By Stephen G. Brooks
International relations scholars have tended to focus on realism’s common features rather than exploring potential differences.(1) Realists do share certain assumptions and are often treated as a group, but such a broad grouping obscures systematic divisions within realist theory. Recently, some analysts have argued that it is necessary to differentiate within realism.(2) This article builds on this line of argument. The potential, and need, to divide realism on the basis of divergent assumptions has so far been overlooked.(3) In this article I argue that realism can be split into two competing branches by revealing latent divisions regarding a series of assumptions about state behavior. The first branch is Kenneth Waltz’s well-known neorealist theory;(4) a second branch, termed here “postclassical realism,” has yet to be delineated as a major alternative but corresponds with a number of realist analyses that cohere with one another and are incompatible with Waltzian neorealism.(5)
Neorealism and postclassical realism do share important similarities: both have a systemic focus;(6) both are state-centric; both view international politics as inherently competitive; both emphasize material factors, rather than nonmaterial factors, such as ideas and institutions; and both assume states are egoistic actors that pursue self-help. But these common features describe how states behave only in very general terms: they are silent about, for example, how international competitiveness varies or how and when states prioritize military security over economic capacity. Specifying how states engage in self-help requires making additional assumptions about state behavior. These further assumptions, I contend, divide realism into two branches and result in competing sets of hypotheses about how states will act in a given environment.
Three assumptions differentiate these two branches of realism. Most significant is whether states are conditioned by the mere possibility of conflict or, alternatively, make decisions based on the probability of aggression.(7) Neorealism holds that the possibility of conflict shapes the actions of states, who are seen as always adopting a worst-case perspective. Postclassical realism does not assume states employ worst-case reasoning; rather states are understood as making decisions based on assessments of probabilities regarding security threats.
Two other differences regarding assumptions naturally follow from this possibility/ probability distinction. The first related disagreement concerns the discount rate. Neorealism’s emphasis on the possibility of conflict reflects the view that actors heavily discount the future, favoring short-term military preparedness over longer-term objectives when they conflict. In contrast, postclassical realism does not regard long-term objectives as always subordinate to short-term security requirements; here, states often make intertemporal trade-offs.
The second related disagreement concerns state preferences. All realists agree that military security is the state’s prime responsibility and that relative military capacity ultimately depends on a state’s productive base. But realists diverge when these priorities conflict: common realist assumptions underspecify state preferences. All agree that defending the state from military threats takes first priority, but neorealists and postclassical realists disagree about the degree to which states favor immediate military preparedness over economic capacity. Within neorealism, military preparedness always trumps economic capacity if the two conflict. In postclassical realism, rational policymakers may trade off a degree of military preparedness if the potential net gains in economic capacity are substantial relative to the probability of security losses.
The first two sections of this article differentiate between neorealism and postclassical realism; the remainder of the article outlines some implications of this differentiation. In the third section I argue that awareness of realism’s divergent assumptions opens up avenues for cumulative intra-realist debates and should lead to a better conceptualization of the theory. The significance of these divergent realist assumptions is shown by comparing the competing hypotheses they imply about (1) German and Japanese foreign policy, (2) nuclear proliferation in Ukraine, and (3) regional economic cooperation among developing countries. In the final section I argue that a clear understanding of realism’s two branches makes it easier to understand why past interchanges with liberals, constructivists, and domestic-level theories have generally been disappointing and also opens up the possibility of a more constructive future dialogue.
The Neorealist Conception of State Behavior
This section outlines the three assumptions that undergird the neorealist conception of state behavior. The first assumption is the most significant, with the next two assumptions following naturally from the first.
Possibility versus Probability
For neorealists, the international system always has high security pressures. As John Mearsheimer contends, neorealists view the international system as a “brutal arena where states look for opportunities to take advantage of each other . . . International relations is not a constant state of war, but is a state of relentless security competition.”(8) Similarly, Waltz argues that “threats or seeming threats to . . . security abound. Preoccupation with identifying dangers and counteracting them become a way of life. Relations remain tense; the actors are usually suspicious and often hostile.”(9) Neorealists envision the system to be “one of high risk,” although “this is meant not in the sense that war constantly occurs but in the sense that, with each state deciding for itself whether or not to use force, war may at any time break out.”(10)
Waltz argues that “in the absence of a supreme authority, there is then constant possibility that conflicts will be settled by force.”(11) Yet what would seem ultimately important is not that conflict is always possible in anarchy, but rather the relative chances that it will occur. Because failing to fully balance the capabilities of potential military aggressors does not necessarily result in conflict, it would seem a rational decision maker might operate on the margin of what is safe regarding defensive preparations when the likelihood of aggression is low, especially since military preparedness is generally quite costly. This is not the neorealist perspective. Within neorealism, a rational state never lets down its guard: states adopt a worst-case perspective and always aim to balance the military capabilities of potential aggressors.(12) For neorealists, states are conditioned by the mere possibility – and not the probability – of conflict.
Neorealists regard states as adopting such a worst-case perspective for three principal reasons. First, neorealists point to the potential costs of war as causing actors to focus on the mere possibility of conflict. As Mearsheimer maintains, “political competition among states is a much more dangerous business than economic intercourse; it can lead to war, and war often means mass killing on the battlefield and even mass murder of civilians. In extreme cases, war can even lead to the total destruction of the state.”(13) In the neorealist view, the mere possibility of conflict induces a high degree of caution, given the extreme potential costs of neglecting to be defensively vigilant.
Second, neorealists argue that states will ultimately focus on other state’s underlying potential for aggression – as measured by material capabilities – because “intentions are impossible to divine with 100 percent certainty,” and the possibility always exists that “a state’s intentions can be benign one day and malign the next.”(14) In the neorealist framework, rational states adopt a worst-case focus because this is the only way to ensure against being caught off guard.
Third, neorealists maintain that rational states will focus on the possibility of conflict because defensive precautions are considered the only true assurance against aggression. For neorealists, war can be prevented – or at least forestailed – only by direct choice to pursue appropriate defensive preparations; “only the vigilance of defensive coalitions stands in the way of an Orwellian nightmare of huge clashing despotisms, or even a single world empire.”(15) States are thus seen as adopting a worst-case perspective because it is assumed that they “do not enjoy even an imperfect guarantee of their own security unless they set out to provide it for themselves.”(16)
The neorealist perspective that rational actors adopt a worst-case/possibilistic focus does not mean neorealists believe war is always highly likely. Waltz clearly argues that “world politics, although not reliably peaceful, falls short of unrelieved chaos.”(17) Although neorealists certainly do not maintain that international relations is a constant state of war, they nevertheless hold that the combined effects of the aforementioned three factors cause states to adopt a worst-case/possibilistic focus.
Neorealists would likely argue that the preceding three factors – which they assert can be traced to the anarchic state of the international system – necessarily induce rational states to adopt a worst-case/possibilistic focus. However, even if these three factors are manifested, this situation by no means compels a rational state to adopt a worst-case/possibilistic perspective – neorealists simply assume that rational states will react in this manner. These three factors merely provide a justification for neorealism’s worst-case/possibilistic viewpoint; whether this assumption is useful remains open to examination.
Although this worst-case/possibilistic view is only an assumption, it plays a pivotal – although usually unrecognized – role in neorealist theory. As Robert Powell and Alexander Wendt argue, balancing behavior and attitudes toward cooperation are conditioned not by a lack of hierarchical authority in the international system per se but by the perceived relative likelihood that force will be used.(18) Absent the worst-case assumption undergirding neorealism, even in an anarchic environment there is no logical reason to infer that balancing behavior will constantly recur and that states will be highly averse to cooperate. Ultimately, this worst-case/possibilistic assumption-and not the condition of anarchy – performs the bulk of the explanatory work in the Waltzian neorealist framework.
Ironically, although neorealists are leading critics of classical realist explanations of international behavior predicated on particular conceptions of human nature,(19) the internal coherence of the neorealist framework itself depends fundamentally on the psychological assumption that actors are characteristically highly fearful. For Hans Morgenthau, actors are guided by a rapacious quest for power that is the result of an aggressive, animal-like craving to dominate one’s fellows.(20) Morgenthau’s view of human nature thus emphasizes aggression, whereas the neorealist conception emphasizes wariness and anxiety. These two conceptions of human nature may not be so distinct from one another: if actors are understood to be aggressive, it makes sense to assume they will also be characteristically highly fearful. In this respect, neorealists may adopt a worst-case/possibilistic perspective precisely because they implicitly accept Morgenthau’s argument that actors are inherently aggressive. Taken to its logical conclusion, Morgenthau’s view of human nature implies that (1) actors will seek to take advantage of weaker states whenever they have the chance, and hence (2) military preparedness is the only true assurance against being exploited. This is remarkably similar to neorealism’s view that “conflict is common among states because . . . [of] powerful incentives for aggression,”(21) and where defensive vigilance is the only guarantee against aggression.(22) Neorealists thus emphasize a different aspect of human nature than Morgenthau, but the net result is that both view the world in very similar, highly pessimistic terms. In the end, neorealism does not move beyond the human nature arguments of classical realism; neorealists simply swap one aspect of human nature (aggression) for another (fear).(23)
To be clear, neorealism’s reliance on this worst-case/possibilistic assumption is not fatal for the theory. Whether this assumption is a useful approximation of international decision making remains an open question; neorealists may be correct that states are most productively characterized in this manner. My task at this juncture has not been to evaluate whether this assumption is useful, but rather to emphasize its centrality in the neorealist architecture.
Short Term versus Long Term
Because anarchy provides no guarantees against elimination, one could argue that states will always seek to first maximize their military security in the short term, even if doing so has less-than-ideal repercussions for the state’s long-term priorities. In practice, a rational state will not necessarily discount the future in this manner, even in an anarchic system. Rather, how rational states weigh short-term military security against long-term goals depends on the strength of security competition in the international system.
In a system with high levels of security competition, a rational state’s first concern will be to maximize the likelihood of its continued existence, even if focusing on short-term security has negative long-term repercussions. But, if security pressures are not as strong, a rational state will give more weight to long-term priorities. In sum, the more competitive the system, the more a rational state discounts the future.
Seeing the international system as a relentless competition for security in which states adopt a worst-case/possibilistic perspective drives neorealist assumptions about the discount rate. Given this view, a rational state will always seek first to maximize its short-term military security from potential rivals, even if this has negative long-term repercussions for other state priorities. Of course, neorealists do not think long-term state goals are unimportant, but they do view such concerns as subordinate to short-term military security requirements when the two conflict. To be clear, neorealism’s view that states heavily discount the future follows not from the anarchic nature of the international system per se, but rather reflects the theory’s assumption that states are shaped by the mere possibility of conflict and hence seek to be prepared for all contingencies regarding the short-term use of force by potential rivals.
Military Security versus Economic Capacity
Waltz contends that states will be concerned, above all else, with securing their survival.(24) While Waltz’s argument appears to say a great deal, this notion of survival is only a precursor to an understanding of state preferences rather than a satisfactory conception on its own. Surviving in anarchy requires both a potent military deterrent and a dynamic productive base. As a result, identifying survival as states’ ultimate goal is insufficient: how states balance the military security component of a survival strategy in relation to the economic capacity element must still be established. Will a rational state always maximize its military security, even if doing so sometimes significantly constrains its economic capacity? Alternatively, will a state sometimes attempt to enhance its economic capacity, even if it thereby reduces defensive vigilance and potentially exposes a state’s vital interests to some military danger? Waltz’s argument that states pursue survival thus contains an underlying tension between military security and economic capacity; however, Waltz skirts this trade-off by describing “moves to increase economic capability, to increase military strength” as both being elements of an internal balancing strategy.(25)
Of course, economic capacity and military preparedness are not always incompatible. Yet, heightened military preparedness often leaves less resources available for economic priorities, especially over the long term. This is not to say a fixed inverse relationship exists between the two objectives, where increasing attention to military security always necessarily causes a corresponding decrease in economic capacity.(26) Policymakers in wealthy states may not always be overly concerned even where an inverse relationship does exist. Nevertheless, it is undeniable that significant tension often exists between these two goals, especially for states with highly constrained economic resources. Moreover, as John Lewis Gaddis, Paul Kennedy, and William Wohlforth emphasize, decision makers very often perceive that they face such a trade-off and hence design policy based on the assumption that they cannot fully satisfy both objectives.(27)
In practice, a fundamental difference exists between pursuing state survival by emphasizing military preparedness substantially more than economic capacity (as North Korea has done in the postwar period and the Soviets did until the mid-1980s) and the opposite, advancing economic capacity at the expense of providing the highest level of protection from potential short-term military rivals (as the British did in the mid-1930s, as the Soviets did in the late 1980s, and as Ukraine has done recently). Lumping such radically different strategies under amorphous headings such as “survival” and “self-help” is highly problematic because it leaves neorealism with very little explanatory content.
Going beyond Waltz’s underspecified notion of survival to a more precise neorealist conception of state preferences requires examining how states trade off military security and economic capacity when the two conflict. What is the marginal rate of substitution between these two goals that characterizes state behavior? In other words, how much economic capacity will a state be willing to give up in order to have more military security? The trade-off between these two goals will frequently be an intertemporal one between short-term military security and long-term economic capacity.(28) For this reason, assumptions about the discount rate substantially affect how the trade-off is understood.
How do neorealists envision states making this trade-off? Neorealism’s worst-case/possibilistic perspective and high discount rate reflects the view that rational states will seek first to ensure military security before advancing other objectives, such as economic capacity. This is not to claim, as Richard Rosecrance does, that neorealists maintain “economics need not be included in a valid conspectus of international politics.”(29) Neorealists certainly do not view economic capacity as unimportant. However, neorealists consider it irrational for a state to focus on the enhancement of economic capacity to the extent that the likelihood of being subjected to a military defeat by potential rivals increases to any degree. As Mearsheimer maintains, “states operate in both an international political environment and an international economic environment, and the former dominates the latter in cases where the two come into conflict. The reason is straightforward: the international political system is anarchic.”(30)
This neorealist view that security priorities trump economic capacity whenever the two conflict implicitly presumes states favor short-term military security over long-term military security. Why? Economic capacity ultimately provides the foundation for future military security; as a result, engaging in intertemporal trade-offs between short-term military security and long-term economic capacity is also a choice between military security in the short term versus the long term. This implicit neorealist argument that protection from short-term potential threats trumps long-term military security is consistent with the theory’s underlying view that actors heavily discount the future. Given that long-term economic capacity and long-term military security overlap, the most precise articulation of neorealism’s conception of state preferences is consequently that short-term military security requirements supersede both short-term and long-term economic capacity.
It would be an incorrect caricature to portray neorealists as simply maintaining that states will always take every conceivable step to ensure their short-term military security vis-a-vis potential rivals. Waltz very carefully argues that “to say, then, that international politics is a game the general rules of which are disregarded at the peril of the player’s existence does not necessarily mean that every state must bend all its efforts towards securing its own survival.”(31) Waltz recognizes states have other important objectives besides military security, and he does not argue that states will devote all, or even anywhere near all, of their resources just to ensure this objective.(32) The neorealist perspective allows states to pursue economic capacity to a significant degree but not to the extent that doing so leaves a state potentially vulnerable to military exploitation by a possible military rival.
Another mischaracterization would be to portray neorealists as arguing that states will always maximize their short-term military security by balancing the capabilities of potential aggressors. Waltz carefully argues that just as firms are free to pursue other objectives to the detriment of profit maximization, states similarly have latitude to pursue other goals to the detriment of their chances for continued existence.(33) Neorealists should thus not be stereotyped as arguing that states will never pursue economic capacity to the extent that military security from possible rivals is potentially jeopardized. What Waltz does argue is that “balance-of-power politics is risky; trying to ignore it is riskier still,” because international politics is a realm where “any state may at any time use force [and] all states must constantly be ready either to counter force with force or to pay the cost of weakness.”(34) Just as firms that ignore market forces will be punished, Waltz similarly argues that “a unit of the system can behave as it pleases,” but “the international arena is a competitive one in which the less skillful must expect to pay the price of their ineptitude.”(35) Neorealists recognize states may decide to pursue economic capacity at the expense of military preparedness vis-a-vis potential rivals, but such decisions are seen as having been caused by a lack of proper understanding of the fundamental competitive tenets that govern international politics.
This analysis makes it easier to understand why neorealists resist broadening the concept of security beyond military factors to include an economic element, as David Baldwin suggests.(36) Neorealism’s worst-case focus and conception of the discount rate culminates in a particular understanding of state preferences – where short-term military security concerns always trump those of economic capacity whenever the two conflict – which then leads neorealism’s proponents to focus on the military aspect of security while downplaying the economic element. Broadening the concept of security to include economic factors simply cannot be accomplished within the neorealist framework, since it would require dismantling the underlying assumptions that provide neorealism with its internal coherence. Not surprisingly, therefore, neorealists prefer to restrictively define the concept of security in military terms. Taking this line of argument to its logical conclusion makes it easier to understand why Waltz ultimately concludes that it is possible, indeed productive, to ignore the economic domain when theorizing about international behavior.(37)
What Is Neorealism’s Theory of Decision Making?
Expected utility, the standard view of decision making in mainstream international relations, is employed, although often not explicitly, by a wide spectrum of analysts in varying degrees of formality.(38) Up to this point, neorealism has normally been understood to share this perspective that actors weight the utilities of different outcomes by their probabilities, calculate costs and benefits of all alternative policies, and choose the option with the highest utility. On closer examination, however, neorealism falls outside this mainstream view; neorealism does not have expected utility foundations. As Robert Keohane argues, “in a standard expected utility formulation, states will not let mere possibilities determine their behavior.”(39) Instead, in the expected utility framework, the utilities of outcomes are weighted according to their subjective probability. Yet, as emphasized here, neorealism’s worst-case/possibilistic focus is insensitive to probabilities.
Since neorealism is inconsistent with the typical expected utility framework, it becomes useful to speculate as to what theory of decision making neorealism represents. One possibility is “minimax.” Actors pursuing a minimax strategy do not pursue aggregate expected utility per se, but instead choose options that minimize the maximum loss that they can suffer.(40) As R. Duncan Luce and Howard Raiffa emphasize, actors that use the minimax risk criteria are “ultraconservative (or pessimistic) in that, relative to each act, they concentrate on the state having the worst consequence.”(41) In the expected utility framework, actors factor the opportunity costs of precautionary measures into their decisions and take chances when the potential gains of doing so are large relative to the probability of losses. In contrast, actors that always pursue a minimax strategy do not take chances under any circumstances: the overriding goal is to minimize the likelihood that the worst-case scenario will occur. In this respect, minimax is highly compatible with neorealism’s worst-case emphasis. The minimax criterion also does not require the actor to estimate the probabilities associated with different outcomes and is consequently consistent with neorealism’s possibilistic focus.
A second alternative is that neorealism is representative of prospect theory.(42) Prospect theory argues that actors give more weight to losses than to gains, and also that actors will often exaggerate the likelihood of rare events; “dramatic events which come readily to mind . . . are perceived to be more likely than they actually are.”(43) This perspective that actors are dominated by loss aversion and that they will exaggerate rare and dramatic events parallels neorealism’s view that actors focus on the worst case with respect to potential losses of military security. Moreover, proponents of prospect theory argue that actors will be “willing to pay far more to reduce the risk of a catastrophic loss from .10 to 0 than from .20 to .10, even though the change in expected utility is the same.”(44) This contention that actors will pay a steep price to reduce the risk of a catastrophic loss down to very low levels is compatible with neorealism’s conception of state preferences, where military security requirements always trump those of economic capacity whenever the two conflict.
The “Postclassical Realist” Conception of State Behavior
Does realism admit any other understanding of state behavior that can be contrasted to the Waltzian neorealist conception? Robert Gilpin argues that all realists share an assumption regarding “the essentially conflictual nature of international affairs.”(45) Gilpin’s argument is accurate but underspecified; the crucial question is the extent to which international affairs are conflictual. Does the level and form of conflict vary over time? If so, to what degree and according to which factors?
A variety of current realist writers provide very different answers from those advanced by Waltzian neorealists. Although some of these realist scholars have attempted to merge their analyses into the Waltzian framework, their analyses do not follow from neorealism’s worst-case/possibilistic assumption. Instead, they can best be understood as reflecting a competing branch of realism – what is termed here “postclassical realism” – which views actors as being conditioned by the probability of conflict.
Although the emphasis in this section is to delineate how postclassical realism differs from Waltz’s neorealist theory, the postclassical realist conception of state behavior outlined below simultaneously stands in contrast to classical realism. The term “postclassical realism” was chosen to designate a branch of realism that does not share four important characteristics that are held in common by classical realism and Waltz’s neorealist theory. Specifically, neorealism and classical realism share the following characteristics: (1) they have a highly static conception of international relations; (2) they rely on particular aspects of human nature – aggression for classical realists, fear for neorealists – to generate hypotheses; (3) they assume that states tend to rely primarily on the use or threat of military force to secure their objectives; and (4) they concentrate on the balance of military capabilities, with neorealists excluding and classical realists generally downplaying other international-level influences on state behavior. Postclassical realism does not subscribe to these four characteristics.
Possibility versus Probability
Consistent with their worst-case/possibilistic perspective, Waltzian neorealists place the emphasis on a single endogenous factor as affecting the likelihood of conflict: the balance of military capabilities.(46) Aggression is seen as less likely when states balance the capabilities of potential aggressors. Do any other variables besides the distribution of military capabilities cause the likelihood of conflict to vary systematically? This question has been addressed extensively by nonrealist theories. Liberals argue that the probability of conflict depends on whether or not states are democratic,(47) and on institutional linkages between international actors.(48) Constructivists argue that the likelihood of conflict depends on the nature of shared understandings regarding norms and identities between actors.(49)
That realists often disagree with liberals and constructivists should not obscure the fact that an expanding literature exists in which a variety of realist scholars – all of whom, as will be seen, are best understood as falling outside Waltzian neorealism – have sought to answer the very same question that guides nonrealist analyses: which factors besides the distribution of military capabilities systematically cause the probability of conflict to vary? These realist scholars point to three material factors other than the distribution of capabilities that affect the probability of conflict: technology, geography, and international economic pressures. First, the importance of technology is identified by realists such as Robert Jervis, Barry Buzan, Charles Glaser, and Stephen Van Evera. Technological innovation affects a range of factors that have a beating on conflict, such as: (1) the offense-defense balance and offense-defense differentiation;(50) (2) “interaction capacity,” that is, the volume, speed, range, and reliability of communications;(51) (3) reconnaissance capabilities;(52) and (4) the ease of extracting economic resources from conquered territory, which depends substantially on whether productive capacity is based on knowledge-intensive/technological industries or more traditional “smokestack”/natural resource industries.(53) Second, realists, including Stephen Walt and Stephen Krasner, identify the importance of geography, which affects both the utility of employing military force(54) and access to raw materials.(55) Third, realists such as Gilpin and Wohlforth underscore the significance of international economic pressures, which lead to fluctuations in the economic opportunity cost of an assertive foreign policy stance(56) and affect whether or not states can most cost-effectively influence other states through informal economic, as opposed to formal military, means.(57)
None of the aforementioned material factors involves ideas or institutions, which are the focus of nonrealist theories such as liberalism and constructivism.(58) These material factors make conquest more-or-less cost effective – and therefore more-or-less likely – irrespective of institutional characteristics or shared understandings about norms and identities. I am not suggesting that these material factors provide a comprehensive understanding of the causes of conflict, nor that they are of equal importance. Rather, I merely wish to underscore that realist scholars identify a range of material factors other than the distribution of military capabilities that influence the probability of conflict.
Until now, there has been confusion whether the preceding realist arguments can be incorporated into the Waltzian neorealist framework. Some of these realist writers portray themselves as amending and revising, rather than critiquing, the Waltzian neorealist approach.(59) As emphasized, neorealism’s conception of state behavior is based on the assumption that states are conditioned by the mere possibility of conflict. The preceding analyses simply do not follow from neorealism’s worst-case/possibilistic assumption; the underlying argument advanced by these realist scholars – that the probability of conflict varies systematically according to factors other than the distribution of capabilities – instead serves to undermine this neorealist assumption. With their implicit emphasis on probabilities, these analyses are incompatible with the Waltzian framework: incorporating any of these factors into neorealism would gut the theory’s worst-case/possibilistic assumption, thereby destroying neorealism’s internal coherence. This helps to explain why Waltz is so resistant to include any additional international-level variables beyond the distribution of capabilities into his theory. In the end, each of the preceding arguments can be more productively incorporated into a competing realist framework – postclassical realism – that is based on the underlying assumption that actors make decisions based on the probability of conflict.
Before outlining postclassical realism’s remaining two assumptions about state behavior, it should be noted that the possibility/probability distinction outlined here subsumes an emerging debate within realism between what Jack Snyder terms “aggressive” and “defensive” realists.(60) Snyder argues that aggressive realists, who are sometimes also referred to as “offensive” realists, believe offensive military action often contributes to security, whereas defensive realists reach the opposite conclusion.(61) In turn, Benjamin Frankel maintains that aggressive realists “posit that security in the international system is scarce,” whereas defensive realists contend security is more plentiful and are “more optimistic about the likelihood of avoiding war.”(62)
In practice, an implicit worst-case assumption regarding the expected future behavior of rising powers leads aggressive realists like Mearsheimer to assert that states “seek opportunities to weaken potential adversaries” as a means of enhancing security and hence to regard security as being scarce in the system.(63) Within a worst-case framework, states focus on the balance of capabilities and are conditioned by the fear that a rising power might become aggressive at some point. In this view, states are inclined to undertake offensive military action to remove or cripple rising powers in order to forestall the possibility of coercive behavior once potential competitors become predominant.
In contrast, adopting a probabilistic view leads to a different conclusion: given that preventive war is risky and costly, acting preemptively contributes to security only when a significant probability exists that the rising power will aggressively use military force when it becomes predominant in the future – which, for defensive realists such as Jervis, Glaser, and Walt, depends not just on the balance of capabilities, but also on technology, geography, and international economic pressures. Ultimately, defensive realists have a much more conditional view than aggressive realists as to whether offensive military actions enhance security – and hence see security as often being plentiful – exactly because of their implicit focus on the probability of conflict.
That the aggressive/defensive debate ultimately reflects a deeper divergence within realism regarding assumptions does not in any way diminish its significance. Understanding that the aggressive/defensive debate is rooted in the possibility/probability distinction is vital, however, because it helps explain why this intra-realist dispute is in fact occurring, and, more importantly, it underscores the need to delineate the assumptions according to which realists diverge. Tracing empirical disputes such as the aggressive/defensive realist debate back to underlying differences regarding assumptions is imperative to promoting cumulation within realist theory; focusing on assumptions makes it possible to interconnect a whole series of empirical disputes within realism and promote aggregation across a broad range of cases.
Short Term versus Long Term
As mentioned earlier, when security pressures are high, a rational actor will significantly discount the future in order to guard against elimination by stronger rivals. The focus of postclassical realism on the probability – and not the possibility – of conflict results in a conception of the international system as often having lower security pressures than neorealists assume. As a result, postclassical realism expects states to often discount the future to a lesser extent than is assumed by Waltzian neorealists. For postclassical realism, long-term state objectives are not necessarily subordinate to short-term military security requirements; instead, actors are seen as regularly making intertemporal trade-offs.
Regarding intertemporal trade-offs, Gilpin’s realist analysis provides an excellent contrast to Waltzian neorealism.(64) Gilpin’s arguments reflect the view that actors do not make worst-case assumptions but are instead conditioned by the probability of conflict. Perhaps not surprisingly, Gilpin also argues that international actors will often make trade-offs between short-term and long-term objectives. Indeed, much of Gilpin’s work is devoted specifically to examining those factors within a state’s control that may potentially alter its long-term relative position.
In comparison to Gilpin’s more dynamic realist framework, Waltzian neorealism has often been criticized on the grounds that it is a very static theory.(65) Yet, this should not be seen as an omission on the part of Waltz and his followers; the static nature of neorealism reflects its worst-case/possibilistic focus. Given such an understanding of the world, a rational state will have a very short-term focus, and consequently neorealist theory itself not surprisingly exhibits this same characteristic. In contrast, Gilpin’s probabilistic-based analysis leads to the view that states often have the discretion to look beyond their short-term security requirements and worry about the longer-term consequences of their actions. Ultimately, it is Gilpin’s underlying probabilistic focus that lends itself to an examination of long-term changes.
Military Security versus Economic Capacity
Postclassical realism’s probabilistic underpinnings result in a rival understanding of state preferences from that advanced by neorealists. In this respect, Gilpin’s analysis again provides a useful contrast to Waltzian neorealism. Until now, however, the prevailing tendency has been to portray Gilpin’s conception of state preferences as reflecting Waltzian neorealism, rather than to explore how, or if, Gilpin might be advancing a competing understanding.(66) This has partly been the result of several ambiguities in Gilpin’s presentation. My aim here will be to outline Gilpin’s underlying conception of state preferences more precisely and to provide a clear contrast between neorealism and postclassical realism regarding their expectations about state behavior.
Gilpin argues that all realists are united in assuming “the primacy in all political life of power and security in human motivation.”(67) Although this may seem a clear exposition of the primary objective of decision makers, it is actually a much more ambiguous statement. Gilpin’s argument that all realists view states as pursuing power and security is complicated by the fact that the pursuit of power and security do not always perfectly overlap. Gilpin defines power as resources: the combined “military, economic, and technological capabilities of states.”(68) Thus, power includes – but is not restricted to – military capabilities; more specifically, power contains within it two different elements – military preparedness and economic capacity – that will sometimes be incompatible. Viewing states as pursuing power leads to very different hypotheses about state behavior than if they are seen as maximizing security. Maximizing security implies that military security requirements will trump those of economic capacity whenever the two conflict – as neorealists argue. In contrast, the argument that states pursue power suggests that states will make trade-offs between economic capacity and military preparedness, given that power is ultimately a function of both military and economic might. In the end, however, it is unclear whether Gilpin views states as ultimately pursuing power or security.(69)
Waltz is very clear on this question, arguing “the ultimate concern of states is not for power but for security.”(70) Waltz recognizes that the pursuit of security and power do not always coincide and makes explicit his view that a rational state will seek power only if the security objective has first been ensured.(71) Yet, Mearsheimer appears to contradict Waltz on this point; Mearsheimer asserts that states ultimately “aim to maximize their relative power.”(72) Mearsheimer’s statement notwithstanding, his own analyses and hypotheses reflect the view that states ultimately pursue security, not power, as Waltz argues. Mearsheimer focuses on the military basis of power, asserting that states aim “to acquire more military power at the expense of potential rivals.”(73) Of course, military might is not the only basis of power, with economic capacity also being a vital component, as Gilpin and Kennedy emphasize.(74) In the end, Mearsheimer can most accurately be viewed as arguing that states above all seek to maximize their security but, in pursuing this goal, will sometimes also enhance their power. The pursuit of power and security coincide only under certain conditions – notably when capturing the economic resources of another state is both the preferred route to security as well as the most cost-effective means of increasing power. It is important, however, not to confound – as Mearsheimer apparently does – the notion that the pursuit of power and of security can sometimes overlap with the idea that states ultimately pursue power, since there will be many circumstances when the pursuit of power and of security will not only fail to overlap, but will actually conflict, as Waltz recognizes. In circumstances where the pursuit of power and security do, in fact, conflict, Mearsheimer would undoubtedly end up agreeing with Waltz’s view that states first seek to ensure their military security before they pursue power. For neorealists, therefore, states ultimately pursue security, not power.
Another source of confusion within Gilpin’s analysis – which relates directly to the power versus security question – concerns whether he views states as having a hierarchy of objectives. In some places, Gilpin contends states do have a hierarchy of goals; be argues that although states have a wide variety of objectives, all of the “more noble goals” of society “will be lost unless one makes provision for one’s security in the power struggle among groups.”(75) Inherent in this formulation is the idea that security requirements trump the pursuit of other state objectives, and hence that Gilpin’s analysis is compatible with the Waltzian neorealist conception of state preferences.(76) At other points, however, Gilpin maintains that states do not have a hierarchy of objectives; instead, he argues that theories that “assume that one can speak of a hierarchy of objectives” actually “misrepresent the behavior and decision-making processes of states.”(77) Rather, Gilpin asserts that “it is the mix and trade-offs of objectives rather than their ordering that is critical to an understanding of foreign policy.”(78)
Gilpin is thus ambivalent about whether to assume that states have a hierarchy of preferences or to criticize such a perspective. Gilpin equivocates because he actually occupies a middle-ground position: although he clearly regards military security as the vital priority of the state, he does not view rational states as privileging short-term military preparedness over economic capacity in the manner that neorealists do. In the end, Gilpin can best be understood as arguing that states do have a hierarchy of objectives, yet it is a much more flexible hierarchy than neorealists envision.
This understanding of Gilpin’s argument can be generalized to produce a postclassical realist conception of state preferences. Whereas neorealism views the pursuit of power as secondary to that of security, postclassical realism regards rational states as ultimately seeking to increase the economic resources under their control – and thus their long-term power given that “wealth . . . is a necessary means to power”(79) – subject to the constraint of providing for short-term military security. For postclassical realism, therefore, power – and not security – is the ultimate goal of states.
The postclassical realist view that states pursue power is compatible with Gilpin’s apt contention that it is the trade-offs, and not the ordering, of objectives that is important, since the concept of power itself contains within it an inherent tension between economic capacity and military security. In marked contrast to neorealism, postclassical realism argues that a rational decision maker may decide to trade off military preparedness to some degree when the potential net gains in terms of enhanced economic capacity are substantial relative to the probability of security losses. States will be especially likely to make such trade-offs when their economic resources are highly constrained. Perhaps the most prominent example of such a scenario is Gilpin’s and Kennedy’s argument that a declining power may rationally decide to retrench from some of its external positions, enhancing long-term economic capacity at the cost of somewhat reducing short-term protection vis-a-vis potential military adversaries.(80)
The idea that states seek power is not new in realist thought; it provided the foundation for the classical realism of Morgenthau, Rienhold Niebuhr, and others. Is the postclassical realist conception of power as the dominant goal of international actors distinguishable from the classical realist understanding? In practice, a crucial difference exists between the two approaches concerning their view of power. Specifically, although both theories regard the international system as a competition for power, they disagree as to the form and intensity that the pursuit of power will take.
Classical realism views decision makers as constantly striving to dominate others: policymakers are characterized as being guided by a rapacious lust for power. Within classical realism, therefore, power is regarded as an end in itself: for Morgenthau, one’s “lust for power would be satisfied only if the last man became the object of his domination.”(81) As a result, states are viewed as always actively seeking any possible means to advance their power over other nations, taking military advantage of weaker states whenever they have the chance.(82)
Postclassical realism advances a very different perspective: state decision makers do not maximize power because of an insatiable desire to dominate others; rather states pursue power because doing so allows for maximum flexibility in achieving the nation’s instrumental interests. In other words, postclassical realism holds that decision makers pursue power because it is the mechanism by which to achieve the state’s overriding objectives.(83) States are seen as seeking to enhance their share of economic resources, and hence their power, because it provides the foundation for military capacity, and furthermore because economic resources can themselves be used to influence other international actors. Because power is viewed as a mechanism, and not an end in itself, states are expected to pursue power subject to cost-benefit calculations. In this respect, Gilpin argues that “there have been many cases throughout history in which states have forgone apparent opportunities to increase their power because they judged the costs to be too high.”(84) Compared to classical realism, therefore, postclassical realism envisions states as being more deliberative in their pursuit of power.
The postclassical realist assertion that states pursue power does not mean that states are seen as necessarily engaging in conquest. Engaging in conquest in order to capture economic resources controlled by other states is a method to increase power, but it is not the only method. States can also enhance their relative share of economic resources, and hence their power, through nonmilitary means, such as: (1) by actively seeking changes in international trade patterns;(85) (2) by creating more efficient institutions to reduce transaction costs and better ensure property rights;(86) (3) by using economic leverage to secure supplies of inexpensive raw materials and other supplies from weaker states;(87) and (4) by reducing nonproductive expenditures to free up resources for economic advancement.(88) The postclassical realist conception of international behavior asserts that in situations when the first strategy, conquest, is the most cost-effective means to increase power, states will be prone to use military force in order to enhance their power. In contrast, in situations when one or a combination of the earlier mentioned nonmilitary strategies provides the most cost-effective means to increase power, international competition will still be endemic, but states will not be prone to use military force to further their power; rather, as is the case today among the major industrialized powers, states will rely on nonmilitary means to enhance their power. In comparison to neorealism, therefore, postclassical realism advances a much more conditional view of whether incentives for military aggression exist in the system; this conditional assessment undergirds the postclassical realist view that states will not make worst-case assumptions but will, instead, be conditioned by the probability of conflict.
In summary, postclassical realism regards states as ultimately seeking to increase their share of economic resources and, hence, their power. This focus on power resembles classical realism, but postclassical realism has a very different understanding of how and why states pursue power in the international system.
Implications for Realist Theory
Given realism’s focus on competition in the international system, it is curious that realists have so far refrained from engaging in intellectual competition with each other regarding their theory’s assumptions. Thus, although realists have frequently engaged in spirited debates with nonrealists regarding assumptions about state behavior,(89) they have not yet similarly directed their sights in an inward direction. The framework advanced here should promote some much needed debate among realists regarding their theory’s underlying assumptions.
As will be seen, the assumptions undergirding the analyses of neorealists such as Waltz and Mearsheimer lead to one set of hypotheses about state behavior, whereas the contrasting assumptions of postclassical realism result in a rival understanding. For reasons of space, the empirical analysis here is an abridged comparison intended to underscore the importance of explicitly taking into account the variance in realist assumptions about state behavior. Throughout this section, I do not mean to imply that realist-based analyses of the phenomena in question provide the only useful explanation, just that exploring these contrasting realist hypotheses is a useful exercise.
Neorealism advances very few hypotheses about state behavior; the three principal hypotheses are: (1) balancing behavior constantly recurs, (2) states will be constrained from engaging in cooperation, and (3) states copy the advances made by rival powers (the “sameness effect”). The current empirical manifestations of these three neorealist hypotheses – and the contrasting postclassical realist understanding – are outlined here. These cases were thus chosen because they provide a range of examples that is broad but falls well within the restricted explanatory scope of neorealist theory.
1. Balancing Hypothesis: The Future Behavior of Germany and Japan
Neorealism’s main hypothesis is that balancing behavior constantly recurs.(90) With respect to the current environment, neorealists argue international incentives will compel Germany and Japan to rise to great power status to balance the United States’ military preponderance.(91) For neorealists, “the hegemon’s possession of actual or latent military capabilities will result in balancing regardless of its intentions.”(92) Thus, even if the United States does not actively seek to threaten the interests of Germany or Japan, neorealists argue that balancing behavior will occur nevertheless.(93) This view that Germany and Japan will strengthen their militaries in order to guard against the mere possibility that the United States might act coercively clearly reflects neorealism’s worst-case perspective.
Christopher Layne pointedly criticizes those who argue that Germany and Japan will “eschew military strength in favor of economic power.”(94) He argues that “eligible states that fail to attain great power status are predictably punished,” and hence that Japan and Germany will strive to become military powers even though pursuing such a policy will entail substantial economic costs.(95) In this respect, neorealists maintain that Germany and Japan will not continue down the path they have taken since World War II namely, focusing on economic capacity while avoiding large expenditures on military security.
For postclassical realism, in contrast, Germany and Japan will not likely balance the United States. Even if Germany and Japan grow economically in the future, developing a military force capable of credibly deterring the economically much larger United States will continue to be very costly. A choice by Germany or Japan to bolster their military forces to balance the United States would also entail considerable economic opportunity costs: each would risk a reduction both in their substantial international export markets and in their access to certain technologies. Given these large economic costs, postclassical realism asserts that balancing behavior will occur only if there exists a significant probability – and not just the possibility, as neorealists argue – that the United States will use its superior military capabilities in a coercive manner. In the end, postclassical realism does not expect Germany and Japan to move in a unilateral direction, since the economic constraints of doing so are more salient than the probability of military exploitation by the United States.
It is perhaps too soon to definitively conclude whether the neorealist or postclassical realist understanding is relatively more useful. However, the latest evidence favors postclassical realism. German defense expenditures have declined in absolute terms every year since 1990.(96) In November 1995, the Japanese Cabinet approved a new defense plan outlining significant military spending cuts.(97) Furthermore, neither Germany nor Japan has sought to break free of U.S. influence by respectively weakening NATO or the U.S.-Japan Security Treaty; rather, both countries have sought to reaffirm their security commitments with the United States.(98) Layne notwithstanding, it appears that Germany and Japan have decided that they can best advance their international influence by continuing to enhance their economic, and not military, strength.
2. “Sameness Effect” Hypothesis: Ukraine and Nuclear Weapons
The second principal neorealist hypothesis is that the “possibility that conflict will be conducted by force . . . produces a tendency towards the sameness of the competitors,” – that is, rational states will imitate successful international technological, organizational, and other advances that have been adopted by competing powers.(99) With respect to the current international environment, the most important illustration of this “sameness effect” hypothesis is the neorealist argument that there will likely be significant nuclear proliferation in the post-Cold War era.(100) The most prominent case in point concerns Mearsheimer’s 1993 argument that Ukraine would retain the nuclear deterrent it inherited following the breakup of the Soviet Union.(101)
Mearsheimer likely advanced his unequivocal claim with respect to Ukraine for two principal reasons: first, at that time Ukraine already possessed nuclear weapons; and second, Ukraine faced a substantial security threat from Russia, a country with substantially greater military resources at its disposal and with which Ukraine has a history of tense relations, including several unresolved border disputes, particularly over Crimea. However, maintaining a nuclear deterrent would have been very expensive, requiring a substantial portion of Ukraine’s gross domestic product. Furthermore, the costs to Ukraine of remaining a nuclear power were not limited to direct budgetary outlays, since decision makers had to consider the economic opportunity costs of pursuing proliferation, namely forgone financial compensation from Russia and loss of Western aid and markets. Consistent with neorealism’s conception of state preferences, Mearsheimer denied that these substantial economic costs would dissuade Ukraine from maintaining a nuclear deterrent.(102)
With respect to the Ukrainian case, Steven Miller directly rebutted Mearsheimer’s hypothesis by employing an argument that is consonant with the postclassical realist conception of state behavior. Miller underscored that “despite the myth that nuclear weapons are cheap, they are in fact quite expensive for most states” and “given Ukraine’s economic needs and constraints, there will surely be incentives to minimize the resources devoted to defense, and the sums associated with a medium nuclear capability will surely be painful for Kiev.”(103) Retaining the nuclear deterrent was costly due to high maintenance requirements; in particular, the liquid fuel for Ukraine’s SS-19 intercontinental ballistic missiles made them particularly troublesome and costly to maintain.(104) The economic costs were not restricted to direct budgetary outlays; as Miller emphasized, pursuing proliferation would also mean “prospects for aid from and trade with the West would be harmed, a setback that has security implications given the importance of economic strength to national power.”(105) In the end, the United States pledged over $900 million in direct financial assistance to Ukraine to renounce its nuclear weapons,(106) as well as additional equipment and resources to aid Ukraine in the process of dismantling its nuclear stockpile.(107) Russia also offered Ukraine extensive economic incentives to forgo proliferation, including forgiveness of Ukraine’s multibillion dollar oil and gas debt to Russia.(108) Moreover, Ukraine was pledged substantial future transfers of nuclear fuel (estimated at a value of $1 billion) for its civilian reactors as part of a trilateral agreement with Russia and the United States.(109) Ultimately, therefore, a decision to maintain Ukraine’s nuclear deterrent would have resulted in a significant reduction in the country’s economic capacity due to (1) direct budgetary outlays, (2) loss of aid from Russia and the West, and (3) forgone integration into global markets.
In the end, although security tensions with Russia were certainly considerable, the huge economic costs to Ukraine of maintaining a nuclear deterrent were apparently even more significant. Ukraine has now ceased to be a nuclear power altogether – in early June 1996 the last of its nineteen hundred nuclear warheads were sent to Russia for destruction.(110) This Ukrainian decision to renounce its nuclear deterrent to advance its economic capacity is a marked anomaly for neorealism.
3. Cooperation Hypothesis: Regional Trade Blocs in the Developing World
The third important neorealist hypothesis is that states will be very reluctant to cooperate due to fears about how the gains will be distributed.(111) As Waltz argues, “States do not willingly place themselves in situations of increased dependence. In a self-help system, considerations of security subordinate economic gain to political interest.”(112) It would be a caricature, however, to say that neorealists regard international cooperation as impossible – they merely view it as greatly constrained.
With respect to current circumstances, the neorealist perspective suggests that developing countries will be very unlikely to pursue cooperation, especially given that security issues are often quite salient in this region.(113) Yet, many cooperative efforts have been initiated in the developing world in recent years, including the Southern Cone Common Market (Mercosur), the Andean Pact, the ASEAN Free Trade Agreement (AFTA), and the Central American Common Market (CACM), to name a few.(114) Significant security issues exist within all of these organizations: (1) within the Andean Pact, Peru and Ecuador engaged in direct military hostilities in late 1994; (2) within ASEAN, defense expenditures have increased dramatically in recent years and several serious territorial disputes exist among its members, most notably over the oil-rich Spratly Islands; and (3) there is a history of strong military rivalry between Brazil and Argentina within Mercosur and also between El Salvador and Honduras in CACM. The decision of these developing countries to initiate attempts at cooperation despite these security issues significantly contradicts neorealism.
In contrast, although postclassical realism sees states as being constrained from cooperating when security issues are salient, cooperation is still regarded as being feasible if the gains in economic capacity are even more significant than the potential security risks. For many developing countries, it does appear the economic benefits of cooperation are significantly higher in the current international environment compared to earlier periods. Specifically, being a member of a bloc: (1) augments negotiating power vis-a-vis larger economic actors that advance assertive unilateral trade policies; (2) acts as a “safety net” – regional trade partners could serve as alternative export markets if the European Union and/or the North American Free Trade Agreement turn aggressively protectionist; (3) enhances the chance of attracting foreign direct investment; and (4) allows member states to reduce transaction costs and acquire economies of scale at a time when the number and efficiency of exporters have increased dramatically in recent years. For many developing countries, engaging in regional cooperation can thus help promote international competitiveness. For the developing country trade pacts mentioned earlier, these four potential economic benefits of cooperation appear to supersede the constraining impact of relative gains concerns, thereby making cooperation possible.
The decision of these developing countries to pursue cooperation with potential rivals is incompatible with neorealism’s underlying assumptions about state behavior. In contrast, for postclassical realism, such behavior is consistent with the view that rational states make trade-offs and will favor economic capacity over security concerns in situations where the potential for enhanced economic competitiveness from regional cooperation outweighs the probability of security losses.
What Should Neorealists Do Now?
This brief empirical review is intended to demonstrate the usefulness of pitting the two branches of realism against each other in a direct debate, given that they produce such different hypotheses. Although the empirical analysis presented here is not definitive, one might reasonably ask how neorealists should respond, especially given that the cases reviewed represent current empirical manifestations of neorealism’s three primary hypotheses.
Although these cases do not support neorealism, the theory’s proponents should not respond by engaging in post hoc explanations in light of empirical evidence that contradicts neorealism’s expectations; doing so greatly constrains theoretical progress. Waltz engages in such a post hoc attempt to explain why many states are now emphasizing economic capacity to a greater degree vis-a-vis military security than neorealism would lead us to expect. Waltz points to the impact of nuclear weapons, arguing that they “make balancing easy to do” and enable states “to concentrate attention on their economies rather than on their military forces.”(115) By arguing in this manner that many states are currently focusing on economic capacity because nuclear weapons have made military security easier to achieve, Waltz thus adheres to the original neorealist conception of state preferences and avoids the need to consider whether many current states may be emphasizing economic capacity because it can, in fact, rival military security in importance.
Waltz’s nuclear weapons argument is ultimately unsatisfactory. Even if Waltz is correct that nuclear weapons will reduce the importance of security concerns and allow states to focus more on economics, the fact remains that a great many states are emphasizing economic capacity to a greater degree than neorealism suggests despite lacking a nuclear deterrent. The cases outlined earlier are clear examples of this fact. Germany and Japan are focusing on economic capacity but do not possess nuclear weapons. Ukraine’s decision to pursue economic capacity did not result from the stabilizing impact of nuclear weapons – rather, the method to advance economic priorities was, in fact, centered around Ukraine’s unilateral relinquishment of nuclear weapons. Finally, none of the developing countries who are now cooperating have nuclear weapons, yet they are nevertheless collaborating. It would seem the dependent variable that Waltz seeks to explain – namely, the decision of many current states to focus on economic capacity to a greater extent than neorealist theory would lead us to expect – has occurred before his primary independent variable – the presence of nuclear weapons – has come into play. Ultimately, therefore, Waltz’s post hoc explanation is empirically implausible.
Do neorealists have any other options? Faced with empirical irregularities, a logical neorealist response would be to ask for more time to evaluate their hypotheses. A very different neorealist response would be to concede that states are acting contrary to their theory’s expectations but to declare that such states are (1) acting contrary to system incentives and (2) that these decisions will ultimately have negative repercussions due to the competitive tenets of the international system. However, both of these potential neorealist responses pose a considerable danger: neorealism’s proponents could employ such arguments in order to indefinitely extend the testing period of neorealism’s hypotheses. In this respect, Layne argues that Germany and Japan will balance the United States within a “reasonably short time,” but he then goes on to point out that we may have to wait for up to fifty years to see the behavior he expects.(116) In the end, it is incumbent on neorealists to provide a means by which to assess the usefulness of the theory’s hypotheses; neorealists cannot simply ask for more time without identifying indicators of the trends that they foresee.
Finally, neorealists might respond by arguing that their theory is very parsimonious and hence cannot be expected to be empirically accurate. Such a claim would be problematic in two respects. First, the preceding empirical analysis involves cases that directly address neorealism’s primary hypotheses; yet, neorealism does not even identify the correct tendency in these examples. Second, as Stephan Haggard argues, “the claim for superiority of a theory on the basis of its parsimony cannot stand alone. What use is a parsimonious theory that is wrong or that explains only a small portion of the variance? The issue, therefore, is not one of parsimony per se but of the trade-off between parsimony and explanatory power.”(117) Postclassical realism does represent a step away from Waltz’s sparse architecture, but it is not a dramatic shift away from parsimony. Specifically, postclassical realism is parsimonious in three important respects: (1) it focuses on material factors that function independently of shared social understandings and institutional characteristics; (2) it operates with a state-centric, unitary actor assumption; and (3) although not a structural theory, postclassical realism is systemic, since it focuses on international-level factors and does not examine domestic political variables. Hence, if the preceding empirical analysis is any indication, the gain in explanatory power of a postclassical realist approach may turn out to more than compensate for the relative loss of parsimony compared to neorealism.
Implications for Nonrealist Theories
Although I conducted this analysis within the confines of realism, the repercussions extend to nonrealist theories as well. Recognizing the possibility/probability distinction within realism helps to explain why the debate with nonrealist theories has so far been couched in highly competitive terms and also indicates potential for a more productive future dialogue.
Liberalism and Constructivism
Up to this point the exchange between neorealism and nonrealist theories such as constructivism and liberalism has not been very productive. The essential reason is that neorealism’s possibilistic focus guarantees that debate with such probabilistic, nonrealist theories will necessarily be a zero-sum exercise. Neorealists cannot recognize the validity of any of the factors nonrealist theories identify as influencing the probability of conflict. To admit that such factors have credence would be a tacit recognition of the deficiency of neorealism’s worst-case/possibilistic assumption.
Recognizing that neorealism’s possibilistic focus automatically leads to a zero-sum contest with probabilistic, nonrealist theories helps to explain why, for example, neorealists have been so averse to accepting the validity of the democratic peace proposition, even as empirical evidence mounts in support of this finding.(118) Significantly, democratic peace proponents such as Bruce Russett do not argue that war between democracies is impossible, but rather that shared democracy significantly reduces the probability of war.(119) Liberals also do not posit that shared democracy is the only factor influencing the likelihood of conflict, just that it is an important factor that must be taken into account.(120) The democratic peace finding is unimpressive to neorealists precisely because of their possibilistic focus. For neorealists, the relevant question is not whether shared democracy reduces the probability of conflict, but instead whether the possibility that democracies might engage in war continues to exist. This makes it easier to understand why the existence of even a small number of potential outliers to the democratic peace proposition is regarded by neorealists as such a serious deficiency.(121) Significantly, Russett does not rule out the potential existence of outliers, asserting that it should be “enough to say . . . that wars between democracies are at most extremely rare events.”(122) For neorealists, however, the existence of even a small number of potential outliers is all that matters; outliers would mean that war between democracies is still possible and hence, in neorealism’s worst-case framework, that states will continue to focus on the capabilities of potential aggressors irrespective of institutional characteristics to be prepared for all contingencies.
In contrast, the probabilistic focus of postclassical realism is more amenable to productive discussion with nonrealist theories. Analyses based on postclassical realism, on the one hand, and those of nonrealist theories such as liberalism and constructivism, on the other hand, differ in important respects, notably as to whether they emphasize material or nonmaterial factors. Yet, this should not overshadow that the two sides ask similar questions and are united in the view that factors exist besides the distribution of military capabilities that cause the probability of conflict to systematically vary. Both sides are thus critiques of neorealism, albeit from different perspectives. The analyses of postclassical realism and those of liberalism or constructivism are not necessarily antithetical; indeed, each may turn out to complement the others by contributing different elements of the answers to certain questions that no single theory can satisfactorily solve on its own. In this respect, some nonrealists are willing to admit that realist-based analyses continue to have an invaluable role to play.(123) The probabilistic underpinnings of postclassical realism make feasible the adoption of a reciprocal stance with respect to nonrealist theories.
Domestic-level Explanations
As with the dialogue with liberalism and constructivism, neorealists have couched the interchange with domestic-level theories in essentially zero-sum terms. In large part, the debate with domestic-level theories has been characterized in this manner because of neorealism’s worst-case/possibilistic focus. Neorealism’s worst-case perspective leads to the view that states have highly constrained policy options: states always seek to balance the military capabilities of potential aggressors, military security trumps other goals (such as economic capacity) where conflict exists between them, and so on. Neorealism thus admits essentially no need to look at domestic-level processes because actors are understood to have minimal discretion regarding the strategies they adopt. Within the neorealist framework, the only consequential impact domestic politics can have is to prevent the state from taking steps, such as balancing, to maximize military security. This explains the zero-sum nature of the debate with domestic-level theories – neorealism admits that domestic factors are important only to the extent that they prevent the state from responding to international incentives.
In contrast, postclassical realism views states as having a wider range of policy options; it is emphasized that states often make trade-offs between different priorities, notably between military security and economic capacity. As a result, postclassical realism accepts an important, non-zero-sum role for domestic-level understandings under some circumstances. Systemic constraints will often cause states to discern a clear strategy to advance their overriding objectives; this will especially be the case when the international influences on states are very strong and/or when the state is relatively susceptible to these systemic factors. At other times, the strategy to pursue the state’s primary objectives may be more ambiguous. In circumstances where the preferred strategy is unclear, the mechanism by which particular policies are selected may be significantly influenced by domestic-level bargaining.(124) In such situations, the systemic focus of postclassical realism may need to be supplemented by domestic-level analyses to sufficiently comprehend international behavior. In this respect, Michael Mastanduno, David Lake, and G. John Ikenberry advance a framework for understanding how decision makers balance the pursuit of power against domestic strategies for achieving this objective.(125)
Neorealism and postclassical realism thus differ markedly regarding the role that domestic-level arguments have in explaining international behavior. For postclassical realism, analyses should be undertaken in stages: first, at the systemic level, and then, if necessary, complemented and extended at the domestic level. An essential difficulty with neorealism is that it is not self-conscious of its own limitations; the theory admits no need for undertaking analyses in stages in this manner. Rather than regarding unit-level analyses as potentially useful clarifying devices, neorealists instead tend to regard domestic understandings as either competitive to neorealism or essentially spurious. Both perspectives are too extreme and impede progress in understanding international behavior, but these reactions are understandable once one recognizes that neorealism’s worst-case/possibilistic focus leads it to view states as possessing limited discretion regarding their choice of strategies.
This is not to suggest that Waltz is wrong to assert that focusing on international forces is valuable; a systemic focus is useful, but this does not necessarily mean that domestic-level factors should be ignored. A more reasonable justification for adopting a systemic focus is that concentrating first on international forces often provides a sufficient understanding of behavior, both for analysts as well as for policymakers. Moreover, in those circumstances where a systemic analysis is insufficient on its own, such an initial focus is nevertheless still useful because it helps ensure as detailed an understanding as possible of the international pressures on actors. Ultimately, a thorough understanding of external constraints is, as Keohane emphasizes, an essential precursor to analysis of international behavior at the domestic level.(126)
Conclusion
My analysis partitions realism into two branches by revealing latent disputes within the theory regarding a series of assumptions about state behavior. In particular, realism diverges regarding whether the mere possibility of conflict conditions decision making, as neorealism assumes, or whether actors decide between policy options based on the probability of conflict, as postclassical realism asserts. Neorealists view the international system as a relentless competition for security; in contrast, postclassical realism is agnostic regarding security competition in the international system: within postclassical realism, the strength of security pressures fluctuates according to a variety of material factors besides the distribution of capabilities, namely technology, geography, and international economic pressures. This possibility/probability distinction harbors two related disagreements within realism. First, neorealism views states as discounting the future to a greater degree than is assumed within postclassical realism. Second, realism is divided regarding state preferences: neorealism emphasizes military security as the overriding priority, whereas postclassical realism maintains that states ultimately pursue power – a concept that contains an inherent tension between military security and economic capacity, where neither goal is necessarily subordinate to the other.
In differentiating these two branches of realism, I do not mean to imply that the theory necessarily needs to be permanently split into two camps; however, at this point the two branches must be distinguished because, as this article shows, realism does not have a unified set of assumptions about state behavior. As long as realists continue to derive their hypotheses from different sets of assumptions, the theory must be divided in order to be clear about exactly what is being tested.
As the underlying divergence within realism becomes apparent, the two branches may engage in an empirical “duel” in order to determine which set of assumptions about state behavior is relatively most useful. Alternatively, the two branches of realism may refrain from engaging in an extended empirical debate. Until now, the worst-case/possibilistic underpinnings of neorealism have been hidden from view and have thus far escaped critical examination. International relations scholars – including many neorealists – may simply come to view this worst-case/possibilistic perspective as being too inflexible.
If neorealists do conclude that their worst-case/possibilistic assumption is too restrictive – and that a probabilistic focus is more useful – this would not mean that neorealism’s current view of state behavior is completely wrong. To be clear, to say that actors are conditioned by the probability of conflict does not deny that actors will under some circumstances behave in a highly cautious and conservative manner, that is, take extensive and costly measures to ensure their military security. Neorealism and postclassical realism are differentiated in terms of how often – and not whether – actors are expected to behave in a highly cautious and conservative manner: the former says always, the latter, sometimes. Neorealism appears unsatisfactory not because the theory’s conception of state behavior is wrong all of the time, but rather because the theory inflexibly claims that this perspective is operative under all circumstances. If, in the end, neorealists decide a shift toward a probabilistic perspective is appropriate, postclassical realism would essentially come to subsume neorealism.
Partitioning realist theory as suggested holds out hope for progress in three respects. First, dividing realism should lead to a more precise conceptualization of the theory and help provide a better understanding of the varying impact of different material factors on state behavior.
Second, dividing realism in this manner makes it easier to understand exactly why interchanges with nonrealist theories have been so disappointing and suggests the possibility of more productive future dialogue. If realism were to settle on a probabilistic focus, it would be possible to move beyond arguments that respectively assert the universal primacy of either realist or nonrealist explanations and to start exploring the conditions under which each is more or less helpful and, in turn, how they can sometimes usefully supplement each other.
Finally, the analysis serves as a counterpoint to the pessimism that is typical of realist theory. Neorealism’s worst-case focus and emphasis on capabilities to the exclusion of other variables leads its proponents to see little hope for progress in international relations. However, it would be premature to conclude that all realist analyses view international relations as necessarily consisting of relentless security competition in which defensive vigilance is the only hope for forestalling aggression. The range of material factors that postclassical realism identifies as influencing the likelihood of conflict are arguably currently quite favorable for low security tensions among the major powers. Unlike neorealism, therefore, postclassical realism provides some reason to be sanguine that the current positive state of security relations between the major powers need not be merely a temporary anomaly.
This material is based on work supported under a National Science Foundation Graduate Fellowship. I thank Deborah Jordan Brooks, Joseph Grieco, Ian Hurd, David Lumsdaine, Sylvia Maxfield, John Odell, Barry O’Neill, Jonathan Rodden, Frances Rosenbluth, Allan Stam, Stephen Skowronek, Nigel Thalakada, Shaoguang Wang, Brad Westerfield, the members of Yale University’s International Relations Reading Group, the anonymous reviewers for International Organization, and especially Bruce Russett and Alex Wendt for their helpful comments on earlier drafts.
Notes
1. See, for example, Gilpin 1984, 290-91; Keohane 1986, 164-65; Grieco 1995, 27; Vasquez 1983, 26-28; Stein 1990, 4-5; Kapstein 1995, 753; Keohane and Nye 1977, 23-24; Grieco 1990, 3-4; and Gilpin 1996, 7-8.
2. See, for example, Snyder 1991, 11-12; Lynn-Jones and Miller 1995, x-xiii; Deudney 1993, 8; Glaser 1994-95,378-81; Mastanduno 1996, 2; Wait 1992, 474; Mearsheimer 1994-95, fns. 20, 27; Frankel 1996, xv-xviii; Grieco 1997; Wohlforth 1993, 11-14; Zakaria 1992, 190-93; Miller 1996; and Jervis 1993, 55-56.
3. Some analysts identify the need to make distinctions within realism but emphasize only those assumptions that realists hold in common; see, for example, Meatsheimer 1994-95, 11-13; Lynn-Jones and Miller 1995, ix-x; Glaser 1994-95, 54-55; Wait 1992, 473; and Mastanduno 1996, 3-5. Other scholars outline divisions within realism only with respect to particular issue areas; see, for example, Deudney 1993; Miller 1996; and Desch 1996. Some suggest that realists may diverge regarding certain assumptions but do not address whether realism should be divided on this basis; see, for example, Grieco 1997; and Wayman and Diehl 1994, 9. The division between “aggressive” and “defensive” realists is the most comprehensive outlined so far; see, for example, Snyder 1991, 11-12; Zakaria 1992, 190-93; and Frankel 1996, xv-xviii. Section two argues that this aggressive/defensive distinction is useful but is reflective of the deeper divergence within realism over assumptions outlined in this article.
4. Waltz 1979.
5. In choosing the term “postclassical realism” to distinguish the two branches of realism, I do not mean to imply that the theory necessarily must be permanently divided, just that dividing realism is currently necessary to improve and clarify the theory. The rationale for selecting the term is outlined in section two.
6. A “systemic” approach to understanding international behavior is sometimes seen as being interchangeable with a “structural” one. I use the term systemic to refer to analyses that focus on international influences and do not examine domestic political variables. “Structural” analyses – which focus on the role of polarity – are best seen as a form of, rather than being synonymous with, “systemic” theory.
7. For arguments discussing the distinction between possibility and probability in a different context, see Wendt 1992, 404; and Keohane 1993, 282-83.
8. Mearsheimer 1994-95, 10.
9. Waltz 1989, 43.
10. Waltz 1979, 111, 102.
11. Waltz 1959, 188.
12. Keohane and Martin and Wendt reach a similar assessment, although they attribute this worst-case assumption to realism in general; Keohane and Martin 1995, 43; and Wendt 1992, 404. As will be seen, not all realists adopt such a worst-case focus.
13. Mearsheimer 1994-95, 12.
14. Ibid., 11.
15. Liberman 1993, 125.
16. Waltz 1959, 201.
17. Waltz 1979, 114.
18. See Powell, 1991, 1993; and Wendt 1992, 1995. Although Wendt and Powell advance the same underlying argument that anarchy by itself does not shape decisions regarding balancing and cooperation, they point to different underlying factors as affecting the perceived likelihood of force: Powell examines the impact of military technology, whereas Wendt outlines the importance of intersubjective understandings.
19. Waltz 1959, 16-42.
20. Morgenthau 1946, 192-94.
21. Mearsheimer 1990, 12.
22. Waltz 1959, 188.
23. For a clear example of how neorealists rely on human nature, see Fischer’s analysis of feudal Europe; Fischer 1992, 465.
24. Waltz 1979, 91.
25. Ibid., 118.
26. For a good summary as to why an ironclad trade-off does not exist between military security and economic capacity under all circumstances, see Friedberg 1989.
27. See Gaddis 1982, 58, 133-35; Kennedy 1983; and Wohlforth 1994-95.
28. See Kennedy 1987; and Gilpin 1981.
29. Rosecrance 1981, 693. Although Rosecrance did not specify the meaning of “economics” in the preceding quote, he appears to have been referring – correctly – to the fact that Waltz ignores international economic interactions as unit-level processes; see Rosecrance 1982, 682. This is an important critique of Waltz’s theory; unfortunately it is clouded by Rosecrance’s use of the blanket term economics rather than referring more specifically to economic interactions between states.
30. Mearsheimer 1992, 222 (emphasis added).
31. Waltz 1959, 206.
32. Waltz 1979, 92.
33. Waltz argues that states may “seek goals that they value more highly than survival; they may, for example, prefer amalgamation with other states to their own survival in form”; Waltz 1979, 92.
34. Waltz, 1959, 221, 160.
35. Waltz 1986, 331.
36. Baldwin 1995.
37. Waltz 1979, 79-80.
38. Examples of analyses that are explicitly based on expected utility include Gilpin 1981; Huth and Russett 1984; Bueno de Mesquita 1981; and Lipson 1984.
39. Keohane 1993, 282.
40. Luce and Raiffa 1957, 278-89.
41. ibid., 282.
42. See Farnham 1994; and Tversky and Kahneman 1986.
43. Levy 1994, 14.
44. Ibid.
45. Gilpin 1984, 290.
46. Neorealists recognize that domestic factors – such as hyper-nationalism – can also have an important impact, but such factors are excluded from their theory. Although some neorealists have attempted to incorporate domestic factors such as hyper-nationalism into neorealism (see, for example, Mearsheimer 1990, 21), doing so would be a post hoc, degenerative means of increasing the theory’s explanatory power; on this point see Haggard 1991, 421; and Wendt 1995, 78-79.
47. Russett 1993, 1995.
48. Keohane and Martin 1995.
49. Wendt 1992, 1995.
50. See, for example, Jervis 1978; Quester 1977; and Glaser 1994-95.
51. Buzan 1993, 72-80.
52. Gaddis 1987, 232.
53. See Van Evera 1990-91, 14-16; and Kaysen 1990, 49, 53.
54. See, for example, Walt 1987, 23-24; Jervis 1978, 183-86, 194-96; and Bueno de Mesquita 1981, 40-44.
55. See, for example, Krasner 1978; and Jervis 1978, 179.
56. See, for example, Gilpin 1977; Wohlforth 1994-95; and Kennedy 1987.
57. See, for example, Gilpin 1981, especially 132-33, 138-40; Huntington 1993; Lake 1996; and Pollins 1994.
58. Wendt and Friedheim 1995, 691-92.
59. See, for example, Glaser 1994-95; Buzan 1993; and Walt 1987.
60. Snyder 1991, 11-12.
61. Ibid., 12.
62. Frankel 1996, xv-xvi.
63. Mearsheimer 1990, 12.
64. Gilpin 1981.
65. See, for example, Ruggie 1983; and Nye 1988, 245.
66. See, for example, Grieco 1990, 39; Ordeshook and Niou 1991, fn. 1; and Grunberg and Risse-Kappen 1992, 113.
67. Gilpin 1984, 290.
68. Gilpin 1981, 13. Gilpin’s conception of power will be used throughout this article; however, technological capacity is understood to be an element of economic capacity. I recognize – as does Gilpin – the limitations of such a definition of power, since it leaves out a myriad of nonmaterial factors, such as prestige, public morale, and leadership qualities; see Gilpin 1981, 13-14.
69. Joseph Grieco similarly points to this ambiguity within Gilpin’s analysis – and within realism in general – regarding whether states pursue power or security, although Grieco frames the power/security distinction differently than I do here; see Grieco 1997.
70. Waltz 1989, 40.
71. Waltz 1979, 126.
72. Mearsheimer 1994-95, 12; see also Mearsheimer 1990, 12.
73. Mearsheimer 1994-95, 13.
74. See Kennedy 1987; and Gilpin 1981.
75. Gilpin 1984, 290-91.
76. For this interpretation of Gilpin’s analysis, see Grieco 1990, 39.
77. Gilpin 1981, 19.
78. Gilpin 1981, 22 (emphasis added).
79. Mastanduno, Lake, and Ikenberry 1989, 462.
80. See Gilpin 1981, 194-97; and Kennedy 1987; see also Wohlforth 1994-95.
81. Morgenthau 1946, 193.
82. Some modem realists have returned to this classical realist conception of power maximization as being a useful assumption, at least as applied to certain states; see Zakaria 1992, 194; and Schweller 1994, 104-105.
83. For this perspective on power, see Gilpin 1996, 6; and Gilpin 1975, 23.
84. Gilpin 1981, 51.
85. See, for example, Krasner 1976; Lake 1984; Huntington 1993; and Gowa 1994.
86. This is the foundation of the New Institutional Economics; see, for example, North 1990; Alt and Shepsle 1990; and Kang 1995, 563-66.
87. See, for example, Hirschman 1980; and Krasner 1978.
88. See, for example, Kennedy 1987; and Gilpin 1981.
89. An excellent recent example is the relative/absolute gains debate; see Powell 1994.
90. Waltz 1979, especially 118-21, 128.
91. See Waltz 1993; and Layne 1995.
92. Layne 1995, fn. 34.
93. Waltz 1991, 669.
94. Layne 1995, fn. 12.
95. Layne 1995, 134-35.
96. SIPRI 1995, 390.
97. New York Times, 29 November 1995, A9.
98. Regarding Japan, see U.S. Department of State Dispatch, 22 April 1996, 200; and Far Eastern Economic Review, 2 May 1996, 14-16. Regarding Germany, see Smyser 1994; and Weekly Compilation of Presidential Documents, 13 February 1995, 216-24.
99. Waltz 1979, 127.
100. See Mearsheimer 1990; and Waltz 1993.
101. Mearsheimer 1993, 58, 66.
102. Mearsheimer maintained that “The United States should recognize that Ukraine is going to be a nuclear power, irrespective of what the West does”; Mearsheimer 1993, 66.
103. Miller 1993, 77.
104. Reiss 1995, 107, 126
105. Miller 1993, 79.
106. See New York Times, 4 March 1994, A3; and New York Times, 23 November 1994, A8.
107. Arms Control Today, October 1994, 21.
108. Reiss 1995, 117, 129.
109. Arms Control Today, Jan.-Feb. 1994, 12-20.
110. Economist, 8 June 1996, 6.
111. Waltz 1979, 104-107.
112. Ibid., 107.
113. Powell argues that states will be concerned about relative gains “when the possible use of force is at issue”; Powell 1991, 1316.
114. See Haggard 1995.
115. Waltz 1993, 74, 52.
116. Layne 1995, 142, 176.
117. Haggard 1991, 417.
118. For the democratic peace argument, see, for example, Russett 1993, 1995. For neorealist critiques, see, for example, Layne 1994; Mearsheimer 1990, 49-51; and Waltz 1993, 78.
119. Russett 1996, 178.
120. Russett 1995, 166.
121. See Waltz 1993, 78; Mearsheimer 1990, 51 and Layne 1994, 40-44.
122. Russett 1995, 169.
123. See, for example, Russett 1996, 176.
124. That strategy choice will often reflect both international incentives as well as domestic-level bargaining is the essence of Putnam’s argument regarding two-level games; Putnam 1988.
125. Mastanduno, Lake, and Ikenberry 1989.
126. See Keohane 1984a, 25-26; and Keohane 1984b, 16.
References
Alt, James, and Kenneth Shepsle. 1990. Perspectives on Positive Political Economy. Cambridge: Cambridge University Press.
Baldwin, David. 1995. Security Studies and the End of the Cold War. World Politics 48:117-41.
Bueno de Mesquita, Bruce. 1981. The War Trap. New Haven, Conn.: Yale University Press.
Buzan, Barry. 1993. Rethinking System and Structure. In The Logic of Anarchy: Neorealism to Structural Realism, by Barry Buzan, Charles Jones, and Richard Little, 19-80. New York: Columbia University Press.
Desch, Michael. 1996. Why Realists Disagree About the Third World (and Why They Shouldn’t). Security Studies 5:358-84.
Deudney, Daniel. 1993. Dividing Realism: Structural Realism and Security Materialism on Nuclear Security and Proliferation. Security Studies 2:7-36.
Farnham, Barbara. 1994. Avoiding Losses/Taking Risks: Prospect Theory and International Conflict. Ann Arbor: University of Michigan Press.
Fischer, Markus. 1992. Feudal Europe, 800-1300: Communal Discourse and Conflictual Practices. International Organization 46:427-66.
Frankel, Benjamin. 1996. Restating the Realist Case: An Introduction. Security Studies 5:ix-xx.
Friedberg, Aaron. 1989. The Political Economy of American Strategy. World Politics 41:381-406.
Gaddis, John Lewis. 1982. Strategies of Containment: A Critical Appraisal of Postwar American National Security Policy. New York: Oxford University Press.
—–. 1987. The Long Peace: Inquiries into the History of the Cold War. New York: Oxford University Press.
Gilpin, Robert. 1975. U.S. Power and the Multinational Corporation. New York: Basic Books.
—–. 1977. Economic Interdependence and National Security in Historical Perspectives. In Economic Issues and National Security, edited by Klaus Knorr and Frank Trager, 19-66. Lawrence, Kans.: Regents Press of Kansas.
—–. 1981. War and Change in World Politics. Cambridge: Cambridge University Press.
—–. 1984. The Richness of the Tradition of Political Realism. International Organization 38:287-304.
—–. 1996. No One Loves a Political Realist. Security Studies 5:3-28.
Glaser, Charles. 1994-95. Realists as Optimists: Cooperation as Self-Help. International Security 19:50-90.
Gowa, Joanne. 1994. Allies, Adversaries, and International Trade. Princeton, N.J.: Princeton University Press.
Grieco, Joseph. 1990. Cooperation Among Nations: Europe, North America, and Non-Tariff Barriers to Trade. Ithaca, N.Y.: Cornell University Press.
—–. 1995. The Maastricht Treaty, Economic and Monetary Union, and the Neo-Realist Research Programme. Review of International Studies 21:21-40.
—–. 1997. Realist International Theory and the Study of World Politics. In New Thinking in International Relations, edited by Michael Doyle and G. John Ikenberry. New York: Westview Press.
Grunberg, Isabelle, and Thomas Risse-Kappen. 1992. A Time of Reckoning? Theories of International Relations and the End of the Cold War. In The End of the Cold War: Evaluating Theories of International Relations, edited by Pierre Allan and Kjell Goldman, 104-46. Dordrecht, Netherlands: Martinus Nijhoff.
Haggard, Stephan. 1991. Structuralism and Its Critics: Recent Progress in International Relations Theory. In Progress in Postwar International Relations, edited by Emanuel Adler and Beverly Crawford, 403-37. New York: Columbia University Press.
—–. 1995. Developing Nations and the Politics of Global Integration. Washington, D.C.: The Brookings Institution.
Hirschman, Albert. 1980. National Power and the Structure of Foreign Trade. Berkeley: University of California Press.
Huntington, Samuel. 1993. Why International Primacy Matters. International Security 17:68-83.
Huth, Paul, and Bruce Russett. 1984. What Makes Deterrence Work: Cases from 1900-1980. World Politics 36:495-526.
Jervis, Robert. 1978. Cooperation Under the Security Dilemma. World Politics 30:167-214.
—–. 1993. International Primacy: Is the Game Worth the Candle? International Security 17:52-67.
Kang, David. 1995. South Korea and Taiwanese Development and the New Institutional Economics. International Organization 49:555-87.
Kapstein, Ethan. 1995. Is Realism Dead? The Domestic Sources of International Politics. International Organization 49:751-74.
Kaysen, Carl. 1990. Is War Obsolete? International Security 14:42-64.
Kennedy, Paul. 1983. Strategy vs. Finance in Twentieth Century Britain. In Strategy and Diplomacy: 1870-1945, edited by Paul Kennedy, 87-106. London: George Allen and Unwin.
—–. 1987. The Rise and Fall of the Great Powers: Economic Change and Military Conflict from 1500 to 2000. New York: Random House.
Keohane, Robert. 1984a. After Hegemony: Cooperation and Discord in the World Political Economy. Princeton, N.J.: Princeton University Press.
—–. 1984b. The World Political Economy and the Crisis of Embedded Liberalism. In Order and Conflict in Contemporary Capitalism, edited by John Goldthorpe, 15-38. Oxford: Claredon Press.
—–. 1986. Theory of World Politics: Structural Realism and Beyond. In Neorealism and Its Critics, edited by Robert Keohane, 158-203. New York: Columbia University Press.
—– 1993. Institutionalist Theory and the Realist Challenge After the Cold War. In Neoliberalism and Neorealism: The Contemporary Debate, edited by David Baldwin, 269-300. New York: Columbia University Press.
Keohane, Robert, and Lisa Martin. 1995. The Promise of Institutionalist Theory. International Security 20:39-51.
Keohane, Robert, and Joseph Nye. 1977. Power and Interdependence: World Politics in Transition. Boston: Little Brown.
Krasner, Stephen. 1976. State Power and the Structure of International Trade. World Politics 28:317-43.
—–. 1978. Defending the National Interest: Raw Materials Investments and U.S. Foreign Policy. Princeton, N.J.: Princeton University Press.
Lake, David. 1984. Beneath the Commerce of Nations: A Theory of International Economic Structures. International Studies Quarterly 28:143-70.
—–. 1996. Anarchy, Hierarchy, and the Variety of International Relations. International Organization 50:1-34.
Layne, Christopher. 1994. Kant or Cant: The Myth of the Democratic Peace. International Security 19:5-49.
—–. 1995. The Unipolar Illusion: Why New Great Powers Will Rise. In The Perils of Anarchy: Contemporary Realism and International Security, edited by Michael E. Brown, Sean M. Lynn-Jones, and Steven E. Miller, 130-76. Cambridge, Mass.: MIT Press.
Levy, Jack. 1994. An Introduction to Prospect Theory. In Avoiding Losses/Taking Risks: Prospect Theory and International Conflict, edited by Barbara Farnham, 7-22. Ann Arbor: University of Michigan Press.
Liberman, Peter. 1993. The Spoils of Conquest. International Security 18:125-53.
Lipson, Charles. 1984. International Cooperation in Economic and Security Affairs. World Politics 37:1-23.
Luce, R. Duncan, and Howard Raiffa. 1957. Games and Decisions: Introduction and Critical Survey. New York: John Wiley & Sons.
Lynn-Jones, Sean M., and Steven E. Miller. 1995. Preface. In The Perils of Anarchy: Contemporary Realism and International Security, edited by Michael E. Brown, Sean M. Lynn-Jones, and Steven E. Miller, ix-xxi. Cambridge, Mass.: MIT Press.
Mastanduno, Michael. 1996. Security Engagement and Economic Competition: Realism and United States Strategy After the Cold War. Unpublished manuscript, Dartmouth College, Hanover, N.H.
Mastanduno, Michael, David Lake, and G. John Ikenberry. 1989. Toward a Realist Theory of State Action. International Studies Quarterly 33:457-74.
Mearsheimer, John. 1990. Back to the Future: Instability in Europe After the Cold War. International Security 15:5-56.
—–. 1992. Disorder Restored. In Rethinking America’s Security: Beyond Cold War to New World Order, edited by Graham Allison and Gregory F. Treverton, 213-37. New York: W.W. Norton.
—–. 1993. The Case for a Ukrainian Nuclear Deterrent. Foreign Affairs 72:50-66.
—–. 1994-95. The False Promise of International Institutions. International Security 19:5-49.
Miller, Benjamin. 1996. Competing Realist Perspectives on Great Power Crisis Behavior. Security Studies 5:309-58.
Miller, Steven. 1993. The Case Against a Ukrainian Nuclear Deterrent. Foreign Affairs 72:67-80.
Morgenthau, Hans. 1946. Scientific Man vs. Power Politics. Chicago: University of Chicago Press.
North, Douglass. 1990. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge University Press.
Nye, Joseph. 1988. Neorealism and Neoliberalism. World Politics 40:235-251.
Ordeshook, Peter, and Emerson Niou. 1991. Realism vs. Neoliberalism: A Formulation. American Journal of Political Science 35:481-511.
Pollins, Brian. 1994. Cannons and Capital: The Use of Coercive Diplomacy by Major Powers in the Twentieth Century. In Reconstructing Realpolitik, edited by Frank Wayman and Paul Diehl. Ann Arbor: University of Michigan Press.
Powell, Robert. 1991. Absolute and Relative Gains in International Relations Theory. American Political Science Review 85:1303-320.
—–. 1993. Guns, Butter, and Anarchy. American Political Science Review 87:115-32.
—–. 1994. Anarchy in International Relations Theory: The Neorealist-Neoliberal Debate. International Organization 48:313-44.
Putnam, Robert. 1988. Diplomacy and Domestic Politics: The Logic of Two-Level Games. International Organization 42:427-60.
Quester, George. 1977. Offense and Defense in the International System. New York: John Wiley and Sons.
Reiss, Mitchell. 1995. Bridled Ambition: Why Countries Constrain Their Nuclear Capabilities. Baltimore, Md.: Johns Hopkins University Press.
Rosecrance, Richard. 1981. International Theory Revisited. International Organization 35:691-713.
—–. 1982. Reply to Waltz. International Organization 36:682-85.
Ruggie, John Gerard. 1983. Continuity and Transformation in the World Polity: Toward a Neorealist Synthesis. World Politics 35:261-85.
Russett, Bruce. 1993. Grasping the Democratic Peace: Principles for a Post-Cold War World. Princeton: Princeton University Press.
—–. 1995. The Democratic Peace: “And Yet It Moves.” International Security 19:164-75.
—–. 1996. Counterfactuals About War and Its Absence. In Counterfactual Thought Experiments in Worm Politics: Logical, Methodological, and Psychological Perspectives, edited by Philip Tetlock and Aaron Belkin, 171-187. Princeton: Princeton University Press.
Schweller, Randall. 1994. Bandwagoning for Profit: Bringing the Revisionist State Back in. International Security 19:72-107.
SIPRI (Stockholm International Peace Research Institute). 1995. SIPRI Yearbook 1995: Armaments, Disarmament, and International Security. Oxford: Oxford University Press.
Smyser, W. R. 1994. Dateline Berlin: Germany’s New Vision. Foreign Policy 97:140-57.
Snyder, Jack. 1991. Myths of Empire: Domestic Politics and International Ambition. Ithaca, N.Y.: Cornell University Press.
Stein, Arthur. 1990. Why Nations Cooperate: Circumstance and Choice in International Relations. Ithaca, N.Y.: Cornell University Press.
Tversky, Amos and Daniel Kahneman. 1986. The Framing of Decisions and the Psychology of Choice. In Rational Choice, edited by Jon Elster, 123-41. New York: New York University Press.
Van Evera, Stephen. 1990-91. Primed for Peace: Europe After the Cold War. International Security 15:7-57.
Vasquez, John. 1983. The Power of Power Politics: A Critique. New Brunswick, N.J.: Rutgers University Press.
Walt, Stephen. 1987. The Origins of Alliances. Ithaca, N.Y.: Cornell University Press.
—–. 1992. Alliance, Threats, and U.S. Grand Strategy: A Reply to Kaufman and Labs. Security Studies 1:448-82.
Waltz, Kenneth. 1959. Man, the State, and War. New York: Columbia University Press.
—–. 1979. Theory of International Politics. Reading, Mass.: Addison-Wesley.
—–. 1986. Reflections on Theory of International Politics: A Response to My Critics. In Neorealism and Its Critics, edited by Robert Keohane, 322-45. New York: Columbia University Press.
—–. 1989. The Origins of War in Neorealist Theory. In The Origin and Prevention of Major Wars, edited by Robert Rotberg and Theodore Rabb, 39-52. Cambridge: Cambridge University Press.
—–. 1991. America as a Model for the World: A Foreign Policy Perspective. Political Science and Politics 24:667-70.
—–. 1993. The Emerging Structure of International Politics. International Security 18:44-79.
Wayman, Frank, and Paul Diehl. 1994. Realism Reconsidered. In Reconstructing Realpolitik, edited by Frank Wayman and Paul Diehl, 3-29. Ann Arbor: University of Michigan Press.
Wendt, Alexander. 1992. Anarchy Is What States Make of It: The Social Construction of Power Politics. International Organization 46:391-425.
—–. 1995. Constructing International Politics. International Security 20:71-81.
Wendt, Alexander, and Daniel Friedheim. 1995. Hierarchy Under Anarchy: Informal Empire and the East German State. International Organization 49:689-721.
Wohlforth, William. 1993. The Elusive Balance: Power and Perceptions During the Cold War. Ithaca, N.Y.: Cornell University Press.
—–. 1994-95. Realism and the End of the Cold War. International Security 19:91-129.
Zakaria, Fareed. 1992. Realism and Domestic Politics. International Security 17:177-98.
Source: International Organization, Vol. 51, no. 3 (Summer 1997)
