-making Process

Mathematical models of the U.S. military expenditure decision-making process are constructed, estimated, and analyzed. Each model contains four interdependent policymaking groups. Each group is assumed to have different goals and objectives. These objectives are formalized and are assumed to be a function of actions of other policymakirg groups, relevant fiscal and economic constraints, expectations of actions of other policymaking groups, and the perceived external threat. Given posited objectives, and a decision calculus by which expectations are formed adaptively and partial adjustments made by the groups to their goals, decision rules describing behavior are derived and estimated. Results suggest that policymakers do employ expectations and do make partial adjustments to operational goals.

A Methodology for the Study of Historical Counterfactuals

Counterfactual reasoning is a component in much historical and political research. A proposal for exploring counterfactuals is elaborated, based on philosophical work on modal logic and possible worlds semantics. It is proposed that phenomena have essences which are unchanging in all possible worlds and that counterfactual analysis consists of making inferences about the contingent properties of these phenomena. Essential properties can be expressed as contingent relations bound, in different counterfactual situations, to different contingent properties. This methodology is applied to counterfactual explorations of a particular phenomenon: the “winnability” of high-level United States foreign policy recommendations. In two cases, the question is asked of whether “harder line” U.S. policies regarding Vietnam would have been adopted. Using the methodology elaborated in the first half of the article, it is found that as early as 1961, recommendations for the overt use of U.S. ground combat troops could have been accepted.

Arms races as iterated prisoner's dilemma games

Abstract Many important aspects of arms races are captured by the iterated prisoners dilemma (IPD) game. In this analysis, strategies are developed and analyzed for IPD games which most resemble arms races; those which are finite, uncertain and time varying in length. It is shown that, under certain conditions, it can be rational to unilaterally cooperate following a series of mutual defection. A number of conclusions concerning arms race behavior are drawn.

Simple Choices and Complex Calculations

Bruce Bueno de Mesquitas The War Trap (1981) is an ambitious attempt at putting forward a rational choice theory of war initiation. The book is subjected to a critical reading and found wanting on three criteria: concepts, measures, and empirical analysis.

Agent Mobility and the Evolution of Cooperative Communities

INTRODUCTION W e are concerned with whether and how cooperation evolves in social worlds characterized by the presence of selfish agents engaged in repeated relations without central authority. This fundamental problem has been the focus of many studies across all the social sciences as well as philosophy, biology, and computer science. Based on the seminal works of John Maynard Smith [1], and Robert Axelrod [2], the iterated prisoner’s dilemma (IPD) has become the central metaphor for the evolution of cooperation in populations of selfish agents without central authority. To investigate whether and how agents in an IPD can overcome the individual rational choice to defect to achieve the socially optimal outcome of mutual cooperation, we construct an artificial world and then analyze characteristics of that world via simulation. The artificial world that we construct is an anarchic, competitive place populated by agents with limited cognitive and social capabilities whose resources are being constantly drained. Into this harsh environment we introduce agent movement and constrained interactions in the form of spatial neighborhoods, features that we anticipate will foster cooperation. We are interested in several properties of this artificial world. Does it become highly cooperative? If cooperation emerges, is it stable? Do agent movement and the type of movement facilitate the generation of a stable cooperative world? Are there patterns in the kinds of strategies that help generate a cooperative world or in the kinds of strategies that evolve in a cooperative world?

A rule based model of the United States military expenditure decision‐making process

I argue that the theoretical foundations of the U.S. military expenditure decision‐making (MEDM) tradition, incrementalism and bounded rationality, suggest that policymakers behavior is rule based. I also argue that the standard way that these rules are formalized, outputs are generated by a linear and additive combination of the base and relevant informational variables, is neither descriptively accurate nor consistent with the basic assumptions of procedural rationality. A rule based model constructed of a set of hierarchically and conditionally organized “if‐then” rules is developed as an alternative. Models of this form provide an opportunity to develop and rigorously test more descriptively accurate representations of how individuals and organizations make decisions and which capture hierarchical and conditional relations of the MEDM process. An empirical analysis of the model demonstrates both its explanatory and predictive power.

The role of expectations in arms acquisition processes: The U.S.‐U.S.S.R. case

Recent empirical studies of arms races have led to the disturbing conclusion that for several important arms races, particularly the U.S.‐U.S.S.R. case, there appears to be no interaction. Instead arms acquisition appears to be driven only by internal factors. A number of interpretations for these results have been provided. They include, inappropriate model specification, inappropriate operationalization of military threat, and error ridden military expenditure data. Recent results, coupled with our more descriptive understanding of arms processes, suggest that expectations of current and future arms expenditures by the opponent are relevant to an understanding of current arms expenditures. This implies that such expectations must be taken into account in the arms modeling process. In this paper several alternative expectations mechanisms are formulated and tested for the U.S.‐U.S.S.R. case. The extent of measurement error in Soviet military expenditures and its effect on the empirical tests of the expec...

How foreign policy recommendations are put together: A computational model with empirical applications

This paper presents and applies empirically a computational model of the way in which bona fide high level foreign policy recommendations by U.S. policy makers are assembled. We begin by pointing out that policy making can be seen as the connection of certain strings of words to other strings. We then discuss how these connections constitute certain types of foreign policy making phenomena as such. To theorize about such connections, one first needs to specify essential features of these phenomena, and we do so for one phenomenon: bona fide recommendations. We next turn to a discussion of the theory that links together the categories by which these features are represented. That theory explains how certain strings of words are assembled into new proximate goals, missions, and tools. The theory can be modeled computationally using the programming language Scheme, and we next present that model. We conclude by presenting a run of the model, showing the close fit between actual and generated strings.

Technological Innovation and Cooperation in Arms Races

Cooperation in the form of either tacit or formal agreements is a relatively rare event in both contemporary and historical arms races. Yet, analysts of most arms races suggest that all the participants usually can benefit from mutual cooperation. In this analysis, arms races are represented as Iterated Prisoners Dilemma games of finite but uncertain length. Based on this plausible formulation, the conditions when mutual and/or unilateral tacit cooperation is ‘rational’, in a game theoretic expected payoff sense, are examined. Then the dynamic effects of technological innovation and formal arms control agreements on the payoff structure of the IPD game are introduced. Given these new conditions, the possibilities for cooperation are reexamined. This initial exploration permits a number of tentative conclusions about the conditions under which it is rational to cooperate, and why cooperation is so difficult to achieve in arms races.