Read an Excerpt
Change and Stability in Foreign Policy
The Problems and Possibilities of Détente
By Kjell Goldmann
PRINCETON UNIVERSITY PRESSCopyright © 1988 Princeton University Press
All rights reserved.
Foreign Policy Stability as a Research Problem
We generally assume that there are patterns in the foreign policy of nations and not just single acts. A knowledge of the pattern — the "policy" — of an actor is assumed to be useful for explaining and predicting actions. If we can show that an action fits into a pattern — that is, that the actor behaves as he usually does, or says that it is his policy to do, we have in one sense explained his action. Similarly, if we know the pattern, we may anticipate what the actor will be likely to do in the future. We know, in other words, about a regular feature of international politics — a feature that may, however, be more or less amenable to change.
The assumption of foreign policy patterns is easy to justify. The arguments commonly put forward range from the imperatives of geopolitics to the standard operating procedures of bureaucracies to the inertia of belief systems. However, analysts also find it easy to explain new departures after the fact by reference to, say, the dynamics of international relations, the flux of domestic politics, or the erratic nature of personalities. Such explanations undermine the very assumption that foreign policy is patterned — unless it can be explained why this particular policy was vulnerable to that particular disturbance.
There is an inevitable tension between viewing international politics as the pursuit of policies and seeing it as variable responses to shifting situations. Yet the tradition in foreign policy analysis is to do both. It was to be expected that the United States and China would remain enemies even after the Sino-Soviet rift: their mutual enmity had deep roots in both countries. It was also to be expected that they would come together as a result of the new triangular balance of power. But if both were to be expected, how can either be explained, and how could either have been predicted? What factors determine whether, when, and to what extent pressure for change in a policy will in fact produce change?
This is an unresolved issue in foreign policy theory. It relates to three key problems in the analysis of international politics:
1. The problem of adaptation. On the one hand, nations — and organizations in general — are under pressure to adapt to changing conditions in their environment. On the other hand, they have a tendency to stick to their previous policies. What determines the impact of environmental change?
2. The problem of learning. On the one hand, nations — and organizations in general — take into account the way in which the environment responds to their policies; negative feedback is a pressure for change. On the other hand, there is a tendency to continue as before. Inter-nation action-reaction sometimes operates and sometimes fails to operate. What factors account for the impact of feedback?
3. The problem of domestic change. On the one hand, shifts in domestic politics — and within organizations in general — may place new people in positions of power, people whose views differ from those of their predecessors. On the other hand, new people may find themselves to be the prisoners of past policies. What factors determine the sensitivity of a foreign policy to domestic change?
These are three common themes in controversies over foreign policy. How will U.S. policy be affected by long-term changes in the international distribution of power? Would Western restraint have an impact on Soviet arms policy? Is West German foreign policy dependent on who wins the next election? Nations are assumed to pursue identifiable, stable policies. At the same time, a number of factors are assumed to be sources of change in policy. The issue concerns the likely outcome of the confrontation between the two. Foreign policy theory is weak on this point. It is uncommon among foreign policy theorists even to make a strict distinction between pressures for, or motivators of, change and pressures for, or motivators of, lack of change. The theoretical sketch to be presented here is meant to improve our understanding of the stability and instability of foreign policies, and to do it by taking environmental change, negative feedback, and shifts in leadership for granted and focusing instead on the factors accounting for their varying impact.
The term theoretical sketch is deliberate. What follows is theory — it contains not merely definitions but also propositions about the ways in which certain phenomena affect certain other phenomena. It is a sketch in three ways, however: its concepts are imprecise, its propositions are weak, and it has not been exposed to a systematic empirical test. A theoretical sketch is a necessary intermediate step between impressions and theory proper. It paves the way for more sophisticated inquiry. It also, however, serves, in the absence of theory proper, as a tool for asking better questions — as a systematic checklist for the analysis of specific problems.
Such a checklist can be useful. If none exists, the analyst has the option of doing the work ad hoc — that is, without considering the meaning and relevance of the concepts. The analysis risks being overly narrow, biased, or prejudiced as a result; whatever comes to mind at the outset is assumed to be the proper object of study. Or the analyst may make his or her own theoretical preparations. This is a big task, however, and it may be out of the question for the analyst who is concerned with a specific case or an immediate problem rather than with the development of theory; those who need to consume theory cannot always produce it themselves. Now, if a theoretical sketch exists, there is no need to choose between being arbitrary and beginning from scratch. Theory proper is better, but a sketch is better than nothing at all. This is the way in which the sketch that is about to be presented may contribute to the analysis of specific problems of foreign policy stability — as a reasonably systematic, considered, and comprehensive list of matters to be taken into account.
The notion of a theoretical sketch is further discussed in the Appendix. A problem with theoretical sketches is how to evaluate and improve them. They are not always strictly testable, especially not by historical case study. Nor are they arbitrary, however. The Appendix includes some thoughts about the way in which a theoretical sketch may be improved by empirical application, and some improvement of the present sketch is attempted in Parts Two and Three of the book, where it is applied to the analysis of East-West relations and the problem of détente.
The outline of Part One is as follows: First, in the present chapter, foreign policy stability is introduced as a problem of research. In section 1.1a conceptual framework for the analysis of change and lack of change in foreign policy is suggested and the concept of a "stabilizer" is introduced. This conceptual discussion specifies the problem with which the book is concerned. Section 1.2 is devoted to some of the literature expected to be concerned with foreign policy stabilizers. Against this background, an inventory of foreign policy stabilizers is made in chapter 2, and the processes of foreign policy stabilization and destabilization are briefly considered in chapter 3.
1.1. The Framework of Analysis
The basic structure of the framework of analysis is shown in Figure 1.1. A change in policy has "sources," but sources do not produce policy changes directly. Sometimes pressure for change produces change, but sometimes it does not. That is the problem of stability considered here. Our concern is not with the sources of change in foreign policy but with its stabilizers.
Thus, two intervening variables, so to speak, are assumed between sources and policies. A source of policy change is an event tending to start a process of policy change. A stabilizer is a variable affecting (1) the likelihood that such an event will set a process of change in motion, and/or (2) the extent to which a process of change will be completed and produce a change in policy. The chief questions are: What are the stabilizers, and how do they operate?
A more detailed view of the framework of analysis is set out in Figure 1.2. Three sources of policy change are shown. One possibility is that a change in policy is brought about by a change in the environmental circumstances called "conditions." This is what may be called a process of adaptation. When the United States revised its policy toward the People's Republic of China in the late 1960s, this may be seen as adaptation to a situation in which China was no longer an ally but an adversary of the Soviet Union. Policies may also be their own sources of change in the sense that they may change in response to negative feedback. This will be called learning. A typical case would be the abandonment of a conciliatory policy because the response had been nonconciliatory. Another would be the abandonment of a forceful policy toward an adversary because such a policy had been found to render one's allies uneasy. Some instances of policy change are neither adaptation nor learning, however, and there is therefore a need for the category of "residual factors."
Three processes of change are also distinguished in Figure 1.2. Each policy is assumed to be based on a set of ideas on the part of the policymaking system, and each change in policy is assumed to be based on a change in ideas. The latter change may, however, be of three kinds: rethinking by individuals within the policymaking system (represented in Figure 1.2 by the direct arrows from "conditions," "policies," and "residual factors" to "ideas"); a change in the composition of this system; or a change in the balance of power between the members of the system.
In the rest of this section, the concepts and relationships in Figure 1.2 will be discussed in more detail.
Policy It is commonplace to point out that foreign policy research has been imprecise about its dependent variable (Faurby 1976: 141–44, Hermann 1978: 25–32). The same may apparently be said about so-called policy studies. It is emphasized in this literature that the meaning of "policy" is unclear and that "policy" is a "generic symbol pointing toward a field of interest, rather than a scientifically precise concept." A consensus has been suggested to obtain among foreign policy analysts to the effect that foreign policy is "a set of goals, directives, or intentions, formulated by persons in official or authoritative positions, directed at some actor or condition in the environment beyond the nation state, for the purpose of affecting the target in the manner desired by the policymakers" — but this, it has been added, is not to have said very much (Cohen and Harris 1975: 385).
Against this background it has appeared necessary to be relatively precise about what is meant by "policy" in the present study — that is, what the theory about to be sketched is meant to be a theory about. A consideration of this matter now may inhibit confusion later on. An attempt to define the concept of policy in a precise fashion has been made by Kerr (1976), and her definition is useful as a point of departure. Kerr suggests that four conditions are necessary requisites for a public policy and implies that together they are sufficient. Her first condition is:
1. An agent must intend to act in accord with some conditional imperative X, i.e., whenever conditions beta obtain, do action(s) alpha.
Hence, a policy requires both an agent and a conditional imperative. Without an agent, there is no policy, but "some sort of principle." A policy has a specific subject and is not a universal norm.
Kerr then distinguishes between "policy" and "promise." Whereas a promise is specific to one or a specified number of instances, a policy is general:
2. The agent must perceive as likely that the conditions beta will occur more than once or a restricted number of times.
There is, however, one more difference between policies and promises: they are not "equally revisable." To change a promise is to break it. A change in policy does not carry this connotation; "the agent has simply to revise the conditional imperative." Hence this condition:
3. The agent may substitute conditional imperative X' for conditional imperative X, without violating conditional imperative X.
Kerr finally distinguishes between public and private policy by adding this criterion:
4. The agent must declare to the relevant public the agent's intention to act in accord with the conditional imperative X.
The problem of identifying the "relevant" public is extensively discussed by Kerr but is unimportant here. From our point of view, the chief problem is that Kerr defines policies as programs and not as behavior patterns. In foreign policy research, the opposite position is sometimes advocated; the Comparative Research on the Events of Nations (CREON) project has decided "to conceptualize the external actions of national governments in terms of behaviors rather than goal-seeking policies" (Hermann 1978: 32). However, programs are not only very close to what is usually meant by policies but are important objects of study because they can have important consequences. The chief question is whether policies are necessarily programs: major change has occurred in Soviet action on the high seas, but we cannot (for the sake of the argument) find any Soviet document substituting, in Kerr's terms, conditional imperative X' for conditional imperative X. Can we talk of a change in Soviet naval "policy" nevertheless?
One possibility is to assume a change in program behind a change in behavior; Kerr's fourth condition is not met, but the three others are. Allison (1971) and others have demonstrated the questionability of ascribing intentions to complex actors on the basis of their behavior, however. Another possibility is to regard a behavior pattern as a policy regardless of whether it has been the agent's intention to pursue such a policy. In our approach, we shall distinguish between verbalized policy, defined roughly as a publicly declared program, and nonverbalized policy, defined without reference to any program on the part of the agent.
The following definitions are a starting point:
Dl. Policy: an agent's line of action with regard to an object.
D2. Line of action: the agent does alpha whenever situation beta occurs; beta will occur more than once or a restricted number of times.
D1 plus D2 is a variation on Kerr's conditions 1 and 2; the "object" has been added to Dl in order to make clear that a policy concerns either a specific issue area (for example, "trade policy") or the agent's relation to some other actor (for example, somebody's "China policy"). Kerr's condition 3 is unacceptable for the purposes of this study, however. It is misleading to distinguish between promises and policies. Policies in the present approach are in fact promises, commitments.
The essence of Dl plus D2 is that a policy is attributed to an agent and that it consists of a line of action. The subclass "verbalized policy" may be defined as follows:
D3. Verbalized policy: a line of action that an agent declares he is following or intends to follow with regard to an object.
A verbalized policy is identified by — and only by — the study of the agent's declarations. D3 refers to the policy as described by the agent. Alpha and beta are specified by the agent himself.
Nonverbalized policy is a more difficult concept:
D4. Nonverbalized policy: a line of action that is in fact followed by an agent with regard to an object.
In this case, alpha and beta are specified by the observer and not by the agent. Moreover, whereas a verbalized policy can apply to the future in the sense of expressing intentions, a nonverbalized policy is by definition a generalization about past behavior.
Verbalized and nonverbalized policies are different phenomena. They can exist independently. There may be a causal relationship between the two, however. Programs may lead to behavior, and behavior may lead to programs. The correlation is unlikely to be perfect, but it is also unlikely to be zero.
An analysis of policy stability presupposes a definition of policy change. This concept is considered by Rose (1976: 14–23). However, his variable is "progress in terms of a policy objective"; he is thus concerned with a change in outcome rather than in policy as defined here. In terms of the present definition, a change in policy with regard to a given object can only mean a change in alpha or beta: either a new act in a given type of situation or a given act in a type of situation previously associated with a different act. Stabilizers inhibit such changes.
Excerpted from Change and Stability in Foreign Policy by Kjell Goldmann. Copyright © 1988 Princeton University Press. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.