Introduction

Neoclassical realism is a relatively recent attempt to marry structural realism's (or neorealism's) emphasis on scientific rigor and the causal primacy of the international system with the attention to domestic level institutions, problems of perception, and concerns of leadership that concerned classical realists. Taking neorealism as their point of departure, its adherents posit that states respond primarily to the constraints and opportunities of the international system when they conduct their foreign and security policies, but that their responses are conditioned by unit-level factors, such as state–society relations, the nature of their domestic political regimes, strategic culture, and leader perceptions. This essay will discuss the nature of the neoclassical realist challenge to neorealism, the theory of the state that neoclassical realists either explicitly or implicitly subscribe to, the varieties of neoclassical realism, and directions for future research.

The Limits of Neorealism

The father of structural realism, Kenneth Waltz (1959), argued that to understand international politics, we must understand the nature of the international system – the political environment within which states interact. Explanations of patterns of international behavior, such as the recurrence of war, that were based on conceptions of human nature or the nature of states (his first and second images of international politics) were, at best, incomplete, as human nature theories could not explain why war does not always occur, while theories of national difference could not explain why states with different political systems behave similarly in similar circumstances. Furthermore, first or second image theories were insufficient because the prescriptions that followed from them required changes in the relations between states, which implies that the problems themselves stemmed from the nature of the international system, his third image. In contrast, third image theories, focusing on anarchy in the international system and its consequences for states, were the most efficient explanations of war and other macropolitical outcomes between states, and they were sufficient as they required no reference to the types of states involved, human nature, or the particular leaders of particular states. As a result, Waltz (1979) constructed a third image theory of international politics that assumes that under anarchy, states respond to the most important variable in the international system, the distribution of capabilities.

For Waltz (1979) and other structural realists (e.g., Organski and Kugler 1980; Gilpin 1981), differential growth rates, which over time change the relative distribution of capabilities between states, are the driving forces of international politics. Rising states pose a challenge to others and inspire them, almost automatically, to balance against the challenger either internally, by arming, or externally, by allying with other states. In addition, to ensure their long-term survival, states are compelled to anticipate future power shifts and forestall them through policies such as preventive war (e.g., Levy 1987; Copeland 2000). Furthermore, they should seek relative rather than absolute gains vis-à-vis other states and avoid the pursuit of cooperative agreements that provide their rivals with gains that can be converted to military advantage (Grieco 1988). In an anarchic, self-help system, where security and survival are always at stake, states are thus compelled to obey systemic imperatives and do so regularly. While Waltz (1979:118–28) acknowledges that states do not always behave as the international system requires them to, he maintains that, because those that defy systemic imperatives are frequently defeated and eliminated, the international system socializes states over time to balance against rising great powers and to emulate the successful security behavior of their peers (cf. Resende-Santos 2007). Consequently, regardless of their leadership and domestic political differences, neorealists expect states to balance against rising challengers in a predictable and unproblematic manner (e.g., Waltz 1967: 306–11, which argues that the domestic differences between Great Britain and the United States had little impact on their foreign policy behavior).

In this regard, although structural realism is designed to explain international outcomes rather than foreign policy, one can derive the externally driven model of state behavior outlined in Figure 1 from its premises (e.g., Elman 1996). For structural realists, states are compelled to select foreign policies that are most appropriate responses to systemic circumstances. Domestic politics and leader characteristics play no significant role in determining policy, given the great dangers of interfering with systemic imperatives in an anarchic realm.

Open in new tab

Figure 1 A neorealist model of foreign policy

Neoclassical realists agree with structural realists that states construct their foreign security policies primarily with an eye to the threats and opportunities that arise in the international system. Since their very survival is at stake if they fail to secure themselves properly from without in an anarchic international system, where the slightest misstep could lead to defeat in war, the incentives are extremely high for states to focus on external stimuli and craft foreign policies to respond to them appropriately. As Jennifer Sterling-Folker (1997) describes, therefore, neoclassical realists share an environment-based ontology, granting primacy to the political environment within which states interact. Nonetheless, they reject the implication that states necessarily respond as fluidly to changing international circumstances as structural realist balance-of-power theories imply. In particular, they note four important limitations to the neorealist model:

1. States do not always perceive systemic stimuli correctly. The international system may present states with relatively clear requirements, based on the relative distribution of capabilities and differential growth rates. Yet, as William Wohlforth (1993:2) points out, “If power influences the course of international politics, it must do so largely through the perceptions of the people who makes decisions on behalf of the states.” As Robert Jervis (1976) and others (Blainey 1973:35–56; Lebow 1981:101–19; Stoessinger 2005) have noted, leaders, who are only human after all, frequently err in their calculations of relative power, their identification of the options at their disposal, and their assessments of the likely consequences of their actions. Such misperceptions can occur to any leader, particularly when faced with incomplete information about other states intentions, relative capabilities, and the likely consequences of one's behavior (Jervis 1988). But it can also result from a systematic bias in a particular leader's package of images and cognitions that comprise his or her cognitive filter, which is used to evaluate and process incoming information (Jervis 1976:28–31). Therefore, a state's national security behavior may have more to do with its leader's personality and behavior than objective systemic constraints and opportunities (Jervis 1976:18–19).

In this vein, James McAllister (2002) argues that American leaders in the immediate post–World War II period did not behave as Waltz would have expected in a bipolar era – that is, they pursued external balancing by rearming the Western Europeans rather than internal balancing and they sought to unite Western Europe, rather than to foster its dependence on American power – because they perceived not bipolarity but “a latent tripolar system” if German recovery were to occur. Wohlforth (1993) maintains that US and Soviet perceptions of the relative balance of power shaped the nature of superpower relations during the Cold War. While he maintains that, over time, these perceptions tended to follow the actual distribution of power, they led to crises when short-term perceptions masked longer-term trends. And Victor Cha (2000) suggests that perceptions of the United States’ level of commitment, rather than merely objective calculations of the balance of power, were the primary determinants of the ups and downs in relations between Japan and the Republic of Korea. If leader perceptions of systemic constraints diverge from reality and differ from leader to leader, systemic theories of foreign policy and international politics would be, at best, incomplete, as the sources of state behavior may lie less in the external environment than in its leader's background or psychological make-up.

2. Leaders do not always respond rationally to systemic stimuli. Even if they correctly perceive the threats and incentives of the international system, they may follow suboptimal or irrational decision-making processes that could lead to policy responses at odds with systemic requirements. As with perception, these problems could relate to cognitive limits on the ability of human beings to process information, particularly in a crisis, when time is short and the stakes are high (Holsti 1979). As a result, leaders may fail to identify all the policy alternatives available to them or may choose between them in a suboptimal manner, rather than selecting the option likely to maximize the expected payoff at the lowest possible cost. They might even become paralyzed by indecision and fail to react in a decisive manner, as Barbara Tuchman (1994:56–70) claims Russian Czar Nicholas II was on the eve of World War I, when he vacillated between competing mobilization plans in response to Austrian action against Serbia (cf. Lebow 1981:115–19). Nonetheless, particular decision makers might be especially susceptible to failures of rationality, due to their unique temperament, cognitive flaws, eccentricities, or historical experiences (e.g., Jervis 1976:217–71; Byman and Pollack 2001; Hermann et al. 2001). Thus both German Chancellor Adolf Hitler and Soviet leader Josef Stalin had megalomaniacal tendencies that led them to dominate foreign policy decision-making, overrule political and military experts, and deny opinions and information at odds with their views. This led them both to undertake irrational decisions, such as Hitler's decision to declare war against the United States after Pearl Harbor “without consultation with his military strategists […] without anything approaching proper preparation for such a conflict, and, as Donitz recalled, without taking cognizance of immediate logistical considerations” (Kershaw 2007:385) and Stalin's unwillingness to prepare for an impending German attack in June 1941 (Murphy 2005). More surprisingly, Canadian Prime Minister William Lyon Mackenzie King reached his foreign policy decisions, including his decision to enter World War II, after consulting the spirits of his dead ancestors with the help of a medium – hardly a rational decision-making procedure (e.g., Stacey 1976: 182–92)! Irrationality of this sort is, of course, problematic for purely structural theories, which require states to respond to international imperatives in a rather automatic fashion, selecting the most appropriate policy response to meet external conditions.

3. The international system does not always present clear signals about threats and opportunities. In extreme circumstances, when states are faced with a clear and present danger – such as a rapid and imminent power transition – they can easily discern the threat and determine how to counter it given its time frame and the resources at their disposal. Thus, in 1967, when Egypt blockaded the Straits of Tiran, mobilized its armed forces on the Israeli border, and asked the UN peacekeeping forces to evacuate the Sinai, it was clear to Israeli leaders that they were in imminent danger and that a pre-emptive strike would be appropriate (Oren 2002). Most situations are not as clear-cut, however, leaving great ambiguity over the nature of both the challenges the international system presents and the appropriate responses to them. Thus, for example, it was unclear whether the rise of American power and its emerging dominance over the Caribbean Sea in the late nineteenth century constituted a threat to British naval supremacy that should be resisted or an opportunity for the British to retrench and concentrate its naval power in regions of greater strategic importance, especially as American economic policies in the Western hemisphere would further British economic goals (e.g., Gilpin 1981:195–6; Lobell 2003:53–63). It is also unclear whether the rise of China in the post–Cold War era requires the United States to respond in a competitive manner, requiring containment, or whether it necessitates an engagement strategy to moderate Chinese risk-taking behavior (see Friedberg 2005; Ross and Feng 2008). If the international system only rarely provides clear enough information to states to guide their policy responses, then a broad range of foreign policy choices and international political outcomes must lie outside the purview of a structural theory of international politics.

4. Because of domestic political/economic circumstances, states cannot always mobilize the domestic resources necessary to respond as the international system requires. The neorealist model of foreign security policymaking requires a perfectly flexible state that is able to identify systemic imperatives correctly and respond promptly as the international circumstances require. If balancing is required, the state must be able to raise revenues, mobilize resources, and enlist military manpower in timely fashion to prevent a revisionist state from attaining hegemony. The state must be prepared to wage preventive war when faced with a certain power transition (Levy 1987; Copeland 2000; Lemke 2003).

This level of flexibility assumes that states face no domestic constraints when making national security decisions. In practical terms, however, not all states have the ability to direct policy on their own when faced with opposition from powerful domestic interest groups and societal veto players in the legislature and elsewhere (Tsebelis 2002). Despite his preference to provide greater support for Great Britain and France against Nazi Germany, for example, American President Franklin D. Roosevelt was impeded by public and Congressional opposition, which initially slowed down and limited American assistance. Until the Japanese attack on Pearl Harbor in December 1941, he was also only able to offer incremental resistance to Japanese policy in the Pacific, despite his preferences, because of domestic constraints (Dallek 1979:199–313; Casey 2001:30–45). Similarly, three successive French premiers who believed that German rearmament within a European Defense Community (EDC) was a strategic necessity for France nonetheless stalled the measure and prevented a ratification vote on the EDC treaty in the National Assembly because of the magnitude of legislative opposition (Ripsman 2001). Nor do all states have automatic access to the human, financial, and material resources they need to implement their preferred foreign security policies. Instead, less autonomous states must frequently bargain with legislators, power brokers, and societal groups over both the policies chosen and the amount of resources to be devoted to that purpose. Thus, Barnett (1992) demonstrates that the Egyptian and Israeli states had to bargain away state power in order to mobilize resources for war in 1967 and 1973, which, in turn, made them more dependent on powerful domestic actors for subsequent mobilizations. Consequently, a theory of international politics and the nature of state responses to the international system must be able to differentiate between states on the basis of their unique policy-making environments.

Two Varieties of Neoclassical Realism

Because of these important limitations, neoclassical realists have supplemented structural realist theory with unit level variables to explain two key categories of events with which a purely structural theory could not cope. In the first, states occasionally respond inconsistently with systemic imperatives. Even when systemic imperatives are quite clear, such as when the growth of German power in the 1930s threatened an imminent power transition in Europe, there have been notable incidents of what Randall Schweller (2004, 2006) calls “underbalancing.” Significantly, Schweller cites the French failure to prepare itself for a war against rising imperial Germany prior to World War I and British and French appeasement in the 1930s as quintessential cases of a failure to balance appropriately. Similarly, as Jack Snyder (1991) contends, while the international system rarely rewards expansionism and usually punishes it, and while most states avoid expansionism for that reason, history offers us many high profile cases of states that ignore these constraints at their own peril. Thus, Imperial Germany, Nazi Germany, and Imperial Japan all sought imperial expansion, which proved self-destructive in the long run. As Jeffrey Taliaferro (2004) observes, states often engage in costly interventions in regions peripheral to their core interests and, moreover, frequently persist in failing interventions. Classic cases include United States interventions in Korea and Vietnam. Finally, although Geoffrey Blainey (1973:115–24) observes that it would be logical for weaker powers to avoid disastrous wars through compromise, which is also consistent with the bargaining model of war (Fearon 1995; Reiter 2003), states occasionally initiate wars against stronger adversaries, or fail to make acceptable compromises to stronger adversaries in order to avoid war. Thus, for example, Japan initiated hostilities against the United States in 1941 and Iraq defied American demands in 1990, which led to a war with a far superior power.

Schweller, Snyder, Taliaferro, and other neoclassical realists explain these surprising discrepancies from structural realist expectations in terms of domestic politics. Schweller (2006) argues that underbalancing occurs as a result of four domestic political variables: elite consensus and cohesion, which affect the state's willingness to balance; and government/regime vulnerability and social cohesion, which explain the state's ability to extract resources from society to implement a balancing strategy. In other words, when the state is fragmented or weak vis-à-vis society, it is unable to respond to external threats as the system requires. In similar terms, Sten Rynning (2002) argues that the long lag time between French power decline and foreign policy adaptation can be explained by domestic political resistance. Ripsman (2001) explains the long delay in equipping West German forces to participate in Western defense efforts against the Soviet Union in the late 1940s and early 1950s in terms of the lack of state autonomy in Fourth Republic France, which prevented successive French Prime Ministers who favored German rearmament from agreeing to it in the face of domestic opposition, despite their sense of the urgency of the Soviet threat.

Taliaferro argues that to explain great power intervention in peripheral regions we must marry a defensive structural realist theory with an understanding of how leaders process information, particularly their aversion to losses and their willingness to take inordinate risks to avoid losses (Taliaferro 2004). In comparable terms, Kevin Narizny (2003) explains underbalancing in terms of the economic and political interests of societal actors, while Mark Haas (2005) does so in terms of the ideology of the dominant coalition. Snyder (1991) explains pathological overextension as a consequence of the nature of domestic political regimes. Unlike most states, those led by imperialistic cartel regimes and militaristic general staffs are more likely to pursue irrational overexpansion.

In Gideon Rose's terms, therefore, this group of neoclassical realists maintains that the international system sends clear signals to states, but that these signals must inform national policy responses only after passing through the often imperfect transmissions belts of leader perception and domestic politics (Rose 1998). In rare cases, either the signals are misunderstood or national leaders are prevented from responding properly by domestic political constraints. This face of neoclassical realism, therefore, is a theory of suboptimality or pathology to explain what are only understood to be infrequent deviations from neorealist expectations (Zakaria 1992:190–1). In other words, this group of neoclassical realists contends that, most of the time, balancing is rather fluid and automatic; only in unusual circumstances do flawed perceptions or domestic political realities interfere with rational national security responses.

In the second, when the international environment does not present a clear and imminent threat, there is often a range of policy options that states can choose from, rather than a clearly optimal policy dictated by international circumstances (Ripsman et al. 2009). The actual choices states make under these circumstances may have far more to do with the worldviews of leaders, the strategic cultures of the states they lead, the nature of the domestic coalitions they represent, and domestic political constraints on their ability to enact and implement various policy alternatives. When French, Russian, German, American, and Japanese contenders rose in power to present potential challengers to British hegemony in the late nineteenth century, for example, it was not clear a priori how British grand strategy should respond. Instead, as Steven Lobell (2003:43–85) argues, domestic political coalitions competed to determine what degree of threat each challenger posed and whether that threat should be met with cooperative or competitive policy responses (cf. Friedberg 1988:135–208; Brawley 1993:115–37). Similarly, William Wohlforth (2009:32) maintains that nothing in the pre–World War I international system required Germany to wage war against Great Britain, its leading trading partner and the great power that threatened it least. Instead, he argues, the roots of war had their foundations in the preferences of German leaders and their concerns for relative status, which interacted with a multipolar balance of power to favor war.

Sometimes, the range of choice is very limited, with the broad outlines determined primarily by external considerations, leaving domestic political considerations to affect only the style of the policy response or its timing. In this manner, Dueck contends that US military interventions in Korea and Vietnam were necessitated by Cold War exigencies, but their timing and style were affected by concerns about domestic political opposition (Dueck 2009). In other circumstances, however, the range of choice is quite wide, giving the state and key societal actors greater scope to bargain over policy; consequently, policy is more likely to be tailored to suit domestic political circumstances. Thus, for example, Mark Brawley argues that British, French, and Soviet foreign policies in the 1920s were all predicated on the threat of a resurgent Germany, but given the remoteness of the threat in that decade, they were able to adopt considerably different policy responses that reflected the uniqueness of both their strategic situations and their domestic political constraints (Brawley 2009a:81–9; Brawley 2009b:chap. 6).

A broad range of the emerging neoclassical realist literature is, therefore, more about foreign policy choices than pathologies. Jason W. Davidson (2006), for example, argues that the fundamental orientation of a state as a defender of the international status quo or as a revisionist challenger is determined not only by its relative power and position within the system, but also by the degree of influence that nationalists and the military have within the domestic political coalition. Colin Dueck (2008) interprets American foreign policy since World War I in terms not only of the United States’ strategic position, but also of the cultural values of the country's liberal internationalist elites. Alastair Ian Johnston (1995) argues that the nature of the Ming Dynasty Chinese response to the Mongol threat was influenced by a centuries old strategic culture. In other words, this face of neoclassical realism is a theory of foreign policy in its own right, rather than merely a corrective to explain anomalies for structural realism.

Both varieties of neoclassical realism depart from structural realism in that they identify a broad range of unit and sub-unit variables that can intervene between systemic stimuli and foreign policy responses. While they agree with neorealists that policy ought to fit the international strategic environment, neoclassical realists observe that states cannot always tailor their policies to international circumstances due to faulty perceptions of systemic stimuli, decision-making procedures that fall short of rational, or obstacles to policy implementation caused by a failure to mobilize societal resources. As illustrated in Figure 2, these flaws are often influenced by: leader images that interfere with accurate perceptions; strategic culture, which shapes state responses; state–society relations, which affect the state's ability to enact and implement decisions; and domestic political institutions, which can either enable or constrain state leaders when they face societal opposition. As a result, this more complex domestic decision-making environment implies that states do not necessarily select the optimal policy response to satisfy systemic constraints; instead, they choose from a range of policy alternatives to navigate between systemic constraints and domestic political imperatives.

Open in new tab

Figure 2 The neoclassical realist model of foreign policy

In this regard, neoclassical realism constitutes a unique position on the agent-structure debate (e.g., Dessler 1989; Sterling-Folker, 1997). While structure plays a dominant role in determining both agential choice and political outcomes, structure itself must be interpreted by agents and responded to within domestic political institutions that may grant decision-making scope to agents. Consequently, under certain circumstances, agents may play an important role in determining the political effects of international structures.

Distinguishing Neoclassical Realism from Other Approaches

Neoclassical realism, therefore, represents an effort to return structural realism to its classical roots, without losing the important innovations and scientific rigor that Waltz introduced to realism. It revives classical realism's concern for domestic politics and institutions, and its emphasis on the quality of diplomacy as a means of explaining the foreign security choices of states (e.g., Carr 1952; Niebuhr 1959; Aron 1966; Morgenthau and Thompson 1985). It updates classical realism, however, by providing clearly-stated, testable hypotheses, aspiring to the positivistic scientific rigor that structural realism introduced to realism and specifying the causal primacy of the anarchic international system (cf. James 2002:14–20). It also avoids classical realism's emphasis on human nature, passions, and the quest for unlimited power, which is not necessary to explain a broad range of international phenomena and foreign policy choices, nor is it conducive to a rigorous scientific theory of international relations and foreign policy. Moreover, while it acknowledges that perceptions and the ideas leaders bring with them to the table can impact on their responses to international stimuli, it does not buy into the fluidity with which ideas interpret and inform interests that some (e.g., Williams 2004) have claimed classical realists, such as Morgenthau, posited.

While its emphasis on unit and sub-unit variables pertaining to domestic political institutions, decision-making characteristics, and perception have been informed by what Rose labels “Innenpolitik” approaches, neoclassical realism is clearly distinct from the Innenpolitik logic, and it cannot be dismissed as reductionist (Rose 1998). Innenpolitik theories are unit-level approaches, which explain foreign policy primarily in terms of the internal characteristics of states, their domestic political processes and the individuals, parties, and coalitions that lead them. Liberal Innenpolitik approaches assume that domestic coalitions and the public at large determine the content of foreign policy independently of external events, and constrain liberal states from selecting policies outside the domestic consensus. Thus, for example, democratic peace theory assumes that public opinion and the public's representatives in the legislature restrain leaders from using force against other democratic states, regardless of the underlying balance of power and the nature of conflictive issues between them (Doyle 1983a, 1983b; Russett 1993). Similarly, societal strands of commercial liberalism (Keohane 1990:186–7; Doyle 1997:230–50; Mansfield and Pollins 2003:2–3) posit that states will avoid the use of force against trading partners because business sectors that stand to lose as a result of war will compel the government to avoid disrupting the normal economic relationship, regardless of strategic considerations (e.g., Papayoanou 1996). In each case, the liberal state is a pluralist entity that responds to the balance of domestic interests and preferences when conducting foreign policy, selecting the policy option that reflects aggregate societal preferences (Moravcisk 1997:518–20). Neomarxist theories of foreign policy and international politics also assume that the state is captured by the dominant societal coalition – namely, the capitalist class – and consequently that it pursues their interests and preferences regardless of concerns of relative power and strategic conditions (e.g., Kolko 1969; Magdoff 1969; Luxemburg 2003).

In contrast, neoclassical realists have a markedly different conception of the state (Taliaferro et al. 2009:23–8). For both strands of neoclassical realists, the essence of the state consists of the foreign security executive, comprised of the head of government and the key ministers and officials charged with the conduct of foreign security policy (cf. Hermann et al. 1987; Lake 1988; Ripsman 2002:43–4). This executive has access to privileged information about foreign affairs from the diplomatic corps and intelligence services that makes it more aware of the interests of state than other domestic or societal actors. For this reason, the state is distinct from society. Although state leaders are drawn from society, their attitudes and preferences change when they experience “the view from the top,” as the privileged information they receive and the raison d’état culture they become imbued in make state actors more than simply representatives of their societal coalition. Thus, it is not uncommon for a candidate for office to espouse a policy position critical of the incumbent, only to abandon it after assuming power and being briefed by government officials. High-profile examples include President Barack Obama backpedaling on his promise to remove US forces from Iraq in short order and President Bill Clinton's commitment to link the extension of “Most Favored Nation” status to China to Beijing's human rights record.

The state, therefore, occupies a critical position at the intersection between the domestic and international arenas. It is uniquely situated to respond to the challenges and opportunities of the international system because of its privileged information. Yet, it may still misperceive systemic stimuli, such as relative capabilities, whether an adversary's force posture is offensive or defensive, or the time frame of threats they may face. Moreover, while the state is potentially autonomous of societal forces, it is not necessarily so. Depending on domestic political arrangements, states vary in their ability both to enact policy responses to international challenges and to raise revenue and resources to implement policy choices (e.g., Ripsman 2002; Taliaferro 2006; Blanchard and Ripsman 2008). Less autonomous states must frequently bargain with societal groups and key veto players in the legislature over foreign security policy. Thus, for example, both Bush administrations spent months garnering domestic support for their respective strikes against Iraq before they felt capable of waging war. Furthermore, while neoclassical realists maintain that the state conducts foreign policy primarily with regard to the international arena, a matter which puts them squarely within the realist camp, they recognize that leaders also must be attuned to threats to their power position from within the state; therefore, in rare times of extreme domestic instability, national leaders might actually conduct foreign policy with greater attention to the domestic audience than to international exigencies (Ripsman 2009). This can explain why leaders whose power positions at home are threatened may launch diversionary wars that might not be in the national interest (Levy 1989; Smith 1996; McLaughlin Mitchell and Prins 2004). Thus, neoclassical realists have a considerably different view of the state than Innenpolitik approaches.

Furthermore, unlike Innenpolitik theories of foreign policy, neoclassical realists assume that not only the state, but societal actors themselves must frame their policy preferences in reference to the international challenges and opportunities they face. Given the importance of security in an anarchic international environment, states should be unwilling to adopt policies advanced by societal forces that could jeopardize national security, despite the domestic political fallout they expect. Therefore, domestic groups that seek to maximize their interests by directing foreign policy in a manner disconnected from the international environment the state faces are likely to be unsuccessful. Thus, for neoclassical realists, the international system focuses and frames the state's foreign policy response, although the shape it will take will be tailored by the domestic arena. For this reason, the approach remains realist and distinct from a liberal, Innenpolitik approach (Rose 1998). In this vein, Brawley demonstrates that in the 1920s and early 1930s Great Britain, France, and the Soviet Union all identified impending German recovery as the leading threat to their security, but their unique societal considerations led them to adopt distinctly different strategies to respond to the threat (Brawley 2009a).

Key Challenges to Neoclassical Realism

Three key challenges to neoclassical realism are of great importance and should be considered. To begin with Jeffrey Legro and Andrew Moravcsik (1999) have criticized a broad range of recent writings – including many we classify as neoclassical realist – claiming to belong in the realist tradition as involving ad hoc additions of variables such as perception, domestic political arrangements, and international norms to fix what they view as a failing realist core. In their judgment, by shifting the causal logic away from the international system, these theorists are, in effect, abandoning the realist tradition and cannot properly be labeled “realists.” For similar reasons, Vasquez (1997) charges that these emendations of realist theory to save a flawed core theory are testimony to the degenerative nature of the entire realist research paradigm. Even some who would classify these theories as belonging in the realist tradition dismiss neoclassical realism as unscientific and ad hoc. Stephen Walt (2002:211), in particular, charged that “Neoclassical realism tends to incorporate domestic variables in an ad hoc manner, and its proponents have yet to identify when these variables have greater or lesser influence.” Moreover, he continues, it “has yet to offer a distinct set of explanatory hypotheses” and “has given up generality and predictive power in an attempt to gain descriptive accuracy and policy relevance.”

Well-specified neoclassical realist theories, however, are not susceptible to the charge of ad hocism and they are clearly consistent with the realist tradition. Theories such as Randall Schweller's (2006) theory of underbalancing, Jeffrey Taliaferro's (2009) theory of emulation and mobilization, and Ripsman's (2002) theory of structural autonomy, for example, stem from both a deductive theoretical analysis of the circumstances under which state behavior is likely to deviate from structural realist expectations and an empirical investigation of carefully chosen cases. As a result, they offer clearly specified predictions in addition to their explanations of particular historical cases. Indeed, one of the distinct missions of the new generation of neoclassical realists is to generate a set of clearly specified propositions regarding exactly when domestic political and leadership variables will have greater causal effect and when policies and outcomes are determined primarily by systemic variables (e.g., Byman and Pollack 2001; Ripsman 2009).

Moreover, they all share common assumptions that attest to their structural realist origins. Specifically, they assume: (1) that the international system is anarchic and, consequently, that states must rely on themselves to ensure their survival; (2) that security is the most important national interest in an anarchic realm; (3) that anarchy makes cooperation difficult, as it leads states to prefer relative to absolute gains (cf. Frankel 1996:xiv–xvii). Indeed, it is precisely these neorealist assumptions that lead neoclassical realists to take as their starting point the primacy of the international system – that is, that states conduct foreign security policy first and foremost with an eye to the anarchic international system because failing to do so could jeopardize national security, the state's overriding priority – utilizing auxiliary variables as intervening variables between systemic constraints and national policy responses. Hence neoclassical realism is a direct descendant of structural realism and is consistent with the underlying principles of realism (Feaver et al. 2000:175–6, 181; Rathbun 2008). Furthermore, since neoclassical realist theories of foreign policy of the second variety described in this essay seek to do more than simply explain away anomalies in structural realism and aspire to construct clearly testable models, they cannot be labeled “degenerative” (cf. Schweller 2003).

It should be noted, however, that not all eclectic, cross-paradigm approaches are properly classified as neoclassical realist. Those that include a range of international- and domestic-level variables but do not specify the primacy of the international system or the intervening role of domestic or perceptual variables are not consistent with realist theory. In this regard, for example, Peter Katzenstein's (2005) theory of regional orders, which borrows conceptual tools from the realist, liberal, and constructivist paradigms, is more properly labeled as “eclectic,” rather than “realist.” More interestingly, Benjamin Miller's (2007) innovative theory of regional war and peace actually reverses the causal logic of neoclassical realism. By arguing that the war proneness of regions depends primarily on domestic and regional factors, such as the state-to-nation balance, and is only modified by the global balance of power and great power behavior, Miller actually posits only an intervening role for systemic variables, while affording causal primacy to unit and region-level variables. His theory, therefore, while intriguing, is not realist.

A second key challenge that neoclassical realism is susceptible to is the charge that it is comparatively inefficient. By including domestic political and perceptual variables within their models of foreign policy and international relations, neorealist theories are less parsimonious than structural realist theories (Walt 2002:211; Freyberg-Inan et al. 2009:259). For Waltz, who prides himself on parsimony (Waltz 1986:330), explaining important elements of international politics and state behavior with a single variable (the distribution of capabilities – Waltz's other variable, system structure, does not actually vary, since the system has been anarchic for centuries and is not likely to change, as states are reluctant to surrender their sovereignty (Waltz 1979:chaps 5–6)) would be a considerable disadvantage of neoclassical realism. Neoclassical realists counter, however, that while structural realism is a useful starting point, it sacrifices too much explanatory power at the altar of parsimony. The inclusion of unit-level variables, provided it is done in a careful, scientific manner, can add significantly to our ability to explain past events and predict future state behavior (Taliaferro et al. 2009:23). Since theories of international relations are intended not only for their esoteric value, but to serve policymakers as guides for action, the more precision we can generate in a systematic manner in our theories about how a particular state is likely to react in a particular set of international circumstances, the more useful the theory is.

Benjamin Fordham (2009) directs a third critique against neoclassical realism. He charges that it is impossible to separate international and domestic variables, as neoclassical realists attempt to do. Instead, he maintains that domestic interests are themselves conditioned by the international system, just as international threats and opportunities must be interpreted through the lenses of domestic interests. Thus, for example, Soviet power after World War II was not an objective structural threat to France and Italy. Certainly, French Communists did not view it in that manner. Had the Communists dominated the governing coalition, these countries would have defined American power as more threatening than the Soviet Union. For this reason, Fordham rejects what he calls the “additive” approach of neoclassical realism in favor of an “interactive” approach.

Fordham's critique is not, in fact, restricted to neoclassical realism, but to neorealism and Innenpolitik approaches, which also assume that the international and internal influences on foreign policy can be separated. Nonetheless, there is reason to believe that the state, in its privileged location between the domestic and international arenas, can separate international and domestic stimuli to a significant degree. Indeed, anecdotal evidence suggests that the world looks different to those in power. In part because of their access to the privileged information discussed above, in part because of the heavy responsibilities of office, leaders share a “view from above” that diverges significantly from the viewpoints of their social cohorts outside of government. This “reasons of state” focus often leads individuals who achieve power to adopt policies that diverge sharply from their previously expressed preferences (Ripsman 2009:172–3). Thus, for example, Conservative Foreign Secretary Anthony Eden judged that Labour Foreign Secretary Ernest Bevin – a former union leader who was the ideological opposite of the aristocratic Eden – pursued a policy that was completely in line with his own views of the realities of Cold War Europe (Eden 1960:5). Similarly, when Bill Clinton was running for office, he opposed the George H.W. Bush administration's policy of constructive engagement of China, preferring a more aggressive strategy of promoting human rights. After winning the 1992 election and meeting with the outgoing administration, however, he comprehended the wisdom of constructive engagement (Myers et al. 2001). Consequently, there is evidence that what Fordham calls an additive model can be more appropriate than an interactive model, which may overstate the degree to which parochial domestic considerations affect the manner in which governments evaluate geostrategic developments once in power.

Future Research Agenda

Adherents of neoclassical realism thus contend that the approach represents a significant improvement on existing approaches to international relations and foreign policy. There remain, however, a few key avenues for future research. To begin with, it would be useful to test the claim that neoclassical realism adds sufficient explanatory power to warrant the inclusion of a raft of unit-level variables. To this end, studies that engage in careful and systematic testing of neoclassical realist models vis-à-vis leading structural realist and systemic liberal alternatives would be extremely helpful. Similarly, it would be useful to test the assumption that states conduct foreign policy primarily with a view to the international arena with careful tests of the comparative utility of neoclassical realist theories vis-à-vis various Innenpolitik approaches.

Another fruitful avenue of research would be to build upon recent efforts to generate well-specified neoclassical realist theories of foreign policy. Although some of the recent scholarship mentioned in this article has begun to address Walt's charge that neoclassical realism focuses more on post hoc idiosyncratic explanation of particular historical cases than on the construction of generalizable theories, for neoclassical realism to develop into a useful approach for explaining categories of events and to serve as a guide for policymakers, it will be necessary to demonstrate that it is not merely a realm of case narratives, but can also operate as a basis for contextually informed theory. In this regard, Nicholas Kitchen's (2010) efforts to construct a neoclassical realist theory of grand strategic formation and change, utilizing domestic culture and ideas as intervening variables, is encouraging, as it is purely a deductive theoretical enterprise, rather than case driven.

Finally, neoclassical realists would do well to heed Shiping Tang's (2009:799–800) call for more research on the domestic politics of international cooperation in order to shed what he refers to as the “competition bias” of neoclassical realism. While a handful of studies do utilize neoclassical realist approaches to explain cooperative issue areas, such as interstate peacemaking (e.g., Ripsman 2002) or compromise on territorial disputes (e.g., Fravel 2008), Tang is correct that much of the literature on neoclassical realism focuses on interstate competition rather than cooperation. Since realism does not preclude cooperation (e.g., Oye, 1986; Feaver et al. 2000:174), there is no reason to believe that neoclassical realism would be better suited as a tool to explain conflict than cooperation. A more balanced set of empirical foci is, therefore, warranted.