Political Realism has been described as the “oldest theory” of international politics, as well as the “dominant” one. Central to the realist tradition is the concept of “security.” Realism sees the insecurity of states as the main problem in international relations. It depicts the international system as a realm where “self-help” is the primary motivation; states must provide security for themselves because no other agency or actor can be counted on to do so. However, realists offer different explanations for why security is scarce, emphasizing a range of underlying mechanisms and causal factors such as man’s innate desire for power; conflicts of interest that arise between states possessing different resource endowments, economic systems, and political orders; and the “ordering principle” of international anarchy. They also propose numerous factors that can intensify or ameliorate the basic security problem, such as polarity, shifts in the overall balance of power, the “offense–defense balance,” and domestic politics. Several alternative approaches to international relations have challenged the basic realist account of the security problem, three of which are democratic peace theory, economic liberalism, and social constructivism. Furthermore, realism outlines various strategies that states can pursue in order to make themselves more secure, such as maximizing power, international alliances, arms racing, socialization and innovation, and institutions and diplomacy. Scholars continue to debate the historical roots, conceptual foundations, and predictive accuracy of realism. New avenues of research cover issues such as civil war, ethnic conflict, mass violence, September 11, and the Iraq War.
Political Realism is a philosophical approach to the study of politics – and especially international politics – that is widely regarded as the most enduring and influential tradition in the field. As Robert Keohane put it in 1983, “for over 2000 years, what Hans J. Morgenthau dubbed ‘Political Realism’ has constituted the principal traditions for the analysis of international relations in Europe and its offshoots in the New World” (Keohane 1983; also Walt 2003). Michael Doyle (1997) agrees, describing realism as the “oldest theory” of international politics but also the “dominant” one.
There are many different realist theories within that broad tradition, but each of them sees states as the central actors in world affairs and emphasizes that they coexist in an anarchic social order where there is no central authority to protect them from one another. As a result, realist theories see the insecurity of states (or in some cases, substate groups) as the central problem in international relations. In short, realism depicts the international system as a realm where “self-help” is the primary motivation; states must provide security for themselves because no other agency or actor can be counted on to do so.
In general, realist theories define “security” as the security of the state and place particular emphasis on the preservation of the state’s territorial integrity and the physical safety of its inhabitants (Walt 1991). A state is thought to be secure if it can defend against or deter a hostile attack and prevent other states from compelling it to adjust its behavior in significant ways or to sacrifice core political values. This conception may be contrasted with alternative definitions of “security” that focus on either the individual or the global level and do not privilege the state, or those that include nonviolent threats to human life (such as disease or environmental degradation), domestic crime, economic hardship, or threats to cultural autonomy or identity (Buzan 1983; Booth 2007).
Thus, a fairly narrow concept of “security” is central to the realist tradition. Indeed, one might argue that this narrow conception of “security” (i.e., protection against violent attack or coercion) has been inextricably linked to realist thought since its inception. In his famous history of the Peloponnesian War, for example, Thucydides traced its origins to the fear induced in Sparta by the growth of Athenian power ( 1996 :16). For Niccolo Machiavelli , writing in the Italian Renaissance, the Prince’s key object must be to preserve his position and the security of his realm in a world filled with wicked men who may threaten his position. As a result, rulers must be feared rather than loved and must be ready to act ruthlessly or treacherously if that is what “reason of state” demands (Haslam 2002:28–33). Working in the shadow of the English Civil War, Thomas Hobbes famously concluded that the natural condition of man was the “warre of every man against every man,” although this bleak condition might be remedied for individuals by a strong government – the Leviathan – that could establish among human society a “common power to feare” (Hobbes 1651/1968:187–8). Among states, however, there was still no overarching authority that could protect them from each other and prevent conflict and war. In his Discourse on Inequality ( 1754 ), Jean-Jacques Rousseau agreed that the absence of a central authority inhibited efforts to cooperate and so made the state necessary, in partial contrast to the Kantian view that “well-ordered republics” might overcome the incentives for rivalry inherent in anarchy and establish a “pacific union” (Doyle 1983).
Modern versions of realism proceed from a similar foundation. The central idea common to all modern versions of realism is that “the presence of multiple states in anarchy renders the security of each of them problematic and encourages them to compete with each other for power and/or security” (Walt 2003). For most realists, the imperative of obtaining security exerts far-reaching effects on states, encouraging them to act in certain predictable ways and eliminating those states who fail to compete effectively. If security were not a problem – either because humans or states ceased to care about it or because it was reliably guaranteed – realist theory would lose much of its analytic power and potential relevance.
Even scholars who do not advertise themselves as “realists” embrace key elements of this picture of the world. For example, the extensive literature on power transitions (Organski and Kugler 1980) implicitly assumes that states react to shifts in the balance of power largely from security concerns, and the so-called bargaining approach to international conflict models decisions for war as actions undertaken by states who are free to use force to secure their aims and are aware that their opponents are able to do the same (Fearon 1995; Powell 2002).
This essay explores the relationship between realism and security by considering three main topics. First, how does realism explain security and insecurity in world politics? In other words, why is security a problem, and what factors or conditions make this problem more or less intense? Second, what does realism tell us about the different ways that states can address this problem? In the “self-help” world that realism depicts, what are the different strategies that states can employ in order to try to make themselves more secure? Third, what security topics is realist theory currently addressing and what theoretical puzzles continue to attract attention?
As noted above, a central theme in virtually all realist writing is the idea that the existence of more than one state or “conflict group” (Gilpin 1986) in a condition of anarchy renders the security of each problematic and encourages them to compete with each other. Yet different realists offer different explanations for why security is scarce and focus our attention on different underlying mechanisms and causal factors. As recent works on the intellectual underpinnings of realism suggest, these different emphases reflect the historical conditions at the time different works were composed, the intellectual backgrounds and life experiences of particular authors, and the state of the broader intellectual discourse that accompanied these works (Schmidt 1998; Donnelly 2000; Haslam 2002).
For “biological realists” such as Machiavelli, Reinhold Niebuhr (1932), and especially Hans J. Morgenthau (1946; 1948), the ultimate taproot of insecurity is human nature, and in particular man’s innate desire for power. As Niebuhr put it, “the will to power of competing national groups is the cause of the international anarchy which the moral sense of mankind has thus far vainly striven to overcome.” Or more simply: “the ultimate sources of social conflicts and injustices are to be found in the ignorance and selfishness of men” ( 1932 :19, 23). For these writers, international anarchy is a permissive condition that allowed human aggressiveness – what Morgenthau termed the animus dominandi, or desire to dominate – to express itself. States were insecure because men craved power and sought to get more of it, and there was no central authority to prevent them from attempting to do so. Human nature is a constant and cannot be amended, which means that conflict is a central part of political life and cannot be eliminated. Indeed, given the primitive passions that he believed drove political behavior, Morgenthau himself remained ambivalent about whether a “rational” science of politics was even possible or desirable (Guilhot 2008).
By contrast, the English scholar E.H. Carr traced the security problem to the inevitable conflicts of interest that arise between states possessing different resource endowments, economic systems, and political orders. Carr’s chief work in the realist vein, The Twenty Years’ Crisis (1946), was a trenchant critique of the idealistic belief that international law, global opinion, or institutions like the League of Nations could effectively eliminate conflict and insecurity between states. In Carr’s words, “it is profitless to imagine a hypothetical world in which men no longer organize themselves in groups for purposes of conflict (1946:231). For Carr as for other “classical” realists, the absence of an effective central authority allowed power politics to continue and rendered all states potentially vulnerable to the predations of others, and especially revisionist powers. But Carr sees the motivating force behind security competition not in some innate human drive for power but in the governing systems, ruling ideologies, or personalities of individual leaders. Carr’s broader historical and sociological approach to world politics is now most closely reflected in the so-called English School (Bull 1977; Watson 1992; Linklater and Suganami 2006), which both draws on the realist tradition and challenges some of its gloomier predictions.
By contrast, “structural” realists emphasize a different “root cause” of the security problem, placing greater causal weight on the “ordering principle” of international anarchy. In this view, the absence of a central authority encourages states to compete even when they might not want to do so, a tendency observed by several writers well before the development of the modern “neorealist” version of this argument (Dickinson 1916; Schwarzenberg 1941). Thus, Herz (1950) noted that in the absence of central authority states face a “security dilemma” whereby actions undertaken by one state to increase its own security (such as building armaments or forming an alliance) tend to leave other state(s) less secure and prompt a counterreaction, which in turn leads to heightened suspicions and thus leaves both parties less secure than before. Moreover, Herz believed that the existing international order was even less stable than the idea of a security dilemma suggested, given the fragility of legal and social institutions and the ever-present possibility of evil (Stirk 2005).
This conception of structure as an active causal force was laid out with particular clarity in Kenneth Waltz’s landmark Theory of International Politics (1979). Where Morgenthau and some other early realists were ambivalent about the possibility of a “science of politics” (due to what they saw as the inherently unpredictable nature of human passions), Waltz sought to put realist theory on a more rigorous scientific basis. To do so, he drew on contemporary philosophy of science, microeconomics, and systems theory, with the aim of developing a purely “systemic” theory.
Waltz began by assuming only that all states sought to survive, while acknowledging the possibility that some states might also have more ambitious goals. Because these states were placed in an anarchic realm, however, they had to rely on their own resources and strategies in order to survive. For Waltz, the international system was not merely a passive arena in which men strove for power or states pursued independently derived “national interests.” Rather, he saw international structure – in which anarchy is a key ordering principle – as an active force that “shaped and shoved” the states whose existence constituted the system and who were in turn induced to behave in certain ways or suffer the consequences (Waltz 1979; 1986:343–4; Buzan et al. 1993).
Here it is worth emphasizing that Waltz relied primarily on the causal mechanism of competitive selection to explain why states tended to act in similar ways (i.e., to compete). Just as poorly managed firms in a competitive marketplace were likely to go bankrupt, states that did not heed the imperatives of the system were more likely to be conquered or to drop from the ranks of great powers. Yet Waltz also maintained that the “death rate” among states was quite low, which implied there was ample latitude for suboptimal behavior. Fazal (2007) amends this basic picture, suggesting that the likelihood of “state death” varies considerably across space and time. She finds that “buffer states” lying between great powers experience especially high death rates historically, that “state death” declined dramatically after 1945 , and that there is only a weak relationship between a state’s relative power and its prospects for survival.
Moreover, despite his emphasis on the autonomous role of system-level forces, Waltz’s “neorealist” theory still relied on unit-level factors to account for the security problem. As Schweller (1996) emphasized, if one accepts Waltz’s assumptions that the system is anarchic and states merely seek to survive, then there is in fact no logical necessity for them to worry about each other and no inherent reason for conflict to arise. And, as Mearsheimer (2009) has recently shown, Waltz did not in fact assume that states were rational but emphasized that great powers often behave in aggressive and reckless ways for various domestic political reasons. In order to explain why conflicts arise and states are insecure, in short, Waltz ended up saying one needed a separate theory of “foreign policy,” which is merely another way of saying that one must add unit-level factors to fully explain why states in anarchy are insecure.
The impact of systemic pressures becomes even clearer in Mearsheimer’s (2001) own elaboration of “offensive realism.” He explicitly assumes that states are rational actors that seek to survive, possess some capacity to hurt each other, and cannot know each other’s intentions with 100 percent confidence. Even if all states were convinced that no other state harbored any dangerous intentions at a given moment, they could not be sure that some other state might not become hostile or aggressive in the future. Similarly, even if a state is strong enough to defend itself now, it must continue to compete lest some other state catch up and then seek to use its power to extract concessions (or worse). Thus, Mearsheimer concludes that states wishing to survive must constantly look for opportunities to increase their power, so that they are in the best position to thwart an attack should one arise at some point in the future. Because all great powers also know that potential rivals are facing the same incentives, they are forced to compete for power even if they do not wish to, for fear of falling behind and becoming vulnerable to others.
Thus, offensive realism traces the security problem directly to the anarchic condition of world politics and the “unknowability” of intentions, though Lobell (2002–3) argues that domestic politics still plays a key role in shaping each state’s specific security strategy. It sees security competition and power-maximization as “hard-wired” into any system of world politics where central authority is lacking, intentions cannot be foreseen, and states have some significant capacity to harm one another.
Given the central place that security has in realist thought, it is not surprising that scholars working in this tradition have devoted considerable attention to identifying factors that can intensify or ameliorate the basic security problem.
Classical realists maintained that the danger of war was lowest and the security of states greatest in a multipolar system containing several (i.e., more than three) major powers. They argued that this permitted flexibility of alignment, precluding the development of bitter and enduring enmities and facilitating the formation of balancing coalitions which would help deter especially strong or aggressive states and thus make war more likely. Even if multipolar systems did not prevent war, they would enhance the security of the great powers and make it less likely that any of them would be eliminated from the system (Morgenthau 1948; Kaplan 1957; Deutsch and Singer 1964).
By contrast, Waltz (1964; 1979) argued that multipolarity created a greater risk of miscalculation and compounded the collective action problem, leading to inefficient balancing and thus to greater opportunities for aggression. He maintained that bipolarity was the most stable structure, because the two leading powers already controlled most of the assets in the system, were less prone to miscalculate the likelihood of opposition, and had a greater capacity to keep client states under control. Mearsheimer (2001) extends this line of argument, agreeing that bipolarity is the most stable configuration of power, while arguing that balanced multipolarity makes states more prone to war and that “unbalanced multipolarity” is the least stable structure of all. In particular, Mearsheimer claims that unbalanced multipolarity tempts the strongest powers into making bids for regional hegemony and is thus most likely to trigger hegemonic wars. A further refinement is Schweller (1998), whose analysis of alliance dynamics in World War II suggests that unbalanced tripolar systems are a particularly dangerous special case. Copeland (1996a) offers a dissenting analysis, suggesting that bipolar systems where one pole is declining will be prone to crises and war because the declining state has no alignment prospects that can reverse unfavorable power trends.
A second strand of theorizing identifies shifts in the overall balance of power as a key source of security competition and war, although there is as yet no consensus regarding the key causal mechanisms linking shifts in power to insecurity and war (Organski and Kugler 1980; Gilpin 1981; Levy 1987). Shifts in the balance of power are seen as dangerous because: (1) rising states become more ambitious and initiate conflicts in order to revise the existing status quo; (2) declining states fear the rising power and wage preventive wars to arrest its ascent; or (3) large shifts in the balance of power make it harder for potential rivals to gauge the balance of power and increase the risk of miscalculation. Copeland (2000) argues these problems are more severe in bipolarity than in multipolarity, because a declining state in bipolarity needs only to defeat a single rising power, but a declining state in multipolarity must believe it is strong enough to take on both the rising power and its potential great power allies.
Where Waltz and other “structural” realists focus on polarity (defined by the distribution of overall power resources), another influential strand of realist theory explains the intensity of security competition by focusing on the “fine-grained structure of power” (Van Evera 1999) and the effects of geography, diplomacy, and technology.
This approach – which is a key element in what is sometimes termed “defensive realism” – relies on the core concept of the offense–defense balance. This balance is generally defined as the relative ease or difficulty of conquest (Quester 1977; Jervis 1978; Glaser and Kaufmann 1998; Van Evera 1999). In the same way as other realist theories, defensive realism recognizes that anarchy forces all states to worry about security, but the intensity of this concern (i.e., the level of insecurity) will depend on whether conquest is easy or hard. When military technology, geography, the character of diplomacy, etc., combine to make conquest difficult, then security is plentiful and the danger of war declines. When these various factors combine to favor conquest, however, states will be less secure, cooperation will be elusive and wars will be more frequent and intense. It follows that states can enhance their security by adopting defensive military postures (especially when defense has the advantage) and status quo states can signal their benign intentions by eschewing offensive capabilities (Glaser 1994–5; Kydd 1997). Thus, defensive realists implicitly challenge the belief that security is scarce because states cannot gauge the intentions of others and must therefore assume the worst.
Although “offense–defense” theory was logically coherent and intuitively plausible, other prominent realists questioned its core claims. Critics argue that the core concept of the theory (the “offense–defense balance”) is impossible to measure and can change unpredictably, which means that states cannot and do not base important national security policy decisions on this factor (Levy 1984). Lieber (2005) has also challenged the empirical basis for the theory, arguing that national leaders rarely agree on what the “offense–defense balance” is – even after major technological revolutions – and do not seem to rely on assessments of the balance when making decisions for war and peace.
While emphasizing the incentives for competition induced by anarchy, some realists have focused on how different domestic orders, social structures, or individual leaders may respond to these pressures in radically different ways. Snyder (1992) and Van Evera (1984) argue that “cartelized” or “militarized” polities are prone to exaggerate security threats and adopt overly aggressive responses to them; more recently, “neoclassical” realists (Lobell et al. 2009) have sought to incorporate a variety of domestic or individual variables to explain specific foreign policy decisions, thereby sacrificing parsimony for the sake of descriptive accuracy.
Over the past two centuries, several alternative approaches to international relations have challenged the basic realist account of the security problem, and especially its conclusion that competition and insecurity are an inevitable condition for sovereign states coexisting in anarchy. In particular, these alternatives accept the idea that anarchy may encourage competition between states, but conclude that the picture of relentless security competition portrayed by realism is at best incomplete and at worst dangerously self-fulfilling.
First theorized by Immanuel Kant in his essay Perpetual Peace ( 1795 ), democratic peace theory acknowledges the potential for security competition in an anarchic order comprised of independent states, but argues that liberal or democratic states can nonetheless establish enduring relations where security competition is significantly attenuated (Doyle 1986; Russett 1994). The primary basis for this claim is the observation that democratic states do not seem to have fought each other, a claim that has been hotly contested (and defended) on a number of grounds (Elman 1997; Green et al. 2001).
Even if the empirical challenge were resolved, there is as yet no agreement on the precise causal mechanism that might account for the absence of war between liberal states. Scholars have suggested that democracies do not fight each other because: (1) democratic leaders fear electoral punishment; (2) there are powerful “norms of respect” between states sharing liberal values; or (3) because democratic states can make more credible commitments and signal intentions more credibly, thereby lowering the risk of war via miscalculation (Schultz 1999). Rosato (2003) challenges these explanations on theoretical grounds, while Mansfield and Snyder (2005) offer the important qualification that, while democratic states may be less inclined to fight one another, democratizing states are in fact more likely to be involved in war.
Liberal theories of economic interdependence have long posed a second challenge to realism’s depiction of the security problem (Angell 1913). As with democratic peace theory, economic liberalism accepts the primacy of national states and the absence of world government, but suggests that high levels of trade or investment can make it too costly for states to fight each other. In this view, rational self-interest (i.e., the desire for greater material prosperity) thus provides a powerful disincentive to war. High levels of economic trade or investment are also believed to encourage transnational ties and create bonds of familiarity between states, thereby reducing suspicions and minimizing xenophobic stereotyping. Critics point out that high levels of interdependence did not prevent World War I and also encouraged Japanese expansion in World War II, but the theory has been resurrected in more modern forms (Rosecrance 1986) and continues to attract scholarly attention. Copeland (1996b) links economic liberalism and offense–defense theory and suggests that interdependence reinforces peace when conquest is hard; while Brooks (2005) suggests that the integrated nature of global production processes has greatly increased the disincentives for conflict for industrial powers. Mansfield and Pollins (2003) summarize the state of the debate and find that the relationship between economic interdependence and war remains highly conditional.
Over the past two decades, social constructivists have mounted a more fundamental challenge to the realist explanation of the origins of international insecurity. Broadly speaking, constructivists argue that there is no necessary connection between anarchy and insecurity; in Alexander Wendt ’s famous phrase, “anarchy is what states make of it” (Wendt 1992; 1999). In this view, it is realist discourse that convinces leaders and populations to pursue competitive policies, and different results would occur if individuals spoke, wrote, and thought about these issues in a new way (Ashley 1984; Booth 2007). A contrasting view is offered by Fischer (1992), who argued that relations among the heterogeneous political units of feudal Europe revealed the same degree of insecurity and competition that realist theory predicts (including the formation of alliances, aggressive wars, etc.) despite the ideology of Christian universalism that infused the period, the fundamentally different nature of political identity, and the absence of a strong norm of sovereignty.
Constructivists also challenge traditional conceptions of “security” itself, suggesting that new conceptions and discursive practices could lead to a significant shift in state practice and yield more stable or peaceful outcomes (Krause and Williams 2003). Other constructivists emphasize the role that norms like the “nuclear taboo” or the norm against chemical weapons use can play in limiting or regulating interstate competition, thereby reducing levels of insecurity without eliminating it altogether (Price 1997; Tanenwald 2007).
In addition to these “regulatory norms,” some constructivists have argued that shifts in discourse and identity can transform existing conflicts and potentially eliminate the root causes of international rivalry, as illustrated by the Gorbachev revolution in Soviet foreign policy and the establishment of a “security community” within and between North America and Western Europe (Koslowski and Kratochwil 1994; Adler and Barnett 1998). Realists generally disagree with these interpretations, arguing that material decline played a more important role in ending the Cold War and emphasizing the potential for conflict that still remains (e.g., Wohlforth 1994–5).
Though not normally described as constructivist works, Mueller’s important discussions of the obsolescence of war (1986; 2004) also suggest that shifts in discourse and collective attitudes toward war have played a key role in inhibiting large-scale great power warfare since World War II. Mueller emphasizes that these changed attitudes have not eliminated all wars or rendered security competition obsolete, but he clearly believes the change is significant and likely to endure among the major industrial powers.
Less optimistically, a recent constructivist interpretation argues that states seek not just physical security from attack but also “ontological security” defined as the preservation of an existing identity and a set of recurring relations with others (Mitzen 2006). The latter view sees states not as trapped in security dilemmas that they would prefer to escape, but rather as attached to conflictive relationships that help preserve the state’s own identity. Interestingly, by providing a distinct causal mechanism for persistent conflict, this perspective actually reinforces realist views about the inevitability of security competition.
While useful, none of these broad critiques of the realist perspective on insecurity has delivered a fatal blow. This is itself not surprising, insofar as wars continue to occur and security competition still persists, even between democratic states with high levels of economic interdependence.
In addition to explaining why states worry about security, realism also identifies various strategies that states can pursue in order to make themselves more secure. Thus, realism offers both diagnosis and prescription, although the latter element is based on pragmatic considerations rather than on larger moral or ethical foundations. Although individual realist scholars obviously have their own moral convictions, realist theories themselves are essentially amoral. Realism may offer prescriptions for how states can best survive, but it is largely silent on whether the survival of any particular state or government is morally desirable.
In fact, there has been a lively debate among realists on the question of whether structural theories such as “neorealism” can offer specific guidance for foreign policy or not. Waltz insisted that his structural theory did not, and that to do so required a separate “theory of foreign policy.” Other realists challenged Waltz’s view explicitly (Elman 1996; Fearon 1998b) and it is clear that many prominent realists (including Waltz himself) have in fact used realist theory to derive specific recommendations for policy (e.g., Waltz 1981; Mearsheimer 1993; Walt 1987; 2005), a tendency that Oren (2009) has challenged on logical grounds. If structure is the main determinant of state behavior, he observes, then on what grounds can realists criticize what specific great powers are doing?
According to Waltz (1979), the tendency for states to balance power discourages attempts to maximize power and encourages states to seek only enough power to defend their own territory. Echoing this view, other “defensive realists” argue that the increasing difficulty of conquest and the nuclear revolution reinforce this tendency, and permit status quo states to adopt more cooperative strategies and to eschew efforts to maximize power (Van Evera 1999; Glaser 2000). In this view, states can maximize security by cooperating with others in mutually beneficial ways (e.g,. arms control agreements), and by adopting defensive military doctrines, which convey a “costly” (and therefore credible) signal of benign intent and permit “security-seeking” states to avoid needless rivalries. Montgomery (2006) offers an important qualification to this line of argument, pointing out that only when offense and defense are easily differentiated and the balance between them is neutral can states reveal peaceful motives without simultaneously jeopardizing their security.
As noted above, offensive realists (and others) reject this line of argument almost entirely, claiming that conquest is more profitable than defensive realists believe (Liberman 1996) and that it is largely impossible to distinguish between “offensive” and “defensive” weaponry. Accordingly, they argue that great powers should seek to maximize their overall power, because this is still the most reliable way to survive in an anarchic system (Mearsheimer 2001). Among other things, these critics point out that the nuclear revolution did not halt intense security competition either during or after the Cold War, and question whether “costly signaling” can ever be sufficiently credible as to convince states to neglect the balance of power (Lieber and Press 2006).
Realist theory has long seen alliances as one of the primary tools of statecraft (Morgenthau 1959), and recent realist research has devoted considerable attention to exploring the dynamics of alliance formation and their consequences for security. Building on Waltz’s claim that great power tended to balance against the strongest state or coalition rather than “bandwagon” with them, subsequent research found that states were in fact more likely to balance against threats, which were conceived as an amalgam of power, geographic proximity, specific offensive capabilities, and perceived intentions (Walt 1987). Weak states were believed to be somewhat more inclined to bandwagon than the great powers, especially when they were vulnerable and could not locate strong protectors, but bandwagoning was still regarded as rare.
The dominance of external balancing behavior was backed by several subsequent studies (Garnham 1991; Priess 1996) and challenged by others (Barnett and Levy 1991; Labs 1992). Defensive realists (e.g., Van Evera 1999) argued that rapid and efficient balancing behavior made conquest more difficult and thus discouraged aggressive behavior, while other realists argued that domestic impediments and dilemmas of collective action encouraged free-riding behavior and “under-balancing,” thus creating openings for aggression, especially in evenly balanced multipolar orders (Schweller 2004).
Employing a broader definition of bandwagoning that included opportunistic alignment for purposes of expansion, Schweller (1994) argued that bandwagoning behavior was more common than earlier realists had suggested. Schroeder (1994) also questioned the propensity for states to balance and suggested that diplomatic history revealed that states often preferred to do almost anything rather than balance a powerful rival. David (1992) argued that developing countries in the postwar era were more sensitive to internal threats than external challenges, and that their alignment decisions were based on whether potential partners could help them thwart internal rivals and retain power.
The most ambitious application of realist theory to the dynamics of alliance politics is Snyder (1997), which develops a more fine-grained explanation of alliance behavior and its broader consequences for international politics by adding a host of other factors to Waltz’s spare distinction between bipolar and multipolar worlds. Snyder’s analysis also showed that alliance ties could be an ambiguous source of security, especially in multipolar systems, because alliance partners had to worry about being either left in the lurch by an ally (abandonment) or dragged into an unwanted war (entrapment). This “alliance security dilemma” resulted from the fact that hedging against one danger made states more vulnerable to the other (Snyder 1984). Christensen and Snyder (1990) offer a further refinement, arguing that multipolar alliances exhibit different pathologies, depending on the state of the offense–defense balance. When offense is easy, alliance ties will be tight and it will be hard to restrain one’s partners (as in 1914 ), but when defense is believed to be dominant, alliance partners will try to pass the buck and therefore fail to balance efficiently.
In response to these various competing arguments, Vasquez (1997) suggested that realism was a “degenerating research program” that should be discarded. Realists countered by noting that the main evidence for Vasquez’s recommendation was quite sparse, and consisted primarily of his observation that a few realists had disagreed with each other about the propensity of states to balance. They also noted that, if disagreements of this sort were sufficient to call realism into question, this criterion would also call into question most research programs in the social sciences (Walt 1997; Vasquez and Elman 2003). More usefully, Levy and Thompson (2005:4) surveyed five centuries of European great power diplomacy and found that balancing behavior occurred primarily in response to hegemonic challenges but not to lesser threats. They conclude that “balancing was a probabilistic tendency rather than an ‘iron law’ of behavior.” The various essays in Paul et al. (2004) find much evidence of contemporary balancing behavior, but an ambitious multidisciplinary survey of different historical systems by Wohlforth et al. (2007) found that balancing behavior was neither a universal tendency nor a reliable barrier to regional hegemony, and was instead dependent on particular historical circumstances.
Mearsheimer (2001) also questioned whether balancing behavior was the preferred response to external threats, and suggested that buck-passing (i.e., getting others to bear the costs of countering a threat) was the more common strategy. In this view, forming a balancing alliance was the second-best alternative, and bandwagoning with a powerful challenger was by far the least appealing.
More recently, realist discussions of alliance formation have explored the apparent failure of major powers to balance against the US in the aftermath of the Cold War. Structural realists predicted that balancing would soon take place (e.g., Waltz 1993), while other realists suggested that anti-American balancing would be limited and that the more likely response would be some form of “soft balancing” (Walt 2005; 2009; Paul 2006; Pape 2006). By contrast, Wohlforth (1999; 2009) and Brooks and Wohlforth (2008) suggest that absence of overt balancing is itself a structural consequence of unipolarity; by definition, a unipole (in this case the US) is too powerful to be countered by anything less than a coalition of all other major powers, and such an alliance would inevitably face nearly insurmountable dilemmas of collective action. Lieber and Alexander (2005) reach the same conclusion by a different path, arguing that medium powers are disinclined to balance because they do not fear US ambitions and agree with most (though not all) of US foreign policy.
In addition to external alignment, states can also address the security problem by internal effort, such as an arms buildup (Waltz 1979). As noted above, this is directly connected to the core concept of the security dilemma, which explains why unilateral efforts to improve one’s security are often ineffective or even counterproductive. Glaser (1992; 2000) argues that the net effect of arms racing depends on the nature of the weapons being acquired, the motivations of the respective states, and their ability to perceive the opponent’s actions in an unbiased way.
Although some early realists questioned whether nuclear weapons could be a reliable source of security (Kissinger 1957), over time many realists came to see them as an important exception to the logic of the security dilemma. The logic was straightforward: if nuclear weapons are used to deter attack by threatening unacceptable punishment, then it is possible to defend oneself from conquest without simultaneously acquiring the capacity to conquer others. Thus, Jervis (1978; 1989) argued that second-strike nuclear forces eliminated the security dilemma between states, because once each side has clear second-strike capabilities, adding more weapons to either side is strategically meaningless. Waltz (1981) drew on similar logic to argue that the slow spread of nuclear weapons would dampen international competition, though this view was soon challenged and remains controversial (Sagan and Waltz 1995).
Realism portrays international politics as a competitive realm of action, in which states are worried about the potential danger posed by other states and must therefore look for ways to protect themselves. Drawing analogies from sociology and microeconomics, Waltz (1979) argues that states are “socialized to the system” by these competitive pressures. Although states are free to act however they wish, those who behave foolishly or who fail to appreciate the need to compete are likely to be eliminated. As a result, the remaining states will tend to resemble each other over time, as outliers are “selected out.” Waltz also argues that states will consciously imitate each other, to prevent a particular state from gaining or exploiting a permanent advantage, a tendency confirmed by several subsequent studies of the spread of military innovation (Posen 1993b; Resende-Santos 1996; Goldman and Andres 1999).
Furthermore, as Mearsheimer (2001) points out, these same competitive pressures also encourage states to innovate. States not only imitate successful innovations made by others (encouraging greater similarities between them), but they will consciously look for new and improved ways to compete, whether in military affairs or in other realms (thereby encouraging greater diversity). This tendency helps explain why balances of power rarely remain fixed over long periods of time, and why states can never be entirely sure that a seemingly favorable security position will not erode in the face of another state’s innovative breakthrough.
Realism sees institutions as “tools of statecraft” that states can use to advance specific security interests. There is therefore no significant difference between realists and neoliberal institutionalists; each group recognizes that institutions can help states cooperate in specific circumstances (i.e., when there are genuine incentives to cooperate as well as incentives to defect. Grieco (1990) argued that concerns for relative gains made cooperation more difficult than institutionalists originally believed, a point that some leading institutionalists eventually conceded after a protracted scholarly debate (Keohane 1993:283).
Realists have long maintained that formal or informal institutions are strong enough to eliminate all conflicts of interest between states or to prevent great powers from pursuing those interests (Carr 1946; Mearsheimer 1994–5). For realists, however, institutions are largely epiphenomenal: they reflect the underlying balance of power and the interests of the most powerful states (Gruber 1999) and thus do not provide a powerful solution to the core security problem.
Nonetheless, realism recognizes that effective diplomatic institutions can make important contributions to security. An efficient system of diplomatic communication is a prerequisite for effective balancing behavior, which in theory makes it easy for threatened states to attract allied support and thus makes opportunistic aggression more difficult. Although realists are skeptical of the claim that “concert systems” or other types of “collective security” systems eliminate the competitive impulses inherent in anarchy (Jervis 1989; Kagan 1997–8; Rendell 2000), they nonetheless recognize that such arrangements can facilitate diplomatic coordination, encourage some degree of mutual restraint, and allow states to deal with shared security problems such as international terrorism or climate change (Jervis 1985; Van Evera 2008).
Not surprisingly, the end of the Cold War led a number of scholars to anticipate an end to security competition, which they believed would render realist theory obsolete or at least less useful (Kegley 1993; Jervis 2002). Collard-Wexler (2006) argues that realism could not account for the pacification of Western Europe under the aegis of the European Union (an event he correctly judged to be “one of the most significant developments in international relations”) but Rosato (2006) offers a realist account of this process that addressed many of Collard-Wexler’s criticisms. More recently, other scholars have suggested that the emergence of nonstate threats from international terrorism requires a thorough rethinking of the realist approach. To a considerable extent, scholars working in the realist tradition have attempted to rise to this challenge. Although the post–Cold War period has seen a marked decline in interstate violence and a growing concern about terrorism and civil or ethnic wars, realism continues to make important contributions to the analysis of contemporary security problems.
As ethnic conflict and civil war began to dominate the post–Cold War security agenda, Posen (1993a) showed that key elements of realist theory – in particular, the absence of a central authority, the vulnerabilities of particular groups, and the implications of rapid shifts in the balance of power – could explain why some multiethnic societies might be especially prone to conflict in the event of a central government collapse. Posen’s insights were echoed by several subsequent studies (Lake and Rothchild 1996; Fearon 1998a; Rose 2000), expanded by others (Kaufmann 1996; 1998), and qualified by several empirical works (Fearon and Laitin 2003; Toft 2003). This basic approach is also consistent with arguments tracing civil war settlements to credible third party guarantees, which overcome the commitment problems inherent in anarchy (Walter 2002).
Recent scholarship on the origins of mass violence highlights the central role that security considerations play in these tragic events. In particular, Valentino (2005) convincingly shows that mass killings reflect neither ancient hatreds nor purely ideological programs, but rather the strategic logic of leaders determined to preserve their positions by exterminating groups that they believe pose a long-term threat to either their personal positions or the security of the state itself. In addition to demonstrating the “rationality” of such seemingly irrational and horrific acts, Valentino also underscores the vulnerability that stateless peoples face when confronted by malevolent state power.
The end of the Cold War triggered a long debate about its likely implications for great power security competition. Mearsheimer (1990) argued that the lack of a great power rival would encourage US retrenchment and lead renewed security competition in Europe, while Friedberg (1993–4), Roy (1994), and Ross (2006) drew on realist ideas to anticipate renewed great power competition in Asia. Waltz (1993) and Layne (1993; 2006) predicted that a combination of overcommitment and external balancing would soon undermine US primacy, while other realists (e.g., Walt 1997; 2005) suggested that efforts to balance the US would be modest and would not threaten US primacy.
Over time, growing recognition of the scope of US dominance encouraged important theoretical discussions of the novel condition of unipolarity. Wohlforth (1999; 2009) and Brooks and Wohlforth (2008) suggest that unipolarity is even more stable than bipolarity, because the unipole’s superior position will deter both hegemonic challenges and clashes between other major powers. Pape (2009) and Layne (2006) question the durability of the “unipolar” moment and suggest that a combination of overcommitment and external balancing will drive the system back to multipolarity and encourage renewed security competition. Lieber and Press (2006) argued that US nuclear weapons policy reflected a continued quest for nuclear superiority, a policy based on the assumption of continued security competition in anarchy.
The terrorist attacks on September 11, 2001 , lay outside the main tenets of realist thought and some scholars suggested that the new focus on terrorism required a fundamental rethinking of the realist perspective (Brenner 2006). The main threat to state security now seemed to arise not from other states but from nonstate actors such as al-Qaeda, whose political programs reflected not realpolitik but an amalgam of fundamentalist religion and opposition to perceived foreign interference and the supposedly corrupt and decadent regimes that tolerated it.
Over time, however, the relevance of realist ideas for contemporary security problems became clearer. Societies facing terrorist threats did not respond by calling on international organizations like the UN or on nongovernmental organizations (NGOs) such as Amnesty International. Instead, they looked to national governments to devise new strategies for dealing with this new threat. Accordingly, prominent realists called for significant adjustments in US foreign policy, both to address the specific dangers posed by al-Qaeda and its affiliates and to eliminate some of the grievances that have given rise to such movements (Walt 2002). Drawing on realism’s rationalist roots, Pape (2005) sought to explain suicide terrorism as a strategic response to perceived foreign occupation, and similarly prescribed reducing the foreign “footprint” in the Arab and Islamic world so as to retard terrorist recruitment. Not surprisingly, realists were in the vanguard of opposition to the neoconservative campaign to spread democracy and “transform” the Middle East at the point of a gun, correctly arguing that this policy would jeopardize foreign support for the “war on terror” and undermine America’s overall global position (Williams 2005a; Schmidt and Williams 2006; Lieven and Hulsman 2006).
It is therefore not surprising that the intellectual connection between realism and security remains strong. Or, as one scholar put it, “it is difficult to avoid a sense that in the 21 st century realism is resurgent” (Williams 2005b:2). Even though the overall level of global violence – and especially interstate violence – has declined dramatically since the end of the Cold War (Gleditsch 2008), states do not appear to take security for granted. Ironically, the levels of violence may even be lower because states are taking security seriously, but in more intelligent and farsighted ways than they did in the past. If so, the contributions of realist theory may deserve at least some of the credit. Scholars continue to debate its historical roots, conceptual foundations, and predictive accuracy, but realist thought continues to provide a powerful way to think about the security problems that all states face and the strategies they employ in the ceaseless quest to overcome them.