|
|
|
|
CIAO DATE: 7/99
Interests, Schminterests: Precision in Descriptions of and Assumptions About State Motivations *
Department of Government
Cornell University
International Studies Association
40th Annual Convention
Washington, D.C.
February 16–20, 1999
Abstract
Occasionally, terminological problems reflect underlying conceptual confusions that can seriously impede the advancement of discourse and understanding. This paper argues that two such problems currently exist in much of international relations scholarship, focused around the concept “interests.”
The paper argues first that because states may be motivated by a variety of impulses, “interests” and “preferences” must not be considered synonyms. It reviews historical arguments to this effect, discusses examples from recent literature illustrating misuse of the terms, and examines the relationship of this terminological problem to both the descriptive and theoretical sides of the interest-based model of state behavior.
Secondly, the paper examines the distinction between “interests” and “behavior,” an important distinction in recent theoretical debates. It argues that the characterization of a goal as an “interest” rather than a “behavior” depends not on the goal itself but rather on how broadly one defines the particular game or study in which the characterization takes place.
Note to readers: This paper is a modified version of a chapter of my dissertation, Samaritan States: International Altruism, ldentify, and State Motivations.I have attempted to make it self-standing, and apologize for any confusing references to other parts of the dissertation that I may have failed to catch and excise. Readers are encouraged to forward their thoughts, comments, criticisms, and suggestions to me at the address on the cover page.
The concept “interests” is a basic and central concept in international relations scholarship, and one would think that it therefore would be well defined and consistently employed in scholarly literature generally. This paper demonstrates that such is unfortunately not the case. “Interests” is used in a number of different ways, and often used (inadvertently, apparently) in more than one sense within the same work. The result is often a lack of clarity, confusion, and a frequent smuggling in of assumptions.
I address conceptual problems in the literature with full awareness that often, arguing about concepts amounts to a lot of fuss about distinctions that in the end are not terribly important. As John Hall warns,
“it is wise to remember Poincaré’s aphorism: Whereas natural scientists discuss results, social scientists argue about their concepts. The presumption that it points to—that once conceptual and technical tools are really sharp, understanding social life will be easy—is surely so questionable as to induce a healthy skepticism toward the very enterprise of paradigm generation.” 1
Yet attention to conceptual precision does not necessarily reflect the presumption Hall points to, that sharp concepts will make understanding social life “easy.” It may rather reflect the hope that sharp concepts will make the difficult task of analyzing an extremely complex reality at least plausible rather than confused and near impossible.
It is in that spirit that my conceptual arguments are presented. Occasionally, terminological problems reflect underlying conceptual confusions that can seriously impede the advancement of discourse and understanding, impeding communication both among scholars and between scholars and practitioners and the general public. I argue in this paper that a set of such problems currently exists in much of international relations scholarship, focused around the concept “interests.”
The problems with the usage of the concept, I argue, are relatively straightforward, but significant. They hinge on two issues. The first is simply definitional: what does “interests” mean, and what should it mean, i.e. what understanding of the term is most useful for social scientists? The second is methodological and analytical: what should be assumed, if anything, about what drives state behavior? These issues are explored in the following through examination of the relationship of the concept “interests” to two other concepts: “preferences” and “behavior.”
A. Interests versus Preferences: Self-Interest and Alternatives
Definitions
The term “interests” has more than one meaning or sense. In one of those senses, interests are understood to be related to or synonymous with advantage, benefit, or profit for some actor. Interests in this sense can be defined as goals or conditions believed to be positively related to an actor’s own wellbeing. I will refer to this sense of the term as interestsw, with the subscript W standing for “wellbeing.” 2
Our wellbeing is not the only thing that determines or influences what goals we set, however—we also set goals related to or influenced by our passions, morals, tastes, and curiosities, for example. We therefore need another general term to refer to the set of all our different goals, based in our wellbeing or in something else, together. The term that seems to serve this function best is “preferences.” Interests—understood as interestsw, goals or conditions believed to be positively related to an actor’s own wellbeing—are thus just one subset of all possible preferences.
These definitions of interests and preferences are more or less conventional. Unfortunately, other senses of‘interests’ are also conventional. For instance, sometimes the term‘interest’ is used to denote something an actor is simply “interested in” or concerned with or about. The sense here is not one of benefit but rather one of a mental or intellectual preoccupation or attention: my interest in (i.e.curiosity about) the forces behind disaster relief (my dissertation topic), for example. I am pursuing that “interest” through this project, but I do so more from curiosity than from an expectation that the answer will directly affect my wellbeing. I will refer to this sense of the term as interestsA, with the subscript A standing for “attention.”
Practically the same set/subset relationship obtains between interestsw and interestsA as was mentioned above between interests and preferences: generally things that affect our wellbeing—“interestsw=wellbeing” —are things we pay attention to and are interested in—“interestsA=attention”—but not all things that draw our attention (interestsA) necessarily do so because they affect our wellbeing (interestsw). 3 Put another way, we are usually quite concerned with goals related to our wellbeing, but we may also be concerned with goals or things not related to our wellbeing. Just as interestsw are a subset of preferences, as shown above, interestsw belong to, or are a subset of, our interestsA. Our interestsAhowever, include things other than our interestsw. One has only to imagine the previous two sentences without the subscripts identifying the different senses of the term to see the confusion that can result from the interchanging of these different meanings of the term.
History
Discussion of how broadly the term “interests” should be understood is nothing new. Hirschman traces the history of the term, and notes the problem of tautology that comes with too broad an interpretation of it, as follows:
“....the concept has stood for the fundamental forces, based on the drive for self-preservation and self-aggrandizement, that motivate or should motivate the actions of the prince or the state, of the individual, and later, of groups of people occupying a similar social or economic position (classes, interest groups). When related to the individual, the concept has at times had a very inclusive meaning, encompassing interest in honor, glory, self-respect, and even afterlife, while, at other times, it became wholly confined to the drive for economic advantage. Correspondingly, “pursuing one’s interests” can cover—to the point of tautology—all of human action while it will more usefully designate a specific manner or style of conduct, known variously as “rational” or as “instrumental” action.” 4
A criticism of the tautological and thereby unuseful aspects of an overbroad sense of the term comes from as far back as 1829, when T. B. Macaulay wrote the following as part of his response to James Mill’s “Essay on Government”:
“One man goes without a dinner, that he may add a shilling to a hundred thousand pounds: another runs in debt to give balls and masquerades. One man cuts his father’s throat to get possession of his old clothes: another hazards his own life to save that of an enemy. One man volunteers on a forlorn hope: another is drummed out of a regiment for cowardice. Each of these men has. no doubt, acted from self-interest. But we pain nothing by knowing this, except the pleasure, if it be one, of multiplying useless words. . . it is . . . idle to attribute any importance to a proposition. which. when interpreted. means only that a man had rather do what he had rather do.” [underline added] 5
As the first underlined passage indicates, for Macaulay “men always act from self-interest.” 6 Whether this indicates that he has a narrow understanding of human motivations or a broad understanding of “self-interest,” (on the order of preferences or interestsA rather than interestsw) the result is the one he presents: when all behavior is understood to flow from interests, the term “interests” loses much of its usefulness as an explanatory term. As Hirschman puts it, “to explain human action by interest thus [becomes] the vacuous tautology denounced by Macaulay.” 7
Hirschman suggests in the passage quoted earlier that the phrase “pursuing one’s interests . . . more usefully designate[s] a specific manner or style of conduct, known variously as “rational” or as “instrumental” action.” 8 Yet such an understanding of “interests” would add yet a third to the two already discussed, for it refers as he says to “a style of conduct” rather than to the nature of the goal or condition being pursued, which is what distinguishes the first two senses discussed above. By this definition, pursuit of any goal—whether based on benefit, morals, love, hate, revenge, supernatural millenarian instructions, or what have you—would be considered “pursuit of interests” as long as it was pursued instrumentally. While such an approach does somewhat reduce the tautology problem Macaulay and Hirschman point out, it fails to eliminate it completely by keeping the term equally applicable to all types of goals.“Pursuing interests” becomes simply“instrumentally pursuing whatever we want to pursue.” 9
I therefore argue here for the following usages:
Illustration
An example may help to clarify matters at this stage. Suppose there are three countries (let country = society + territory = state + population + territory) all in the same region. A large mountain lies near to the borders of all three of them—where exactly is unimportant. All three want possession of the mountain.
All three of these societies are “interested in” this mountain. Each has as a goal possessing the mountain—that is one of their preferences. If an “interest” is understood as interestA, something about which we are concerned or pay attention to, we can say that possessing the mountain is among the “interests” of each society. While accurate, however, this is to add nothing to the first two sentences of the paragraph. If however, “interest” is understood as interestw, as a goal or condition related to wellbeing, then we can say that possession of the mountain is among the interests of only one of the three societies, the industrialized one. The other two are not pursuing their interests, and indeed may be hurting their interests, by working for possession of the mountain. They are rather pursuing what they perceive as their sacred and ethnic/cultural duty, respectively. 10 Reserving “interests for the narrower interestw sense thus allows us to use it to more precisely describe the situation, making distinctions that might otherwise be blurred, and that would probably have real implications for behavior. Citizens of the religious country may feel it would be an honor to die attempting to take or to defend the mountain. Citizens of the industrial country may be somewhat less eager for such a fate. Using “interests” in the broader, tautological sense, interestsA, to say that all three societies want control of the mountain because of their interests, hides such a difference in motivation.
One might be tempted to simply add adjectives such as “religious” or “ethnic” or “material” to “interests.” To say that people carrying Out a sacred duty are pursuing their “religious interests,” however, is to try to make the concept serve where it fits uneasily. Likewise to use “ethnic interests” mischaracterizes the cultural motivation depicted above: it is more the honor than the wellbeing of the ethnic group that is held to be at stake. The distinction between the industrial and the religious and ethno-cultural motivations is more clearly captured, and mischaracterization and confusion better avoided, by restricting “interests” to the one motivation and using “duty” or some other more appropriate term for the others.
Granted, the example is somewhat contrived, and clearly the motives of the three societies could be clearly and precisely described without once using the term “interests” at all. But the fact is that the term is used, and often, in international relations scholarship, and indeed is put at the center of many theoretical debates. If it is going to be used, it ought to be used consistently and in such a way that it clarifies and advances knowledge and debate, rather than blurring it.
As the following examples from the IR literature show, this has often not been the case. The first example illustrates uses of “interests” and related terms in the overbroad sense that leads to tautology. The others show the ambiguity, confusion, and contradictions that can result from the mingling of different senses of “interests” within a single work, and within the discipline as a whole.
Examples from the literature
This first example comes from Robert Keohane’s chapter, “Empathy and International Regimes,” in Jane Mansbridge’s volume Beyond Self-interest, and it illustrates excessively broad understandings of “interest” -related terms such as “egoism” and “self-interest.”
The first problem is Keohane’s understanding of the term “egoism.” He writes the following:
“To relax the assumption of egoism by drawing a sharp distinction between egoism and altruism, however, would confuse the issue. . . since apparently altruistic behavior can always be reinterpreted as egoistic: the “altruist” may be seen as preferring to sacrifice herself rather than to violate a principle or see someone else suffer. 11 [underline added]
This passage indicates that for Keohane, to act as one prefers is to act egoistically, even when the preferred action is “sacrifice.” By definition, then, with this approach, any voluntary action is egoistic, even when alternatives are limited such as in the kind of situation Keohane describes. As Macaulay and Hirschman have argued, such an understanding of the term would seem to broaden it so far as to blunt its analytical usefulness, to say the least. 12
Essentially the same problem is evident in a subsequent passage, though here with the term “self-interests.” Keohane writes that in situations of empathetic interdependence—characterized by interest in the welfare of others for these others’ own sake, even when this has no effect on our own material wellbeing—
“self-interests would by no means have disappeared. Rather, they would have been redefined so as to depend on the welfare of others being realized as well.” 13
The welfare of the “others,” then, though having no effect on our own material wellbeing, would have been “redefined” into our self-interests. In essence, we would have changed or added to what we want, what we would be happy about, and what we would pursue. We would therefore have changed our preferences, thereby also redefining or adjusting our psychological interests (assuming these are linked to preference satisfaction). In effect, then, what Keohane does is include happiness, preference satisfaction, and psychological wellbeing, in “self-interests.”
At first glance that seems a plausible thing to do, (provided one is clear and explicit about doing it). It substantially enlarges the meaning of “self-interests,” however: with preference satisfaction included in self-interest, self-interests would never“disappear”, would never be acted against, unless someone acted against their own preferences, i.e. involuntarily. As with “egoistic” above, then, the term “self-interest” loses much of its analytical usefulness when used so broadly: with these usages, to say that behavior is egoistic or done out of self-interest is to say basically, as Macaulay said, that “a man had rather do what he had rather do.” 14 They thus lead to the tautology that Hirschman warns against.
The second example illustrates a mingling of different senses of “interest” in one work. Stephen Krasner, in Defending the National Interest, chooses to define the “national interest” inductively, writing that the preferences of central decisionmakers can be called the national interest if:
He goes on to discuss how broader U.S. foreign policy objectives shifted from “interest-based” (economic or strategic) before WWII (except 1912-20) to “ideological” after WWII. During this latter period, he argues, US leaders were “moved by a vision of what the global order should be like that was derived from American values and the American experience. . . . They were more concerned with structuring the international system and the domestic politics of other countries than with pursuing readily identifiable economic and strategic interests.” 16 He writes that the pursuit of ideological objectives was done in nonlogical ways, involving misperception and exaggeration of the importance of Communist elements, and with no clear calculations about means and ends. 17 “Their ideological objective of defeating communism assumed a millennial coloration,” he writes of American leaders. “It often appeared that there were no limits to the costs they were willing to incur to achieve this goal.” 18
Despite all this, Krasner labels these objectives “the national interest,” since they meet his standards of being “general” objectives and of persisting over time at the same level of priority. Krasner thus preserves the traditional phrase “the national interest” while pointing out that it may be based not in interests (meaning interestsw) but in ideology—normative ideas about right and wrong ways of ordering societies. He ends up arguing, then, that “the national interest” after WWII was not “interest-based.” Needless to say, the terminology is less than helpful here: when “interest” is not “interest-based,” something is awry, tradition notwithstanding.
Clearly Krasner (and others who use national interest similarly) intermingles the two different senses of “interest” discussed above. He defines “national interest” as “preferences of central decision-makers,” whether interestw-based or not, making it closest to interestsA, basically ‘what the nation/state is interested in.’ The term “interest” in “interest-based,” on the other hand, seems to refer to interestw. 19
Translated, then, the argument that “the national interest wasn’t interest-based” becomes something like “what the nation/state was interested in wasn’t directly related to its wellbeing.” That argument—that some important state behavior has been ideology-driven and non-logically pursued—is an important one. Krasner succeeds in getting that point across despite the terminological problems because he is careful enough in defining his terms that his meaning is understood: we understand that here, “interest” when preceded by “national” doesn’t really mean interestw but rather “preference.” 20
Clearer terminology, however, would obviate the need for “translation.” Why would Krasner preserve the conventional term? It seems that the main reason for sticking with if national interest” is just that: convention. Part of Krasner’s whole objective, however, is to point Out that in fact the conventional deductive approaches, that use “the national interest” to mean objectively determinable material interests, fall short. It is unclear why he would want to graft old terminology, with all its accumulated and unsatisfactory baggage threatening misunderstanding at every usage, onto his new understandings of state behavior. Within the confines of his own carefully specified book, his argument survives the terminological hazards he creates. In other fora, however, in which terms are assumed to actually mean what they say and conventional baggage is less easily defined away, arguing Krasner’s unconventional point with his conventional terminology would invite confusion and misunderstanding.
Conventions facilitate dialog when they provide a functional lexicon that all can refer to as standard. When they are no longer functional, however, conventions cease to play their facilitative role, and can quickly become counter-productive. That seems clearly to be the case with this conventional use of “national interest.”
Judith Goldstein and Robert Keohane, in the introductory chapter to their edited volume, provide another example of the confusion that can result when different senses of “interests” are used, apparently inadvertently in this case, within the same work. 21
In this chapter, Goldstein and Keohane set out their methodology for taking on the materialist claim that ideas do not really matter for policy outcomes. They distinguish their approach from reflectivists by bracketing the problematization of interests that reflectivists engage in. They analytically separate ideas and interests, and focus on the influence of ideas on policy outcomes. While they “recognize that ideas and interests are not phenomenologically separate,” for them “the important question is the extent to which variation in beliefs . . . affect political action under circumstances that are otherwise similar.” 22 By looking at “circumstances that are otherwise similar,” then, they seem in essence to want to hold “interests” constant.
With that as their aim, then, they first state their null hypothesis as follows:
“variation in policy across countries, or over time, is entirely accounted for by changes in factors other than ideas.” 23 (italics in original, underline added)
When the null hypothesis is restated at the end of the chapter, however, it has changed slightly:
“if... variations of interests are not accounted for by variations in the character of the ideas that people have.” 24 (underline added)
The change from positive to negative phrasing is not important here—what is important is the substitution of the word “interests” for “policies.” Having stated that they want to look for variation in beliefs in “circumstances that are otherwise similar,” (i.e. with interestsw held constant) and thus having been certain to make clear that they do not follow reflectivists in investigating the constitution of interests, they then seem to be doing just that: looking for variations in interests with variations in ideas.
The contradiction seems to flow from changing the sense with which they use the term “interests.” If the “interests” they want not to problematize are understood as interestsw (related to wellbeing), while the “interests” in the last statement of the null hypothesis are understood as interestsA (related to our attention), the seeming contradiction goes away and the original meaning of the null is preserved, more or less. This resolves the contradiction because of the fact that the jump from‘policy’ (used in the first statement of the null) to interestsA is much smaller than that from policy to interestsw, and we can imagine holding interestsw constant, or to put it differently taking them as given, while looking at variation in interestsA. Goldstein and Keohane thus appear to slip between the different senses of “interests” if without notice or signal, resulting in what seems to be a confounding contradiction or inconsistency in their arguments.
A fourth example shows ambiguity similar to some of that reviewed above, as well as some sign of awareness of the problem. Martha Finnemore, writing about humanitarian intervention, writes,
“In this essay I understand norms to shape interests and interests to shape action.” 25
Thus far it is unclear whether she means interestsA or interestsw, and the difference is important for what she is claiming about norms: to argue that norms affect state conceptions of what furthers their wellbeing is one thing, while to argue that norms affect what states pay attention to/are interested in (doing) is another. Barring simplifying assumptions (more about which later) the latter claim need not involve state wellbeing. Finnemore continues:
“Changing norms may change state interests and create new interests (in this case, interests in protecting non-European non-Christians and in doing so multilaterally through an international organization). But the fact that states are now interested in these issues does not guarantee pursuit of these interests over all others on all occasions.” 26
Here “interests” seems to mean the very broad understanding,‘issues a state is interested in,’ what I have called interestsA. Finnemore seems to recognize the terminological problem later when she puts “interest” in quotes a number of times:
As in the other examples, more precise terminology that distinguished more clearly between interestsw, interestsA, and preferences would eliminate analytical ambiguity and blurriness. It would make clearer what exactly Finnemore is arguing. Rather than hinting at alternate meanings with quotation marks, the right terminology would unambiguously put the intended meaning there on paper.
Discussion: questioning the interest-based model
Note that three of these four examples come from work that questions the standard interest-based model of state behavior. Keohane experiments with (what he calls) relaxing the assumption of egoism and introducing empathy to explain moralistic rules and unbalanced exchange relationships in international politics. Krasner depicts and tries to explain “nonlogical,” ideological anti-communist foreign policy. Finnemore presents a norm-based explanation of changes in patterns of humanitarian intervention. In each case, however, they continue to use, at least to some extent, a conceptual/ terminological toolkit designed primarily for explaining behavior within a rationalist, self-interest-based model. 29 Keohane, in fact, argues explicitly for doing so: “ . . drawing a sharp distinction between egoism and altruism . . . would confuse the issue,” he writes. “Rather than argue about egoism versus altruism, we need to ask how people and organizations define self-interest.” 30 The result of such an approach, as we have seen, is often tautology or ambiguity.
There are really two parts to the interest-based model these scholars question, one descriptive and one theoretical. Questioning either part requires some conceptual and terminological adjustment, the lack or insufficiency of which causes the problems discussed above.
Description
The descriptive aspect of the model simply addresses the empirical question of what actually motivates state behavior. Some assert that states by nature (their own nature and/or the nature of the international system) act out of self-interest, always or almost always. The three scholars above, and many others, argue (and provide evidence) to the contrary, that state behavior is sometimes driven by motivations other than self-interest (i.e. other than concern for the state’s own wellbeing). 31 Put another way, state preferences do not collapse to state interests—they are sometimes non-interest based. 32
Assumption
The theoretical part of the interest-based model questioned here is somewhat less straightforward. Robert Keohane and Judith Goldstein write that
“...rationalist analysts of international politics have often recognized that the assumption of rationality, like that of egoism,‘is a theoretically useful simplification of reality rather than a true reflection of it.’ ” 33
The approach illustrated in this excerpt is one of assuming that states act only out of self-interest (and rationally), even when it is acknowledged that this may not actually be the case all the time.
There are three reasons for making such an assumption. One is argumentational strategy. Neoliberals trying to make the case that states can and do cooperate in an anarchic system, and that they do so more than Realists think they can, have often begun with the assumption of purely self-interested states in order to make a kind of “least likely case” argument. By showing that even under this assumption it is often rational for states to cooperate with each other, they take on and show flaws in the Realist model “on its own turf,” so to speak. The idea is then that in the real world in which states may sometimes not strictly follow only self-interest, cooperation is that much more possible and likely. 34
The other two reasons for assuming self-interest are parsimony and theoretical usefulness. As Jon Elster writes, “The assumption that all behavior is selfish is the most parsimonious we can make, and scientists always like to explain much with little.” 35 Not only is the self-interest assumption “little,” i.e. simple and straightforward, it is useful: it explains a lot, or at least seems to. Much state behavior, and human behavior in general, does seem to be driven primarily by self-interest.
The self-interest assumption can become un-useful in two different ways. One is when its use as a theoretical assumption is taken to mean that it also accurately describes reality. In other words, it is sometimes slipped from one role (theoretical assumption) into the other (description of reality) without notice being taken or given.
The other way it can become un-useful is simply by impeding explanation and good theory building rather than aiding it. Much state behavior simply cannot be explained satisfactorily by an interest-based model, or can only be so explained by contorting the model so that it loses its parsimony and analytical usefulness, as illustrated by the examples discussed above. Furthermore, the explanations that flow from such a model often prove not to be in accord with the data, i.e. research seems to show them wrong. While “usefulness” in generating interesting hypotheses is certainly one criterion for judging theoretical assumptions, the agreement of these hypotheses with data is another important criterion, and hypotheses derived from a self-interest based model sometimes fall short in this regard. As Elster writes, despite the parsimony of the assumption of selfishness, “we cannot conclude, neither in general nor on any given occasion, that selfishness is the more widespread motivation. Sometimes the world is messy, and the most parsimonious explanation is wrong.” 36
For some aspects of state behavior, then, we may want to make our theoretical models somewhat less parsimonious in order to increase the empirical accuracy of the hypotheses/explanations they suggest. Hirschman makes this argument in his “Against Parsimony,” noting that others, such as Schelling and Sen, have come to the same conclusion. Here Hirschman refers to Amartya Sen: “Like any virtue, so he seems to say, parsimony in theory construction can be overdone, and something is sometimes to be gained by making things more complicated. I have increasingly come to feel this way.” 37 Later Hirschmann adds, “All these complications flow from a single source—the incredible complexity of human nature, which was disregarded by traditional theory for very good reason but which must be spoon-fed back into the traditional findings for the sake of greater realism” 38 . Assuming self-interest can be useful sometimes, then, but increasing amounts of work showing the importance of other motivations indicates limitations on this usefulness.
Questioning “self-help”
It should also be noted that in the realm of international relations the assumption of purely self-interested actors, when combined with the fact that there is no supreme global power, becomes an assumption about the nature of the international system as a whole, namely that it is a “self-help” system, that in the end it is “each state for itself.” To the extent that the self-interest assumption is relaxed, then, to that extent the self-help nature of the system is no longer deductively implied.
My argument here is similar but not identical to that made by Alexander Wendt in his “Anarchy is what states make of it” and subsequent articles. 39 The fact that there is no world government, i.e. that the international system is “anarchic,” Wendt argues, does not necessarily imply a self-help system, because such a system is contingent on states’ “conceptions of self and other,” their identifications. 40 Wendt argues that states may “identify collectively” rather than or as well as individually, drawing “the boundaries of the self” more widely, and thus may pursue other states’ interests as well as their own. The result of this would be, for states, a non-self-help system. Wendt thus questions the state-self-interest-based model, and the assumption of a self-help (for states) international system, by allowing for change in the inclusivity/extensity of the actor, away from the individual state. Put another way, he allows for changing “definitions of self.” 41
I make a slightly broader argument, questioning the model by allowing for change not just in identifications and the extensity of the actor, but more importantly in the motives of action, away from exclusive pursuit of self-interests (regardless of which “self”, a state or a higher-level collective, is involved). It is the possibility of altruistic motives, rather than a broader “self” in “self-interest”, that lies at the center of my departure from the standard model. 42
This point indicates the “halfway” nature of some of the approaches discussed above that fail to follow the implications of questioning the self-interest model to their conclusion. Keohane, for example, writes the following:
“Empathetic explanations of behavior in world politics are limited to relatively small spheres of activity: situations in which actions do not have obvious explanations in terms of more narrowly defined self-interest. The presumption in a self-help system is that empathy will play a subordinate role.” 43
For Keohane, then, the introduction of empathy does not change the self-help nature of the system. Certainly it doesn’t change the anarchic nature of the system, but what Keohane seems not to appreciate is what Wendt argues: that the jump from anarchy (lack of a supreme power) to self-help itself is an assumption, based in the assumption of purely self-interested states. Keohane’s own argument that empathy sometimes drives state behavior undermines this assumption, and thus undermines the jump from anarchy to self-help, by questioning the totality of self-interest.
Interests versus preferences, pIus...
A final example serves to illustrate a combination of the problems just discussed, as well as their link to the other main conceptual and terminological problem dealt with here, also related to the term “interests.” It comes from Judith Goldstein’s Ideas, Interests, and American Trade Policy. She writes as follows:
“[I]t is not enough to study individual preferences over outcomes—what are commonly called interests—to explain behavior. Rather, analysis must focus on preferences over actions or the choice of strategy.” 44 (emphasis in original)
The first sentence, with its equation of “preferences over outcomes” with “interests, if provides another example of the interest/preference problem discussed already above. Equating the two indicates either an understanding of the term interests in the overbroad tautological sense, meaning “preferences” of any kind, or it indicates the (unstated) assertion or assumption that all behavior is self-interested and no preferences are based in principles, passions, empathy, etc. The passage comes on p.2 of the book, and is preceded by nothing about theoretical assumptions that would indicate which way the term interests is meant, so it is unclear how to understand the passage.
Goldstein does eventually address the question of assumptions, on p.9:
“if . . . I go beyond the assumption that individuals are “self-seeking” to ask how rational actors go about furthering their interests . . . . I do not want to argue about whether “objective” interests motivate trade policy; I assume they do. This book focuses instead on how those interests are conceived by those making policy.”
Goldstein’s unwillingness to argue notwithstanding, one might first question the wisdom of Goldstein’s assumption of self-interested policy. Even in the trade sector, it is easy to find policies, such as those granting special preferences to less-developed or ethnically-related countries, that would seem to be based on other than “objective interests.” That aside, however, knowing that this is her approach, we can conclude that for her, in this book, preferences over outcomes do indeed collapse to objective interests or interestsw, since she rules out any other basis for preferences over outcomes by assumption (her assumption, in the quotation above, that individuals are self-seeking, and that interests motivate trade policy). That conclusion about her assumptions in this book, however, is a decidedly more specific and qualified one than the one she makes initially in the passage quoted above from p.2, that “preferences over outcomes . . . are commonly called interests.” As with Krasner’s book, discussed above, we can eventually get around the conceptual imprecision of the initial passage because of specifications later in the book that help with “translation” and allow the book’s argument to proceed on its own terms. Attempts to take the argument of the book into other fora, however, as with Krasner’s, would be in danger of producing confusion and miscommunication unless accompanied by similar efforts to specially specify terms and concepts in this particular (and unhelpful) way.
B. Interests versus behavior
The second major conceptual issue dealt with in this paper, introduced nicely by the first Goldstein quote, is the distinction between interests—equated by Goldstein with preferences over outcomes—and strategy, equated by Goldstein with preferences over actions. This and related distinctions have been important in recent theoretical discussions. This section subjects these theoretical distinctions to closer scrutiny.
Distinctions, dividing lines
I begin with the distinction between preferences over actions and preferences over outcomes, originally made in those terms by Robert Powell. 45 Central to the distinction that Powell makes is the effect of strategic considerations on the goals actors pursue relative to those they prefer ideally. He uses the concept of a game-theory matrix to illustrate, writing that the preference ranking of the different squares of the matrix, of the various possible outcomes of an interaction, constitute the actor’s preferences over outcomes. After the actor has taken account of how others might act, however, it will rank possible actions from most to least preferred, with that ranking indicating which action the actor believes will produce the outcome highest on its preference ranking given the expected actions/reactions of others. For Powell, then, the distinction only exists if one actor’s actions depend on what it expects from another, that is, if there is strategic interdependence. 46 Legro refers to this distinction when he sets Out the two parts of what he calls the rationalists’ “cooperation two-step”: the first step is establishing preferences over outcomes, while the second step, strategic interaction, would involve formation and implementation of preferences over actions. 47
Goldstein and Keohane use a similar, though not necessarily identical, distinction in their introductory chapter in Ideas and Foreign Policy. While recognizing that, as constructivists argue, ideas inform interests, they nevertheless take interests as given or constant (so as to be able to engage rationalist and materialist arguments that ideas are “simply hooks” used to legitimize interest-based behavior) and focus on the role they see ideas playing in influencing policy choice. 48 They thus distinguish interests from the policy choices designed to pursue those interests, just as Goldstein in her own book distinguishes interests and strategies, as quoted above. Note that these distinctions are slightly more generic than Powell’s, making a straightforward contrast between ends and means, without regard necessarily to the actions of other actors (i.e. to strategic interdependence). Powell’s distinction is thus in a sense a special case of this more generic differentiation.
Goldstein and Keohane thus position themselves between materialist rationalists, on the one hand, for whom both state interests and behavior are determined primarily or exclusively by material circumstances, (and for whom state identities are not an issue) and, on the other hand, constructivists, for whom both state interests and behavior, along with state identities, are influenced by material circumstances and cultural environments (including ideas), and social interaction. 49 While Goldstein and Keohane note their ultimate agreement with the constructivist position, they nevertheless use the interests/strategy distinction to maintain a stance midway between constructivists and materialists so as to better engage the materialist position that ideas have no independent effect even just on policy outcomes. In this position they would be joined by most neoliberals, who take state identities as given but give a causal role to ideas, institutions, or other ideational or cultural aspects of international interaction in influencing state behavior.
Wendt’s view of the theoretical landscape is about the same as Goldstein’s and Keohane’s. Constructivists and other critical international relations theorists, he writes, make two basic claims:
“that the fundamental structures of international politics are social rather than strictly material (a claim that opposes materialism), and that these structures shape actors’ identities and interests, rather than just their behavior (a claim that opposes rationalism).” 50 (emphasis added).
Wendt thus indicates clearly where neoliberals and constructivists seem to concur—in the first claim, against materialism—and where they seem to differ—the second claim, regarding the “depth” of the effects of social structures.
The following table summarizes the various distinctions just reviewed:
Distinctions
|
vs. | preferences over outcomes | ||
|
vs. | strategy / policy choice | ||
|
vs. | behavior |
Goal-strategy chains and nested games
Having reviewed the role these distinctions play in some theoretical discussions, let us now return to Powell and to a small cautionary point he makes in a footnote. While his distinction is useful, he writes, “it should not be pushed too hard. An outcome in one game may be seen as a policy choice in a larger game.” It takes only brief reflection to realize that the same goes for the interests versus behavior/strategy/policy distinction Wendt and the others use: an “interest” or goal in one game may be just a policy choice or part of a strategy in a larger game. This is illustrated by the following table, beginning with a recent policy stance taken by the Clinton administration and tracing downwards through the chain of goals and strategies that (presumably) underlie this policy. At each stage, the left hand column indicates the route that seems to have been followed, while the right hand column indicates alternative strategies that could conceivably have been chosen to meet the chosen goal in the prior stage. 51 The table is perhaps most effective if read from the bottom up.
Means-Ends Chains
A box that encloses two of these levels at a time could be moved along this progression, and at each step, the bottom level in the box could be considered the highest ranked of feasible US preferences over outcomes (a la Powell). It could also be called an interest, or a goal, or an end. At each step, the top level in the moveable box could be considered the highest ranked of US preferences over actions (a la Powell) designed to achieve the outcome below it, or a behavior in pursuit of the interest below, or a strategy to reach the goal below, or a means to the end below. For interest-based behavior, beyond the most basic goals like brute survival, any goal or strategy can be seen as:
Beyond survival, the characterization of a goal/strategy/action as an “interest” rather than a “behavior” depends not on the goal/strategy/action itself but rather on how broadly or narrowly one defines the particular game or study. Within the confines of a given study or scenario, it makes perfect sense and can be quite useful to distinguish what states “bring to the table” from what they “do at the table,” or to distinguish their means from their ends, within that particular arena. For broad theorizing, however, there is no single “table” or arena, except for the largest and most general arena encompassing all international interaction. Within that “outermost game,” however, there are a lot of nested, repeated, simultaneous, and ongoing games. General theoretical models must make sense across all these sub-arenas, and to do so, they must anchor their concepts in the most general, outermost game. The only understanding of the interest/behavior distinction appropriate for such general theorizing, then, would have to be based in the bottom-most level of the diagram above. For state-level theorizing, at least, state survival is the goal at the bottom, the basic goal from which all other interest-based goals flow.
Powell provides an example of such theorizing. He uses his distinction between preferences over actions and preferences over outcomes to clarify differences between neorealism and neoliberalism. Structural theories such as neorealism, he notes, take preferences over outcomes as given: they assume, in other words, that states want to survive, and then try to predict states’ preferences over actions, i.e. how they try to survive, from the strategic setting they face. It is here, he argues, as theories of preferences over actions, that neorealism and neoliberalism have significant differences. This usage of the actions/outcomes distinction in discussing general theory is unproblematic because its context is the largest arena, the outermost of the nested games states engage in. The preferred outcome referred to is survival, the “bottom of the chain,” so to speak.
It is when theorizing moves to smaller arenas and “inner” games, and uses arena-specific terms and distinctions but tries to apply the resulting arena-specific conclusions to the broader general arena encompassing all interstate interaction, that problems arise. If the neoliberal position that ideas and social factors do not affect interests but only affect behavior is understood in the context of the largest arena, it is uncontroversial, since it amounts to the argument that states want to survive, independent of any effects of ideas and social factors. Few would disagree. Since neoliberal, non-materialist rationalist theory generally does not limit itself to the broadest, most general arena, however, the neoliberal use of the interest/behavior distinction is simply untenable. Something that is an “interest” in one study and therefore held to be unaffected by ideas and exogenous to social interaction can become “just behavior” in a larger study and therefore suddenly seen as affected by ideas and endogenous to interaction. A factor such as ideas seen as just a “constraint” on behavior or a “cost” to be added into strategy calculations in one arena can become a determinant of basic goals in another arena.
If neoliberals want to maintain that ideas and social structures only matter down to a certain point, that “fundamental goals” are independent of ideas and social interaction, they need to specify differently than they have heretofore where the line marking “fundamental” lies, in a manner that makes the location of a particular goal or policy with respect to that line independent of the boundaries of a particular study. Beyond the most basic goal of survival, and possibly a level above that dictated by geography, it is difficult to see how such a line could be drawn.
One place one might be tempted to draw the line would be at the point at which judgment enters the picture: when reasonably sane observers could differ on the best route to follow, one might argue, one has left the realm of “interests” and entered the realm of “behavior.” As the right hand column of the above diagram illustrates, however, alternatives exist, and thus judgment enters the picture, practically “all the way down.” The same goes for considerations of strategic interdependence. With the possible exception of situations in which geographical circumstances dictate that a particular policy is very clearly necessary for survival regardless of what other actors might do, judgment and considerations of what other actors will do almost always come into play. In short, once some role for ideas and social factors is conceded, it seems that there is no theoretically convincing way to limit the depth of that role as non-materialist rationalists have tried to do.
Clearly the nature of the goals does change as one moves down the chart, from short-term to long-standing, from single situation actions to general policies meant to serve as rules of thumb for a variety of situations. Here too, however, it is difficult to draw a single clear line between “interests” and “behavior.” “Short-term” versus “long-term” is too vague, and any attempt to define it precisely would very likely be arbitrary and thus theoretically of questionable validity.
The general argument that can be made, then, is that ideas and social factors affect preferences. Whether those are characterized as interests or behavior, or as preferences over outcomes or over actions, will depend on the arena addressed by particular studies or games.
I must emphasize at this juncture that the point made here applies mainly to discussions of interest-based behavior, behavior that ultimately traces back to basic goals like survival and prosperity. Certainly other, non-interest based fundamental goals generate ends-means chains just as interest-based goals do. The arbitrary division of such a chain into “basic” and “other,” however, and the imbuing of this divide with theoretical importance, seems to be problematic only with interest-based (what I have referred to as interestsw-based) behavior.
iImplications
Wendt, Goldstein, and Keohane use the distinction between “interests” and “behavior” to distinguish rationalists from constructivists. If, as I have argued here, this distinction is untenable as a theoretical dividing line, does this mean the distinction between rationalists and constructivists also fails, since in the end both simply argue that social factors affect preferences? No.
For constructivists, culture and social interaction influence preferences because they influence state identities or “properties”, including what it means to be a “state” and what kind of state a state sees itself as being. Here, constructivists clearly do go beyond rationalists. Jepperson, Wendt, and Katzenstein, for example, write that “cultural environments affect not only the incentives for different kinds of state behavior but also the basic character of states—what we call state‘identity.’ ” 53 They emphasize the role of social structure and social interaction in the construction of “actor properties,” in which they include identities, capabilities, and interests if (translatable as goals or preferences). 54 The critique I offer here questions whether seeing social structural effects on interestw-based goals is unique to constructivist thought. It does not so question seeing social structural effects on non-interest based goals, identities and capabilities. Rationalists clearly do take state identities and capabilities as given, so constructivists, by giving this deeper role to social factors, truly do go beyond non-materialist rationalists. Furthermore, the state identity that rationalists take as given is generally that of a self-interested, non-altruistic state, leading to an emphasis in rationalist approaches on preferences related to interests (interestsw). Even when rationalists do give social structure and ideas a causal role, these are seen as affecting interestw-related preferences. Constructivists’ problematizing of identity allows them to more easily consider the possibility of non-interest-based preferences, based in principle or empathy, for example.
The following table summarizes my understanding of the theoretical landscape, taking account of the arguments I make above.
Theoretical Landscape: Summary
| constructivists: | Actor properties—capabilities, identities, and preferences—are constructed by cultural environments, are endogenous to social interaction. Preferences may be non-interestw-based. |
| non-materialist rationalists: | Actor capabilities and identities are exogenous to interaction, ideas, and culture. Preferences (over outcomes or actions, except for the most fundamental like survival and geographically dictated survival needs ) may be endogenous to interaction, affected by ideas and culture, but they remain interestw-based (because of assumed identity). |
| materialist rationalists: | Actor properties—capabilities, identities, and preferences—are given, are exogenous to interaction, ideas, and culture. Preferences are interestw-based. |
In the end, scholarship is about communication. The better able we are to communicate, the better able we are to carry on our scholarship, and the more useful that scholarship can be. This paper has argued and attempted to demonstrate that greater conceptual clarity and precision, especially in the use of a central concept such as “interests,” will do much to improve our ability to communicate, and to reduce our ability to mis-communicate. It will reduce confusion and make assumptions, arguments and theoretical positions more explicit and clear. It will not make understanding social life “easy.” It will, however, greatly increase our chances of progress toward such understanding.
Endnotes
*: Prepared for presentation at the 40th annual meeting of the International Studies Association, Washington, D.C., February 16–20, 1999. Back.
Note 1: John A. Hall,“Ideas and the Social Sciences,” ch. 2 (pp.31-54) in Judith Goldstein and Robert 0. Keohane, eds., Ideas and Foreign Policy: Beliefs, Institutions, and Political Change (Ithaca, NY: Cornell University Press, 1993), p. 31. Hall gives no citation for Poincaré. Back.
Note 2: When it is useful to do so, this can be broken down into material, psychological or other kinds of interests, but what concerns us now is the general category. Back.
Note 3: 0r, as the Bar council of Great Britain is reported to have put the second half of this, “What interests the public is not necessarily in the public interest.” “Barristers: Too Late To Harrumph,” The Economist. March 4, 1 989, p.56. cited in W. David clinton, The Two Faces of National Interest (Baton Rouge: Louisiana State University Press, 1994) p. ix, confirmed in original. Back.
Note 4: Albert 0. Hirgchman, “The Concept of Interest: from euphemism to tautology,” pp.35-55 in his Rival Views of Market Society and Other Recent Essays (New York: Viking, 1986), p.35. Back.
Note 5: ”T. B. Macaulay, “Mill’s Essay on Government: Utilitarian Logic and Politics,” in Jack Lively and John Rees, eds., Utilitarian Logic and Politics: James Mill’s ‘Essay on Government’, Macaulay’s critique and the ensuing debate (Oxford: Oxford University Press, 1978), p.125. Appeared originally in Edinburgh Review, no. xcvii(March 1829), Article vii. Back.
Note 7: Hirschman, “The concept of Interest” (above, note 4). p.50. For a similar argument, specifically with regard to the term “national interest,” see David Halloran Lumsdaine, Moral Vision in International Politics: The Foreign Aid Regime. 1949-1989 (Princeton: Princeton University Press, 1993), p.61. Back.
Note 8: Hirschman,“The Concept of Interest”(above, note 4), p.35. Back.
Note 9: In effect, defining “pursuing interests”(as opposed to pursuing something else) as designating a particular “style of conduct”attempts to force a verb’s direct object (what the verb operates on—here “interests”) into the role of an adverb (how it operates)—a problematic approach that confuses goals with the manor in which they are pursued. Back.
Note 10: If one disaggregates “society” and talks about the states involved here, the situation requires a more complicated treatment. since the general public in each society wants to possess the mountain, it is in the interestsw of each state to take or defend the mountain, since each state’s wellbeing will depend to some degree on popular support. This does nothing to vitiate the value of distinguishing the different motivations of the larger societies, however, so the complication is put aside. Back.
Note 11: Robert 0. Keohane, “Empathy and International Regimes,” ch. 14 in Jane J. Mansbridge, ed., Beyond Self-Interest (Chicago: University of Chicago Press, 1990), p. 228. Back.
Note 12: For similar criticisms of this understanding of self-interest/egoism in the international relations literature, see David A. Welch, Justice and the Genesis of War (Cambridge: Cambridge University Press, 1993), p.20, fn. 54, and Lumsdaine (above, note 7), p.61. Back.
Note 13: Keohane, “Empathy and International Regimes” (above, note 11), p.230. Back.
Note 15: Stephen D. Krasner, Defending the National Interest: Raw Material Investments and U.S. Foreign Policy (Princeton: Princeton University Press: 1978), pp.13, 35. Back.
Note 17: Ibid., pp.15-16. Back.
Note 19: Krasner is to some extent following an unfortunate convention in doing this—I simply use him as an example. The problem is more salient in his case because he takes on a topic that challenges the standard rationalist model. The terminological problems are to some extent kept hidden by rationalist assumptions. Back.
Note 20: Likewise “national” really means “state” or “central decision-makers.” This too is a confusing convention overdue for disposal, but I concentrate on “interest” and related terms here. Back.
Note 21: Judith Goldstein and Robert 0. Keohane, “Ideas and Foreign Policy: An Analytical Framework,” in Goldstein and Keohane, eds., Ideas and Foreign Policy (above. note 1). Back.
Note 25: Martha Finnemore, “Constructing Norms of Humanitarian Intervention,” in Peter J. Katzenstein ed., The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996), p.158. Back.
Note 27: Ibid., p.170, fn. 35. Back.
Note 29: For an example of work that questions the standard model but avoids the terminology problems discussed here, see Jeffrey W. Legro. “culture and Preferences in the International Cooperation Two-Step,” pp.118-137 in American Political Science Review v90, n1 (March 1996). Back.
Note 30: Keohane (above, note 11), p.228. Back.
Note 31: See for example: Jane J. Mansbridge, ed., Beyond Self-Interest (Chicago: University of Chicago Press, 1990); Robert W. McElroy, Morality and American Foreign Policy: The Role of Ethics in International Affairs (Princeton: Princeton University Press, 1992); David Halloran Lumsdaine, Moral Vision in International Politics: The Foreign Aid Regime. 1949-1989 (Princeton: Princeton University Press, 1993); David Welch, Justice and the Genesis of War (cambridge: cambridge University Press, 1993); Masato Kimura and David Welch, “Specifying‘Interests’: Japan’s claim to the Northern Territories and Its Implications for International Relations Theory,” pp.213-44 in International Studies Quarterly v42, n2 (June, 1998); Peter J. Katzenstein, ed., The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996). Back.
Note 32: For a brief discussion of this specific point, see Legro (above, note 29) p.119, fn. 5. Back.
Note 33: Goldstein and Keohane (above, note 21), p.S. They cite Robert 0. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy (Princeton: Princeton University Press, 1984), p.108. Back.
Note 34: See for example Keohane, After Hegemony (previous note), p.67. Back.
Note 35: Jon Elster, Nuts and Bolts for the Social Sciences (Cambridge: Cambridge University Press, 1989), p.54. Back.
Note 37: Albert 0. Hirschman, “Against Parsimony: Three easy ways of complicating some categories of economic discourse,” ch. 6 (pp. 142-160) in his Rival Views of Market Society and Other Recent Essays (New York: Viking, 1986), p.142. Back.
Note 39: Alexander Wendt, “Anarchy is what states make of it: the social construction of power politics,” International Organization v46, n2 (spring 1992). Back.
Note 40: Alexander Wendt, “Collective Identity Formation and the International State,” American Political Science Review v88, n2 (6/94), p.388. Back.
Note 41: Ibid., pp. 387-88. Back.
Note 42: Still, Wendt does challenge the standard model because his actors are no longer individual states promoting their own individual interests, so the system is not “self-help” from the perspective of states. Back.
Note 43: Keohane (above, note 11), pp.230-31 Back.
Note 44: Judith Goldstein, Ideas, Interests, and American Trade Policy (Ithaca: Cornell University Press, 1993), p.2. The distinction between “preferences over outcomes” and “preferences over actions” is from Robert Powell, “Anarchy in international relations theory: the neorealist-neoliberal debate,” International Organization v48 n2 (Spring 1994), pp.318. Back.
Note 45: PoweII (previous note), pp. 318ff, including footnotes 16 and 21. Back.
Note 46: PoweII (above, note 44), p.319. fn. 21. Back.
Note 47: Legro (above, note 29), p.120, fn. 6. Back.
Note 48: GoIdstein and Keohane (above, note 21), pp.4, 26. Back.
Note 49: Note the inclusion of “material circumstances” in this sentence: constructivists do not make a radical idealist argument that culture and ideas determine everything and material factors nothing. Their argument is rather that both are important. See for instance Alexander Wendt, “Constructing International Politics,” International Security v20 n1, summer 1995, p.73; and Ronald L. Jepperson, Alexander Wendt, and Peter J, Katzenstein, “Norms, Identity, and Culture in National Security”, ch.2 in Katzenstein, ed., The Culture of National Security (above, note 25), p.40. Back.
Note 50: Wendt, “Constructing International Politics” (previous note), pp.71-72. Back.
Note 51: Note that the accuracy of my presumptions about the specific elements of this chain is less important than the existence of such a chain.
It should also be noted that while there is a basic grammatical difference between, on the one hand, a condition or state of affairs ( e.g. a strong NATO ) and, on the other hand, an action or behavior (e.g. supporting NATO), one form can generally be expressed in the other form, so that this grammatical distinction is not helpful in making the theoretical distinction discussed here. Back.
Note 52: State survival is sometimes itself seen as a means to a more fundamental end, namely serving the interests of individuals or groups within the state. Where the chain stops depends in part on the level of analysis on which one is operating. Back.
Note 53: Jepperson, Wendt, and Katzenstein (above, note 49), p.33. Back.
Email this citation