## To the Editors (Ronan Tse-min Fu writes):

In “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” Adam Liff and John Ikenberry claim that “a number of recent developments suggest that the region is ripe for, or may already be experiencing, severe security dilemma-driven dynamics, even arms races.”1 They portray China's rise as the main cause of this dilemma and assert that states must adopt measures to reduce military competition in the region while they still can.

I applaud Liff and Ikenberry for the policy relevance of their research, but their fundamental claim about the prevalence of severe military competition in the Asia Pacific region does not match the empirical reality. The real puzzle is why over the last thirty years Asian countries have shown a surprising lack of interest in boosting their military expenditures in response to China's massive increases, whether these expenditures are measured in absolute or proportional terms, and whether they are measured over the past generation or the last few years.

### IS THERE “SEVERE MILITARY COMPETITION” IN THE ASIA PACIFIC?

Liff and Ikenberry ground their analysis in the empirical observation that Asia might already be engaged in severe military competition, but they never define what they mean by “military competition” (pp. 65–82). They imply that such competition exists when parties seek to enhance their military capabilities in response to an external threat (p. 65). More specifically, they state that “there is evidence of a security dilemma-driven spiral gradually unfolding between China and several states that is driving investments in military capabilities and that may worsen significantly in the years ahead” (p. 88).

If Liff and Ikenberry's claim that the Asia Pacific is currently experiencing severe military competition were true, then major countries in the region should have significantly boosted their military spending over the past few decades.2 After all, China increased its military spending by 576 percent in real terms from 1992 to 2013. And as Liff and Ikenberry note, “the PLA [People's Liberation Army] increasingly poses at least a potential threat to China's neighbors and the United States” (p. 67). A generation of Asian leaders, however, has chosen not to respond with similar arms increases. Measured in constant $2011, from 1992 to 2013 Japan's increase was only 13 percent; South Korea's increase was 96 percent; and India's was 192 percent.3 Measured as a percentage of gross domestic product (GDP), the military expenditures of most Asia Pacific countries have steadily declined since the Cold War, despite the emergence of the PLA as a professional and capable fighting force.4 Few countries in the region have been willing to make massive investments in military power at the expense of other priorities.5 In addition, as a proportion of GDP, Asian military spending in 2013 was the lowest it had been since the end of the Cold War. On average, major Asian countries spent 3.32 and 2.83 percentage of their GDP on defense in 1988 and 1992, respectively. In 2013 the average proportion of GDP that major Asian countries devoted to defense was only 1.88 percent.6 In this sense, contrary to Liff and Ikenberry's claim, military competition in the Asia Pacific is becoming less rather than more intense. Perhaps one could argue that twenty-five years is too long an analytical time frame— that China's neighbors only recently decided its growth was a threat and have begun to respond. The most recent data do not support this claim either. Measured in inflation-adjusted constant terms, the military spending of key countries in the region declined in 2013: Japan —0.23 percent, India —0.74 percent, Taiwan —2.63 percent, and Australia −3.58 percent.7 This is not to suggest that Asian countries are indifferent to China's assertiveness. Asian countries, especially those with territorial (or other) claims, certainly care about China's policy in the South and East China Seas, but they have not responded by engaging in severe military competition, at least when measured in the standard manner. ### LIKELY RIVALS Perhaps one could argue that a broad regional perspective masks military competition between specific countries. For example, Japan, the Philippines, Taiwan, and Vietnam all have enduring disputes with China. Yet even these “most likely” rivals have been taking part in muted military competition or arms race.8 From 2003 to 2013, none has significantly increased its military budget, measured in constant 2011 dollars, despite a massive increase in Chinese military spending over the same period (see figure 1). Figure 1. Military Expenditures, 2003–13 (constant$2011)

Figure 1.

Military Expenditures, 2003–13 (constant $2011) Japan's 2015 military budget increased 0.8 percent in absolute terms, hardly qualifying as remilitarization, much less competition with China.9 Even though Prime Minister Shinzö Abe wants to reestablish Japan as a first-tier nation, the five-year plan he put forth in 2013 might total increases of only$9 billion if it is fully implemented— and even those increases are in doubt given Japan's precarious economic situation. Furthermore, the Japanese people remain skeptical of their government's efforts to engage in a more activist foreign policy and the military's desire for greater influence.10 A poll taken by the Asahi Shimbun in April 2014 shows that 65 percent of respondents opposed Abe's efforts to promote closer security ties with the United States, believing these efforts will not benefit regional stability.11

Meanwhile, in 2014 Vietnam and the Philippines spent a combined total of less than $4 billion on their militaries—roughly 4 percent of China's total budget. Although both countries are steadily modernizing their militaries, neither appears to be aiming to match China's military power.12 From 2003 to 2013, Vietnam's annual military expenditure increased 8 percent on average, and in 2013 it increased only 2.4 percent. These restrained increases are particularly important to note because Liff and Ikenberry claim that Vietnam boosted its military spending 70 percent in 2011 (p. 80). The source they cite, however, reports the huge increase in nominal terms and notes that the increase in military spending could easily “be negated by continuing double-digit inflation in Vietnam and the fall in the value of the local currency.”13 Thus, rather than the 70 percent annual increases in defense spending that Liff and Ikenberry claim, the percentage change in Vietnamese military spending (as measured by SIPRI in constant$2011) was an increase of 11 percent in 2010; a contraction of 6 percent in 2011; and an increase of 16 percent in 2012.14 Moreover, although there is insufficient space to address all of Liff and Ikenberry's claims about Vietnam's military modernization, it should be noted, for example, that Vietnam's expansion of its total principal surface combatants from its current two destroyers to four destroyers by 2017 hardly represents a realistic intention to conduct a significant naval campaign of any type against a great power such as China.

The pattern of Taiwan's military expenditure is even more revealing. Despite the threat to Taiwan's security posed by China's sovereignty claim and the accelerating enhancement of Chinese military capabilities, and despite the uncertainty of the availability of alliance partners, Taiwan reduced its defense effort nearly every year from 1990, where it stood at 5.3 percent of GDP and 19.1 percent of government spending, to 2013, where it stood at 2.2 percent of GDP and 11 percent of government spending. As the most-likely actor to engage in severe military competition with China, Taiwan has chosen a different path to deal with the rise of China.15 This suggests that other Asian countries’ propensity to engage in military competition with China might be lower than Liff and Ikenberry predict.

### A U.S. SECURITY GUARANTEE?

The most common explanation for East Asian states’ muted military response to China's rise is the existence of a robust U.S. security guarantee that reassures allies and deters rivals.16 Given that the military expenditures of most Asia Pacific countries have gradually declined for the past quarter century despite “China's surging defense budget” (p. 67), and that the United States continues to “pivot” and increase its forward military deployment (p. 82), it might be that the region feels safer because of the U.S. military presence. If so, U.S.-China security competition would be a cause for concern, but not China-East Asian security competition. Yet Liff and Ikenberry miss this key argument—the bulk of their article focuses primarily on China-East Asian security dynamics (pp. 65–82). Either states are free riding on U.S. defense commitments and hence not spending what they might, or they have doubts about the U.S. military guarantee against all contingencies. Liff and Ikenberry cannot simultaneously argue that states are afraid of China and desperately arming themselves but also unafraid of China given the U.S. security guarantee (p. 88).

Although the U.S. alliance system may mitigate national security concerns for some Asian countries, it is unlikely to reassure every one of them. After all, the U.S. security guarantee is not a blanket guarantee for all military contingencies. There is a simpler explanation for continued low military expenditures for both allies and non-allies in the region: few states fear for their survival; hence they are not arming as much as Liff and Ikenberry suggest.17 Although China's maritime claims are expansive, they are not new, nor are they increasing.18 And most important, China is unlikely to embark on territorial expansion.19 Direct military competition with China is thus unwarranted. The United States and China may be competing for regional hegemony, but few other states feel the necessity to choose sides. Given that the stakes are fairly low, it is not surprising that few states appear willing to make the costly domestic and economic trade-offs required for major sustained investments in their militaries.

### CONCLUSION

A systematic examination of the evidence reveals that countries in the Asia Pacific are making only modest increases in their military spending. Contrary to Liff and Ikenberry's assertion, East Asia is experiencing neither intensifying military competition nor an arms race. Furthermore, by most measures, China has already risen to regional dominance without causing major disruptions to regional stability. The real question, therefore, is why counterbalancing against China or an arms race has not occurred, despite three decades of extraordinary Chinese economic and military growth.

Ronan Tse-min Fu

Los Angeles, California

## To the Editors (David James Gill writes):

In “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” Adam Liff and John Ikenberry argue that a security dilemma-driven spiral is developing in the region that is leading to increased tension and military competition, particularly between China and the United States.1 Liff and Ikenberry deserve praise for providing a useful elaboration of security dilemma theory and for offering an informative empirical survey of Asia Pacific states’ security policies. Nevertheless, their argument would be more convincing if they had addressed two problems, which I discuss below.

### AMBIGUOUS EVIDENCE

The first problem is that Liff and Ikenberry sometimes rely on ambiguous evidence to support their argument that a security dilemma-driven spiral is unfolding between China and the United States. They accept that China's “true intentions are unknowable” (p. 91), and they note that the security dilemma is not the only cause of military competition taking place in the Asia Pacific, pointing to other possible sources such as “concrete conflicts of interest,” “domestic politics,” or the pursuit of “international prestige” (p. 60). Yet Liff and Ikenberry conclude that security dilemma dynamics are a major driver of China's actions in many instances.

To support this claim, they cite rises in military spending, quantitative and qualitative enhancements in China's military capabilities, and “leadership rhetoric and political statements” (ibid.). Liff and Ikenberry also adopt a four-question test to determine whether security dilemmas exist in the Asia Pacific. They do not persuasively show, however, why theirs is the best interpretation of the data that they have assembled. Further detail would help to clarify their key assumptions—namely, how they select evidence, how they distinguish between greedy and security-seeking states and how they separate and assess other drivers of state behavior.2

Much of the evidence cited by Liff and Ikenberry remains ambiguous. An alternative explanation of China's behavior could reasonably attribute it to revisionist ambitions. Rises in military spending and efforts to enhance strategic capabilities might reflect greed more than insecurity, whether regional or global in ambition.3 Walter Russell Mead, for instance, has cogently argued that China is a revisionist power.4 Ikenberry has previously noted that China desires “greater regional influence” and, “where possible, regional domination.”5 Liff and Andrew Erickson have asserted that China holds “irredentist” concerns and has increased military spending in part to “improve capabilities to address outstanding territorial and maritime claims.”6 The rhetoric and political statements of China's leadership are also open to a variety of interpretations. The examples selected by Liff and Ikenberry may not accurately reflect the leadership's motives. Indeed, the authors recognize elsewhere the existence of Chinese political figures’ revisionist rhetoric and the ambiguity of Beijing's statements concerning great power relations.7

Revisionist interpretations of China's military buildup and foreign policy behavior enjoy two clear advantages over those emphasizing the security dilemma. First, they do not assume persistent misunderstanding among experienced policymakers or ignorance of the security dilemma in Beijing or Washington.8 Second, as detailed below, they enjoy stronger support from the historical record, which provides limited evidence for major competition between status quo powers driven by fears of potential insecurity.9 None of this is to suggest that security dilemma theory is irrelevant to the study of strategic competition. Several theorists have argued persuasively that states in competitive environments that are unable to balance deterrence and reassurance appropriately can risk unnecessary escalation.10 Such arguments, however, often explain military competition with reference to a mix of insecurity and revisionist aims rather than as being directly attributable to uncertainty about strategic intentions or the “type-1 strategic setting” that Liff and Ikenberry suggest exists in parts of the Asia Pacific (pp. 63–64, 82, 86). Uncertainty in anarchy matters, but the crucial issue is the extent to which it explains China's behavior. Ambiguous evidence makes it difficult to reach firm conclusions concerning the importance of the security dilemma as a driver of military competition in the region.

### THE SECURITY DILEMMA AND THE HISTORICAL RECORD

The second problem with Liff and Ikenberry's argument concerns their assumption that “history teaches us to be wary of security dilemma-induced military competition and war” (p. 91). They could be clearer as to which events from the past are suitably instructive. Often-cited examples from security dilemma scholarship include the Anglo-German naval buildup before World War I, the outbreak of World War I, and the Cold War. Yet all three cases are contentious. Some political scientists have drawn on historical research to argue that the security dilemma plays a more limited role than revisionist aims and domestic politics in explaining the origins of World War I and the root cause of the Cold War.11 Scholars of international relations, however, have not fully engaged with the historiography concerning the security dilemma's role in the Anglo-German naval arms buildup.12

Conventional accounts based on security dilemma theory of the Anglo-German rivalry cast Britain and Germany as status quo states seeking security. These accounts suggest that Germany perceived Britain's huge navy as a threat because it increased the risk of coastal raids, trade interdiction, and economic isolation. To counter this perceived threat, Germany started building its own powerful navy for defensive purposes. Britain in turn saw Germany's navy as an offensive force. Uncertainty about intentions thus drove escalatory balancing on each side. Yet the Anglo-German naval buildup is far from a clear-cut case of the security dilemma in operation.

Some international relations scholars, though continuing to note the importance of insecurity, recognize that Germany did intend to challenge the status quo.13 Indeed, Liff and Ikenberry claim that a perceived threat was not the primary explanation for Germany's naval buildup, but suggest that its actions could have generated insecurity in others (p. 64, n. 27).14 Few scholars have engaged with the historical literature concerning British naval policy, however, which remains vital for claims concerning an action-reaction cycle. Many historians believe that fear of Germany is an inadequate explanation of British naval policy before World War I.15 Revisionist scholars suggest that imperial interests and domestic-political concerns, especially financial constraints and interservice rivalries, were major drivers of many of the British Admiralty's strategic decisions.16

The historical record does not provide clear lessons about insecurity creating or exacerbating military competition. Historians also offer limited support for the argument that security dilemmas drive serious rivalries between status quo powers. Given the tremendous stakes that would be involved in a U.S.-China conflict, scholars should consider the lessons of the past carefully. Those who use historical examples to highlight the importance of the security dilemma in explaining strategic rivalry would benefit from engaging with the historiography to better support their claims.

Liff and Ikenberry provide a series of policy recommendations for U.S.-China relations based on their research. Their suggestions could be helpful if a U.S.-China security dilemma exists, but ineffective or counterproductive if other explanations for China's foreign policy behavior prove more valid. Establishing and strengthening diplomatic mechanisms for bargaining presents limited risks to either side, though it would do little to stop a revisionist power driven by greedy tendencies. In contrast, increasing the transparency of military capabilities, strategic objectives, and decisionmaking could prove damaging to U.S. security. Neither the available evidence nor the historical record instills sufficient confidence in policy recommendations based on the assumption that the security dilemma is the major driver of military competition between China and the United States. When making such important decisions, scholars and policymakers should not race to conclusions. Further debate is essential to clarify the causes of strategic rivalry.

David James Gill

Nottingham, United Kingdom

## To the Editors (Eric Hundman writes):

Adam Liff and John Ikenberry argue that the security dilemma is an important driver of military competition in the Asia Pacific, one with the potential to worsen and lead to a dangerous arms race in the region.1 In so doing, they provide a useful overview of many military developments in the region and highlight the enormous importance of identifying security dilemma dynamics. Liff and Ikenberry's conceptualization of the security dilemma and their empirical tests for it, however, are flawed in ways that undermine many of their claims.

### PERCEPTIONS, INTERESTS, AND MOTIVES

Liff and Ikenberry characterize the security dilemma as a situation in which two states are driven to pursue costly, potentially destabilizing competitive policies “because of insecurity and uncertainty about the other's true intentions” (p. 54). By under-emphasizing the role of misperception, however, this characterization diverges in important ways from those of the models they cite in their article.

Robert Jervis defines the security dilemma as a situation in which “many of the means by which a state tries to increase its security decrease the security of others.”2 Such attempts can be self-defeating for such a state when others’ reactions reduce its security, even if all actors avoid misperceptions and behave rationally. Charles Glaser, for instance, expanded scholars’ understanding of the security dilemma by elaborating the rational logics through which its dynamics can develop even if misperceptions do not occur.3 Perceptions, however, can also be critical drivers of the security dilemma. Jervis brings them into play through his discussion of a state's subjective security demands: “Decision makers act in terms of the vulnerability they feel, which can differ from the actual situation.”4 For Jervis, perceptions play such a central role in the security dilemma that even though the defense turned out to have an “overwhelming advantage” over the offense in World War I, alliance patterns, defense budgets, mass opinion, and attitudes toward preemption were all better explained by the widely shared, erroneous perception that the offense had the advantage.5 Glaser, too, notes that “the effects of structure are mediated through states’ perceptions, whether accurate or not.”6 Perceptions play an even stronger role in newer models of the security dilemma, which high-light its nature as a socialization process in which two states form impressions of each other; perceptions influence whether this process leads to stability or to the construction of adversarial dynamics.7

Although Liff and Ikenberry do note that their theoretical model describes situations that can lead “one side to interpret the other's defensive measures as offensive” (p. 58), they downplay the possibility of misperception. Instead, they focus on uncertainty as the source of this kind of misinterpretation, arguing that “mistrust and uncertainty about intentions lead one side to interpret the other's defensive measures as offensive” (ibid.). This kind of misinterpretation can result, however, from either uncertainty or misperception. Liff and Ikenberry's model does not clearly distinguish between the two, which undermines their analysis because different drivers of the security dilemma demand divergent policy responses. It is possible, for instance, for an actor to have the knowledge and skill needed to ascertain its opponent's intentions but still perceive them incorrectly; conversely, an actor can perceive an opponent's action correctly, but nonetheless remain uncertain about what that action conveys about the opponent's intentions. This is a core insight of the security dilemma—it is not always possible to clearly signal intentions. Misperceptions can either generate this problem or exacerbate the effects of uncertainty.

In addition, perceptions are critical to security dilemma theory because they govern a security-seeking actor's assessment of its security requirements. For instance, perceptions will determine whether or not a state thinks it needs more territory to be secure. Accordingly, both Jervis and Glaser note that security-seeking states can desire territorial expansion.8 Liff and Ikenberry, though, equate “security-seeking” states with “status quo” states (p. 60), and argue that security dilemmas can exist only between “status quo, defensive-oriented states” (p. 63). In their model, no security dilemma can be present if “one or more states seek changes to the status quo in a fundamentally zero-sum manner” (p. 64). They identify “territorial disputes” (pp. 69, 79), “material and territorial interests” (pp. 82, 86), and “interest-based disputes” (p. 86) as examples of the kind of zero-sum conflicts in which their model of the security dilemma does not apply (p. 64).

Equating security-seeking states with status quo states is problematic as well, because doing so can make sense only when states (1) never perceive a need for more territory to ensure their security and (2) never disagree over the status quo. For Liff and Ikenberry, it seems, security-seeking states can never desire territory claimed by another state or disagree with others about the nature of the status quo, which is inconsistent with the models of the security dilemma they cite. It also does not accord with the reality of ongoing disputes about the status quo in the Asia Pacific.9

More important, conflating status quo inclinations and security-seeking intentions renders Liff and Ikenberry's model unable to distinguish between security dilemma-driven and adversarial dynamics. For example, Liff and Ikenberry argue that because military competition between Vietnam and China is “driven primarily by perceived revisionist behavior” (p. 86), it is not driven by the security dilemma. At the same time, they argue that Australia is “responding to a perceived … threat from China” (p. 69), but that this competition is driven by the security dilemma. They distinguish between the two by arguing that the threat to Australia is “uncertain” (ibid.), whereas the threat to Vietnam is “specific” (p. 86) and rooted in “material and territorial interests” (p. 82). Identifying security dilemma dynamics, however, requires determining if states have security-seeking motives, which Liff and Ikenberry's assessment of threats’ certainty or specificity does not do. Further, if the presence of a “direct conflict of interest” rules out the applicability of Liff and Ikenberry's security dilemma model (p. 64), as they claim, then it should not be possible to apply this model to any of the many states they examine that, in their own words, “may … have long-standing interest-based disputes with China” (p. 86).

### OFFENSE, DEFENSE, AND THE SECURITY DILEMMA

Whereas the potential for security dilemma dynamics centers on actors’ intrinsic motives, the nature and severity of such dynamics depends especially on two “offense-defense variables”: whether the offense is distinguishable from the defense, and whether the offense dominates over the defense. Liff and Ikenberry themselves note that “the distinguishability of offensive and defensive forces and the defense having the advantage is stabilizing” (p. 59), and they initially set out to test for distinguishability and a defensive advantage in the Asia Pacific. Their proposed empirical test for determining the presence of a security dilemma in the region includes four questions, two of which concern offense-defense variables. The first asks, “What is the offense-defense balance in the Asia Pacific?” (p. 64); the second asks, “Are allegedly defensive measures/weapons distinguishable from offensive ones?” (ibid.). Liff and Ikenberry never answer these questions in their empirical analyses, however.

As Jervis noted, “If the defense has the advantage, and if the status-quo powers have reasonable subjective security requirements, they can probably avoid an arms race. Although an increase in one side's arms and security will still decrease the other's security, the former's increase will be larger than the latter's decrease…. Thus a stable equilibrium can be reached.”10 Therefore, because Liff and Ikenberry do not assess the offense-defense balance, they cannot answer Glaser's question about security dilemma dynamics: “Why does not this action-reaction process simply leave the state's security unchanged, since the adversary's reaction could just offset the state's action?”11 Further, because they do not examine the distinguishability of offensive and defensive forces, readers are left even more uncertain as to the severity of any security dilemma dynamics. As Glaser puts it, “Offense-defense differentiation has the potential virtually to eliminate the security dilemma.”12 In some cases, an action-reaction spiral can actually increase security,13 but because Liff and Ikenberry do not investigate the very offense-defense variables they highlight as critically important, they provide little basis on which to judge whether existing dynamics are more likely to increase or decrease security for states in the Asia Pacific.

### EMPIRICAL INFERENCES AND CASE SELECTION

Liff and Ikenberry propose two other questions as part of their test for the presence of a security dilemma in the Asia Pacific. First, are “clashing interests” driving military competition, “with one or both sides being greedy” (p. 65)? Second, does each state have “a good understanding of the other's motives,” especially as indicated by “relative levels of military transparency” (ibid.)? Liff and Ikenberry investigate these questions, but not in ways that support their claims. For instance, they endeavor to identify the presence of Glaser's greedy states by looking for concrete conflicts of interest. But Glaser defines “greed” in this context on the basis of intrinsic interests, as “motives beyond security”;14 a greedy state is thus one that is not intrinsically security-seeking. Having interests that clash with those of another state does not necessarily mean a state has motives beyond security, nor does a lack of “concrete” clashes of interest mean a state is fundamentally security-seeking. Liff and Ikenberry's focus on concrete conflicts of interest at the expense of evaluating the intrinsic motives critical to security dilemma theory therefore further weakens their analysis.

Liff and Ikenberry go on to highlight the critical role of transparency in understanding states’ motives. Their investigation of transparency is flawed in two ways, however. First, although they correctly note that transparency can increase a state's understanding of another's motives (allowing a determination of whether a state is seeking security), they do not acknowledge that in terms of security dilemma theory, transparency exposes motives only to the degree that the offense and the defense are distinguishable. If they are not, a state's motives are likely to remain unclear even in cases of high transparency.15 Transparency is therefore a poor proxy measure for others’ ability to understand motives. Second, Liff and Ikenberry do not examine the level of transparency of actors besides China. Focusing exclusively on China's limited transparency undercuts Liff and Ikenberry's recommendation that both the United States and China should improve their transparency efforts. Doing so also implies that any security dilemma dynamics in the Asia Pacific are solely China's fault, when in fact the security dilemma is a fundamentally dyadic phenomenon. When suggesting improved transparency on both sides, for instance, instead of explaining how the United States might reassure China by increasing transparency, Liff and Ikenberry suggest that “Washington must more effectively explain why exactly China's low transparency harms strategic trust” (p. 90).

Liff and Ikenberry thus only partially apply their proposed test for the security dilemma. The rest of their empirical analysis focuses on assessing “the drivers of … efforts” to increase military capabilities (p. 65). Although they provide no explicit criteria for identifying such drivers, their standard seems to be whether a state says China is driving increases in its military capabilities. For instance, they assert that some of Japan's military “upgrades are driven by the growing missile threat posed by North Korea, but a number of recent trends are a direct response to China's military buildup” (p. 73). Official assessments notwithstanding, though, Japan could in fact be responding to one or both of these pressures—or, as Liff and Ikenberry note, to other pressures such as “domestic politics or pursuit of coveted … international prestige” (p. 60). So, how are we to identify drivers in any analytically consistent way? Further, even if we accept their assessments, because Liff and Ikenberry do not effectively test for the security dilemma, readers are still left uncertain if responses to China have the potential to spiral dangerously. If the security dilemma is not present, for instance, it is possible Japan could be responding rationally to China's military development in a way that could lead to a stable equilibrium.

Liff and Ikenberry's case selection further undermines their analysis in two ways. First, they omit mention of the substantial cooperative exchanges occurring between China and other regional actors, thus foregrounding adversarial dynamics.16 Second, they focus on those among China's neighbors in Asia that are quickly developing their military capabilities. Omitting states such as Malaysia, the Philippines, and Taiwan likely biases Liff and Ikenberry's results, because all of these states also perceive a threat from China and often have overlapping territorial claims, but are developing their military capabilities relatively slowly.17

Ultimately, this selection problem, in combination with the other problems discussed above, leaves Liff and Ikenberry open to charges that their article's analysis is shaped by the very fears, uncertainties, and misperceptions driving the security dilemma that appears to be operating between China and the United States.18 Because their analysis can neither determine when the security dilemma is present nor characterize its relative severity, it cannot support even their qualified claim that security dilemma dynamics are driving a “spiral … that may worsen significantly in the years ahead” (p. 88). That they nonetheless make such claims while also criticizing China (and China only) for lacking transparency leaves them open to accusations of threat inflation. Their article could thus, ironically, fuel the intensification of security dilemma dynamics in the Asia Pacific.

### AN UNCERTAIN ROLE FOR THE SECURITY DILEMMA

Security dilemma dynamics are undoubtedly creating additional challenges in the Asia Pacific, but Liff and Ikenberry's analysis cannot identify where or how strongly the security dilemma is operating in the region. This limitation is perhaps one reason why their policy proposals concern only the U.S.-China relationship and focus on nebulous suggestions such as improving “the wider political and strategic context in which military competition is unfolding” (p. 90). Doing so would certainly reduce tensions, but lacking a clearer picture of security dilemma dynamics in the region, Liff and Ikenberry's analysis cannot provide more specific advice about how best to improve the security situation in the Asia Pacific.

Eric Hundman

Chicago, Illinois

The letters by Ronan Tse-min Fu, David Gill, and Eric Hundman highlight the lively debate on the causes, consequences, and degree of military competition in the contemporary Asia Pacific. Their thoughtful critiques provide a welcome opportunity to engage our critics on these important issues and to clarify several points from our article, “Racing toward Tragedy?”1

In our piece, we set out to empirically assess the validity of an idea too often imputed uncritically in academic and popular discourse: that the contemporary Asia Pacific is rife with arms races and security dilemmas. Drawing on security dilemma theory, we developed a systematic empirical test to assess these claims, particularly whether security dilemma dynamics drive military competition in the region. We broke down strategic interactions dyadically and looked at specific causes and consequences. Noting that “the transition from the realm of abstract theory to the empirical world is often treacherous” and that it “requires simplified abstractions of a complicated world,” we believe that our effort is nevertheless worthwhile given the theoretical and policy significance of China's rise and security dynamics in the Asia Pacific (p. 63). In our view, effective policy responses aimed at ameliorating tensions should be based on an understanding of the underlying drivers of regional frictions.

To ensure analytical tractability, we simplified our approach to look for evidence of two basic, notional “types” of competitive security dynamics between states: Type 1, or situations in which both sides’ security dynamics are basically aligned and action-reaction cycles are caused primarily by mistrust and uncertainty about each other's intentions; and Type 2, or those where the core driver of military competition is a direct conflict of interest (pp. 63–64). Our empirical analysis yielded mixed results, including variation across dyads (pp. 65–88). On the one hand, we found evidence of uncertainty about intentions driving mutual arming. Policies understood as defensive oriented by one state generate insecurity and military responses on the other side that make both sides less secure, triggering new rounds of competition. Insecurity is exacerbated by each side's disparate perceptions about offensive versus defensive capabilities. Both sides see their actions as defensive and the other's as threatening. On the other hand, our analysis also suggests that security dilemmas are not the only driver of regional frictions and action-reaction military competition. In particular, we identified direct conflicts of interests over territory and features in the South China and East China Seas as important drivers. Although our analysis revealed practically significant levels of military competition and grounds for concern that it may worsen, we disagree with those who suggest that the region is rife with arms races or that military conflict is likely, much less inevitable. Moreover, both sides can adopt measures to ameliorate frictions and to manage disputes peacefully (pp. 87–91).

### MISSING THE BIGGER PICTURE? (RESPONSE TO RONAN FU)

Ronan Fu commends our article for its “policy relevance,” but he argues that our assessment of contemporary dynamics in the Asia Pacific is inconsistent with the “empirical reality.” Fu states that “East Asia is experiencing neither intensifying military competition nor an arms race.” He also claims that “counterbalancing against China” is not occurring, despite China allegedly having “already risen to regional dominance.”

We appreciate Fu's extensive analysis of military expenditures and agree with him on at least two important points, both of which we mentioned in our article: (1) defense spending trends are noteworthy; and (2) they provide important evidence of the relatively moderate scale of military competition in the Asia Pacific, especially compared with earlier, far more worrisome periods, such as the Cold War or pre-World War I era (p. 87). Nevertheless, we disagree with Fu's analysis in two fundamental respects. First, he repeatedly mischaracterizes our argument and empirical analysis, neither of which he engages directly. Second, he systematically overlooks the very metrics we argued are essential to measuring regional military competition.

Fu's mischaracterization of our argument begins with the first sentence of his letter. In it, he quotes our observation that many scholars and commentators claim that security dilemmas and arms races are occurring throughout the region. The manner in which he does so misleadingly implies that this is our assessment of regional dynamics, rather than what it was: our motivation for conducting the study. As we made clear, our article's purpose was to empirically assess the validity of these widespread, often uncritical and excessively simplistic, claims. Our findings suggest a more complex reality in the Asia Pacific, and we noted variation in outcomes. Fu's claim to the contrary, nowhere in our article do we “assert that states must adopt measures to reduce military competition in the region while they still can.” Such language implies a cynical fatalism absent from our argument.

We do argue that practically significant military competition is occurring in the Asia Pacific and that moderate security dilemma dynamics play a role in some dyadic relations and realms. Unlike Fu, we are concerned about the current level of military competition, mistrust, and possibility of miscalculation, and we believe that tensions and arming on both sides could worsen. Nowhere, however, did we argue that a destabilizing spiral of the sort Fu suggests, much less war, is inevitable. On the contrary, we concluded that there is no evidence of an “ideal-type security-dilemma conflict,” which we defined as “a full-scale arms race, manifest in surging military expenditures and massive increases in weapons procurement on both sides” (pp. 54, 63, 87). Moreover, we wrote that “increases to defense spending appear roughly consistent with economic growth,” and that in response to perceived threats, states are adopting “targeted measures, rather than reflexive, reciprocal buildups reminiscent of the dreadnought race of a century ago” (p. 87). As history teaches us, “Not all cases of rising powers end tragically. No outcome is inevitable” (p. 91). We also noted that opportunities for cooperation and mutual gains abound, including in the military domain, and that such efforts can significantly ameliorate frictions (p. 90).2

In addition, we do not follow Fu's analysis where he seemingly lays out several false dichotomies, further mischaracterizing our argument. For example, Fu claims that we “cannot simultaneously argue that states are afraid of China and desperately arming themselves but also unafraid of China given the U.S. security guarantee.” Nowhere did we put forth such a claim. Many states in the region are pulling the United States closer while enhancing their military capabilities. Fu's suggestion that states must do one or the other not only presents a false policy choice between internal and external balancing measures; it is also inconsistent with decades of theoretical scholarship on balancing and alliances.

Our second major issue with Fu's critique centers on his attempt to challenge our empirical findings by relying almost entirely on defense spending trends. We agree that defense spending matters—indeed, we incorporated it in our own analysis. Fu's narrow approach, however, exemplifies the potential pitfalls of what we argued is a “focus on the most conspicuous, and often easily quantifiable, metrics” (pp. 59–60). Fu systematically overlooks operationally significant qualitative and quantitative internal and external balancing responses to both abstract and concrete perceived threats. The consequences of Fu's empirical selection bias are significant: he summarily dismisses the idea that any counterbalancing or military competition is occurring in the Asia Pacific. We disagree. Moreover, Fu systematically neglects important changes pursued by various states across the region (including the United States) in doctrine, force structure, and force posture; measures to enhance alliance coordination and military interoperability, to preposition military assets, to expand joint exercises and force rotations, and so forth—often in response to perceived threats vis-à-vis Beijing.3 China's own threat perceptions, conversely, are shaped in part by these measures.

The way in which Fu analyzes his chosen metric is also problematic. He uses defense spending as a percentage of gross domestic product (GDP) over time to claim that military competition in the Asia Pacific is “becoming less rather than more intense.” Fu's analysis thus overlooks important trends that undermine his core arguments. First, rapid economic growth has in important cases allowed major increases in absolute defense spending, which in turn have resulted in practically significant military capabilities enhancements that Fu's assessment of spending based on percentage of GDP does not consider. Second, irrespective of aggregate spending trends, the operational focus of key countries’ military investment has not been static over the past quarter century; rather, it has shifted qualitatively, quantitatively, and geographically in terms of doctrine, force structure, and force posture to prioritize new threats over old ones. Fu problematically sets the 1980s as his baseline—a period when most countries’ investment targeted a perceived existential threat posed by the Soviet Union and when Beijing was for a time aligned with Washington and its allies against Moscow. Since the Soviet Union's collapse, however, the foci of investments have shifted significantly toward new perceived threats (e.g., from China). Fu's decision to start his analysis of defense spending trends before the Cold War's end ensures that it lends minimal purchase to the question of whether states today perceive China as a threat and are balancing against it militarily. It presents a straw man, as we have never claimed that military competition in the Asia Pacific is tantamount in severity to a “new Cold War” or that any state is adopting a containment strategy vis-à-vis Beijing.

Factual errors further weaken Fu's argument. For example, his claim that “military expenditures of most Asia Pacific countries have gradually declined for the past quarter century” is incorrect. Nor is it true that China's neighbors have decided against making “major sustained investments in their militaries.” Taken together, these statements imply that absolute defense spending is declining and that regional militaries are doing nothing to modernize or bolster their capabilities. Both claims are false.

Employed appropriately, defense budgets can be one useful metric for measuring balancing behavior, because they are key enablers of investments in effective military power. At best, however, analyses of regional dynamics based on defense spending trends alone are incomplete. Defense budgets constitute inputs, not outputs. The bigger, more important question for measuring balancing behavior is whether states are adopting operationally significant internal and external measures to improve their military capabilities to deter and, if necessary, defeat specific perceived threats.4 In the cases we examined, the answer appears to be yes, albeit to varying degrees. Fu's restrictive and problematic methodology leads him to categorically dismiss the idea that any balancing against China is occurring—a claim we find significantly at odds with the empirical record.

### THE AMBIGUITY OF UNCERTAINTY (RESPONSE TO DAVID GILL)

Like Fu, David Gill raises empirical questions about our analysis of security dilemmas and arms races in the Asia Pacific. First, Gill claims that we “sometimes rely on ambiguous evidence to support [our] argument.” We are sympathetic to this critique, which highlights a theoretical and methodological issue fundamental to any effort to empirically test for the presence of security dilemma-driven military competition: because perceptions of states’ intentions are fundamental to the security dilemma, much is necessarily left to interpretation. There are four reasons why we have no major issue with Gill's claim that our evidence is “sometimes” ambiguous. First, it implies that sometimes our evidence is unambiguous. Second, Gill acknowledges that the evidence for his counterargument is also contentious (read: ambiguous). Third, our goal was to answer a specific research question as objectively as possible using available evidence about leaders’ perceptions and the policies driven by them, and we note evidence of mixed results: (1) variation occurred across dyads in the degree to which security dilemma dynamics appeared; and (2) both Type-1 and Type-2 dynamics can manifest simultaneously. We did not claim that security dilemmas are the only driver of military competition in the Asia Pacific, and we resisted assigning a “one-size-fits-all” label to describe it. Fourth, regarding Gill's concerns about Beijing, we noted the ambiguity surrounding China's intentions, and we stressed that our conclusions are “preliminary” (pp. 57, 86, 91).

The roles that uncertainty and perceptions of states’ intentions play in worsening tensions are fundamental to security dilemma theory. We argued that these elements manifest on both sides of the strategic interaction—both in internal and external government documents and in public discourse. With reference to China, Beijing's relatively low level of military transparency exacerbates the difficulty of judging its true intentions—not just concerning its controversial sovereignty claims, but also how it will go about pursuing its self-identified national interests—an issue as important for peace and stability as what Beijing defines those interests to be. If they were granted full data access, not even the most prescient of Chinese leaders could accurately predict China's future intentions in every instance (p. 57). Given this uncertainty, we endeavored to answer the following question: Do other states’ defense planners typically see Chinese moves as offensive and threatening while leaders in Beijing see those same moves as defensive, and vice versa? Our analysis suggests that the answer is yes (p. 56). We argued that particularly in light of China's rapidly increasing material wherewithal and military capabilities, its low level of transparency deepens uncertainty and buttresses worst-case assumptions, thereby exacerbating security dilemmas.

Second, Gill takes issue with the historical underpinnings of security dilemma theory, and his analysis suggests that he favors exclusively “revisionist interpretations” of China's military buildup and behavior. With regard to China's behavior, however, Gill's reasoning is unclear. We appreciate his thought-provoking references to the Anglo-German rivalry prior to 1914. But as Gill concedes, scholars have long contested the idea that insecurity was not a driver of tensions between London and Berlin. Regardless, we believe that Gill overreaches when he apparently uses this single case to generalize that history provides no “clear lessons” about whether insecurity can contribute to military competition, and to suggest that states will arm only when they have revisionist/offensive ambitions. We disagree and find such a claim empirically disprovable. Specific to the Asia Pacific, Gill provides no empirically grounded justification for the “China as exclusively revisionist” view or clear bounds on how revisionist it will be.

Finally, Gill claims that our policy recommendations would be “ineffective or counterproductive” if any mechanisms aside from a security dilemma are at play. Disappointingly, he offers little support for this assessment and suggests no alternative means for mitigating the military competition he agrees exists. We want to know more about what he thinks policymakers should do. Regardless, we disagree with his criticism. None of the measures we proposed would be taken in isolation. Rather, they would be part of a comprehensive strategy centered on a nuanced combination of cooperation (win-win and positive inducements) and deterrence aimed at shaping China's rise in a peaceful and order-sustaining direction. As we wrote, none of these measures would be panaceas, but they are worth considering if they can ameliorate destabilizing competition by reducing uncertainty and the risks of miscalculation or unintended escalation (pp. 54, 61, 90–91).

### IN THEORY, PRACTICE IS SIMPLE (RESPONSE TO ERIC HUNDMAN)

Whereas Fu and Gill focus on our empirical analysis, Eric Hundman engages our analysis of security dilemmas and arms races in the Asia Pacific primarily in abstract theoretical terms. Although he offers a vigorous critique of our argument, we see evidence of significant common ground. Most important, Hundman agrees with us on at least two fundamental points. First, he concurs that “identifying security dilemma dynamics” in the region is of “enormous importance.” Second, he contends that security dilemmas are an important driver of regional military competition, conceding in his conclusion that “[s]ecurity dilemma dynamics are undoubtedly creating additional challenges in the Asia Pacific.”

Despite these fundamental agreements, Hundman questions the theoretical model that we developed to identify and empirically test claims about the presence of security dilemma dynamics. What Hundman offers is a useful reflection on the rich and complex literature on the security dilemma. But in critiquing our article, he fails to engage it on its own terms. Most important, he does not show how his preferred theoretical formulations would translate into tractable, much less superior, models for empirical testing. Many of his insights are thoughtful, but nowhere does he show that the theoretical models he favors would in practice better identify, illuminate, and evaluate the regional security dilemma dynamics that he concedes “undoubtedly” exist. Nor does he offer new empirical data that might demonstrate the merits of his preferred approach.

In our article, we drew on the seminal works on security dilemma theory that Hundman cites, but we modified and adapted their models for the purpose of empirical testing. (The editor of International Security put it well in his original correspondence requesting revisions: “The literature of the security dilemma is often confusing…. Develop and apply your own model of the security dilemma.”) We have little doubt our approach is imperfect. Indeed, we warned the reader that “the transition from the realm of abstract theory to the empirical world is often treacherous” and that our findings should be considered “preliminary” (pp. 63, 86).

Hundman's letter draws on important works by Robert Jervis and Charles Glaser and makes reasonable points relevant to conceptual distinctions and debates about security dilemma theory—in particular, the importance of differentiating between uncertainty and misperception as a source of misinterpretation of others’ intentions. He also references important debates about distinguishing offensive and defensive forces. We accept several of his points as valuable for theoretical debate. But here we feel compelled to reiterate that our article's objective is to empirically assess the validity of widespread, often uncritical claims about security dilemmas and arms races occurring across the Asia Pacific—not to engage in a debate about the extent to which our conceptualization of the security dilemma remains faithful to the theory as articulated by others. We simplified our model so that we could develop an analytically tractable empirical test.

Hundman concedes that we recognize misperceptions as important drivers of security dilemmas, but he criticizes our model for “downplaying” them and for not clearly distinguishing between uncertainty and misperception. He also claims that we conflate status quo inclinations and security-seeking intentions, and he criticizes our decision to focus on differentiating between mutual arming dynamics in which states’ interests are aligned but leaders are mutually suspicious largely as a result of uncertainty about intentions and those in which a more basic clash of interests exists. In the very early stages of our project, we tried unsuccessfully to proceed along the lines Hundman implies we should have. Distinguishing these concepts outside the realm of pure theory presents a number of challenges Hundman appears to overlook. We invite him to develop and apply a security dilemma model capable of uncovering evidence of the allegedly “clear” differentiation that he suggests our empirical test lacks.

Further suggestive of the difficulty Hundman may have moving from the realm of theory to the complexity of the empirical world are several simplistic all-or-nothing claims. One is his puzzling statement that based on his interpretation of our argument, the presence of a conflict of interest need be mutually exclusive of security dilemma dynamics. His claim seems based on our brief description of a purely hypothetical Type-2 strategic setting. We recognized explicitly, however, that these categories are notional and that both Type-1 and Type-2 logics can manifest simultaneously in a dyad—a fact reflected in our empirical analysis and discussion (e.g., Japan-China). Simply because an existing interest-based dispute may be the primary driver of action-reaction arming does not necessarily mean that other factors (e.g., a security dilemma) are not also at play, and vice versa. Despite Hundman's assertions, we made no claim to the contrary. Teasing out the causal drivers of mutual arming, as well as their relative weights, is an empirical task—and a difficult one.

In his discussion of distinguishing offensive and defensive weapons, Hundman misattributes to us our summary of Jervis's seminal argument. We did not go into an extensive discussion of the offense-defense variable because disparate levels of military development and the changing nature of military technologies (e.g., long-range precision strike, ballistic missile defense technologies [which can double as antisatellite weapons], cyber, etc.) seem to render the distinction less applicable in shaping contemporary leaders’ perceptions. Because we judged there is no one-size-fits-all explanation for the entire region, the first question in our four-question framework—“what is the offense-defense balance in the Asia Pacific?”—while motivated by Jervis's theoretical insights, proved impracticable to address empirically. Regarding the second question—“are allegedly defensive measures/weapons distinguishable from offensive ones?”—our analysis raised doubts about whether (1) a given technology's offensive or defensive nature is objectively definable based on characteristics of the technology itself, and (2) more importantly for our purposes, whether leaders react based on such objective assessments. Because our approach required us to evaluate perceptions and behaviors from the perspective of policymakers, this was a crucial finding. No clear evidence demonstrated either side systematically distinguishing between offensive and defensive measures based on the actual characteristics of the technologies themselves or on some elusive yet prevailing offense-defense balance in the system. Rather, leaders frequently interpret the procurement of new weapons systems through a sharply political lens, one often tinted by mistrust and suspicion of others’ motives. Perceptions and politics thus often appear to trump any objectively definable reality of such a balance.

Although we did not include an in-depth analysis of these issues in our article, our empirical analysis reflects our awareness of them. For example, capabilities procured or policies adopted by one side for defensive reasons are often interpreted by the other side as offensive and threatening—be it measures the United States and its allies adopt for the stated goal of strengthening deterrence against perceived attempts to change the status quo, or China's allegedly defensive aircraft carrier program and conventionally tipped ballistic missile arsenal.

We disagree with Hundman's claim that the offense and the defense must be distinguishable for transparency to offer leaders any insight into another state's motives. Again, his point seems based more on a theoretical assumption than on how policymakers behave in the dyads we examine. To claim that unless one can judge whether weapons are inherently offensive or defensive military transparency is useless for judging others’ intentions strikes us as not empirically justifiable. Indeed, transparency can be a key factor in shaping leaders’ threat perceptions, often trumping judgments about actual capabilities. For example, among states with mature democratic systems transparency significantly ameliorates what might otherwise be major distrust and uncertainty about other states’ intentions. Policy debates and budgets are relatively open and transparent; policymaking processes are significantly less opaque; and external actors can more easily assess—or even shape—policy outcomes. For example, holding capabilities constant, if U.S. military decisionmaking, capabilities, and expenditures were as opaque as Beijing's, even countries that currently worry very little about their security vis-à-vis the United States would be far more suspicious of its intentions. We therefore disagree with Hundman's desire to write transparency off categorically, in both general terms and specifically with regard to our cases. In the Asia Pacific, transparency—or lack thereof—manifests empirically as a powerful factor shaping contemporary policy debates about China in Washington, Tokyo, Canberra, and beyond.

We are further puzzled by Hundman's inaccurate claim that our article “implies that any security dilemma dynamics in the Asia Pacific are solely China's fault.” He then seems to correct our alleged claim by pointing out that the security dilemma is a “fundamentally dyadic phenomenon.” We repeatedly noted, however, that security dilemmas are dyadic and acknowledged insecurity as a key driver of China's calculus. Space constraints prevented an extensive assessment of Beijing's own threat perceptions, something that one of us has offered elsewhere using Chinese-language sources.5 Furthermore, we noted several instances where others—including the United States— could do more to reassure China about their intentions. That said, China's rapid development and military modernization are the primary drivers of changing power balances in the Asia Pacific. Its capabilities and interests are expanding, and Beijing is increasingly well situated to attempt to achieve objectives coercively or even though military force. This ability is likely to increase over time, and regional states are clearly concerned about China's intentions.

Hundman's related request for us to call on the United States to reassure China by increasing its own military transparency implies a false equivalence of responsibility for political mistrust and uncertainty that we do not accept on empirical grounds. The United States, Japan, Australia, and several other regional actors, though imperfect, are significantly more transparent than China regarding their decisionmaking and military capabilities, budgets, and intentions. Furthermore, throughout the post-Cold War period Northeast Asia's two other major players—the United States and Japan—have a history of engaging China and have shown restraint in not attempting to change the status quo through coercion or force, despite having enjoyed enormous capabilities advantages over China's People's Liberation Army (PLA) for decades. In contrast, and despite regional leaders’ repeatedly stated concerns, Beijing's military transparency remains relatively low. Top PLA experts identify China's military as “the blackest of the black boxes,” while influential China scholars and former high-level U.S. government officials bemoan how much remains unknown about Beijing's basic policy decisionmaking.6 China's budget transparency remains significantly lower than that of the United States and its major Asia Pacific allies, despite some recent progress.7 We readily acknowledge that Chinese leaders have sincere security concerns, but official and unofficial policy discourse suggests strongly that Beijing's opacity disproportionately exacerbates political mistrust and uncertainty.

Finally, we are further perplexed by Hundman's criticisms of our case selection as “undermining [our] analysis.” One should evaluate case selection mechanisms based on a study's research objective. The entire premise of our project was to empirically assess the validity of widespread, often uncritical claims that the Asia Pacific is rife with security dilemmas and arms races. By its very nature our project thus required us to, in Hundman's words, “foreground[] adversarial dynamics” in our empirical survey. We selected the cases most frequently eliciting the claims that motivated our study. We recognize and appreciate numerous positive developments taking place in the region. And we stated that the situation in the Asia Pacific is far more stable than it was in Europe before World War I or during the Cold War—both flawed analogies employed much too frequently in scholarly and public discourse on China's rise. We also concluded that despite ongoing regional competition, a traditional arms race is not occurring and that conflict is by no means inevitable. We mentioned several cooperative U.S.-China exchanges, called for more, and offered five proposals for ameliorating frictions. Given this, Hundman's claim that our article engages in “threat inflation” and “fuel[s] the intensification of security dilemma dynamics in the Asia Pacific” is hyperbolic, to say the least. It also appears empirically groundless, as his largely theoretical letter contains no clear evidence to support his claim that we exaggerated (or, conversely, undersold) anything, much less for holding us responsible for fueling regional military competition. Meanwhile, Hundman's assertion that our selection of certain dyads biases our conclusions makes sense only if we had put forth a one-size-fits-all argument (we did not), or if our stated research question was “what explains variation in the degree of balancing against China among Asia Pacific states?” (it was not). In the interest of case diversity, we included two U.S. security allies cum democracies and two nondemocratic non-allies—including a historical, communist-led adversary (Vietnam). Given the objectives of our article, what alternative dyads could be of greater practical, or even theoretical, importance than those we chose?

Because Hundman shares our view that “identifying security dilemma dynamics” in the Asia Pacific is of “enormous importance,” we are disappointed by the generally negative approach of his letter, in both substance and tone. Particularly frustrating is a core contradiction that seems to undermine his critique: Hundman criticizes our analysis throughout his letter, yet concedes in a single sentence at the end that we are basically correct in arguing that “[s]ecurity dilemma dynamics are undoubtedly creating additional challenges in the Asia Pacific.” This contradiction is a bit jarring given that his letter provides no clear empirical basis for reaching that conclusion. It is also ironic, given our stated motivation for writing our article in the first place.

### CONCLUSION

Our article was motivated by frustration with the tendency of many observers of the contemporary Asia Pacific to play fast and loose with the terms “security dilemma” and “arms races” as descriptors for the sources and consequences of regional political frictions and military competition. In our view, these terms are loaded with theoretical and real-world significance, yet few scholars have developed, much less systematically applied, an analytical framework to evaluate their appropriateness. Our article constitutes a preliminary attempt to do so.

The security dilemma is a fundamental concept in international relations theory. Yet the theoretical and practical shortcomings in its associated literature, and the disconnection between them, have yet to be resolved. We offered a conceptualization of the security dilemma that we considered empirically operationalizable. As we noted at the outset, moving from abstract theory to applied empirical tests presents numerous challenges. It also requires a willingness to recognize the complexities and nuances of how policymakers perceive, and operate in, the real world. We hope that our article and this correspondence help to bridge the gap and move the scholarly debate forward, while shedding light on the challenges associated with such an enterprise. We thank the editors and our critics for the opportunity to contribute to this debate.

Bloomington, Indiana

G. John Ikenberry

Princeton, New Jersey

## Notes

1.

Adam P. Liff and G. John Ikenberry, “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” International Security, Vol. 39, No. 2 (Fall 2014), pp. 52–91, at pp. 54–55. Further references to this article appear parenthetically in the text.

2.

Military expenditure remains the most widely used proxy for measuring the intensity of military competition and the existence of arms races. See Paul F. Diehl, “Arms Races and Escalation: A Closer Look,” Journal of Peace Research, Vol. 20, No. 3 (September 1983), pp. 205–212; and Toby J. Rider, Michael G. Findley, and Paul F. Diehl, “Just Part of the Game? Arms Races, Rivalry, and War,” Journal of Peace Research, Vol. 48, No. 1 (January 2011), pp. 85–100.

3.

Author's calculations using data from Stockholm International Peace Research Institute (SIPRI), “Worldwide Military Expenditures, 2013” (Stockholm: SIPRI, 2013), http://www.sipri.org/research/armaments/milex/milex_database.

4.

Military expenditure as a percentage of GDP is often used as a proxy for measuring a state's defense effort. See Barry Buzan and Eric Herring, The Arms Dynamic in World Politics (Boulder, Colo.: Lynne Rienner, 1998).

5.

Military expenditure as a percentage of government spending provides another way to assess the defense effort of a country. As noted above, however, the shared percentage of government spending that defense represents has also significantly declined in most Asian countries over the past quarter century.

6.

Calculated from SIPRI, “Worldwide Military Expenditures, 2013.”

7.

Calculated from ibid.

8.

Steve Chan demonstrates that most Asian political leaders do not regard military balancing against foreign enemies as a viable way to enhance their political legitimacy. See Chan, Enduring Rivalries in the Asia-Pacific (Cambridge: Cambridge University Press, 2013).

9.

Paul Kallender-Umezu, “Japan Defense Budget Rises 0.8%,” Defense News, January 14, 2015, http://www.defensenews.com/story/defense/policy-budget/budget/2015/01/14/japan-budget-increase-abe-nakatani-china-nansei/21770531/.

10.

“55% Do Not Support Abe's Security Policies: Kyodo Poll,” Kyodo News, December 17, 2014, http://www.japantimes.co.jp/news/2014/12/17/national/politics-diplomacy/55-do-not-supportabes-security-policies-poll-shows/#.VTCVHiHBzRY.

11.

“Asahi Poll: 63% Oppose Abe's Attempt to Lift Ban on Collective Self-Defense,” Asahi Shimbun, April 7, 2014, http://ajw.asahi.com/article/behind_news/politics/AJ201404070067.

12.

Truong-Minh Vu and Nguyen Thanh Trung, “A U.S.-Vietnam Alliance or (Still) a U.S.-China-Vietnam Triangle?” International Policy Digest, October 3, 2014, http://www.internationalpolicydigest.org/2014/10/03/u-s-vietnam-alliance-or-u-s-china-vietnam-triangle/.

13.

Jon Grevatt, “Vietnam's Military Spending to Rise by 70% in 2011,” Jane's Defence Weekly, January 17, 2011; and Jon Grevatt, “Vietnam Announces 35 Percent Defence Spending Rise,” Jane's Defence Weekly, November 16, 2011.

14.

Calculated from SIPRI, “Worldwide Military Expenditures, 2013.”

15.

Chin-Hao Huang and Patrick James, “Blue, Green, or Aquamarine? Taiwan and the Status Quo Preference in Cross-Strait Relations,” China Quarterly, Vol. 219 (September 2014), pp. 670–692.

16.

Stephen G. Brooks, G. John Ikenberry, and William C. Wohlforth, “Don't Come Home, America: The Case against Retrenchment,” International Security, Vol. 37, No. 3 (Winter 2012/13), pp. 7–51.

17.

Evelyn Goh, The Struggle for Order: Hegemony, Hierarchy, and Transition in Post-Cold War East Asia (Oxford: Oxford University Press, 2013).

18.

Michael D. Swaine and M. Taylor Fravel, “China's Assertive Behavior—Part Two: The Maritime Periphery,” China Leadership Monitor, June 24, 2011, pp. 1–29; and M. Taylor Fravel and Alastair Iain Johnston, “Chinese Signaling in the East China Sea?” Washington Post, April 12, 2014, http://www.washingtonpost.com/blogs/monkey-cage/wp/2014/04/12/chinese-signaling-in-the-east-china-sea/.

19.

M. Taylor Fravel, “International Relations Theory and China's Rise: Assessing China's Potential for Territorial Expansion,” International Studies Review, Vol. 12, No. 4 (December 2010), pp. 505–532.

1.

Adam P. Liff and G. John Ikenberry, “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” International Security, Vol. 39, No. 2 (Fall 2014), pp. 52–91. Subsequent references to this article appear parenthetically in the text.

2.

On the complexities of analyzing states with multiple adversaries and distributed interests, see Keir A. Lieber, “Mission Impossible: Measuring the Offense-Defense Balance with Military Net Assessment,” Security Studies, Vol. 20, No. 3 (2011), pp. 458–459.

3.

Concerning the possibility of greedy and insecure states, see Charles L. Glaser, “Political Consequences of Military Strategy: Expanding and Refining the Spiral and Deterrence Models,” World Politics, Vol. 44, No. 4 (July 1992), pp. 497–538.

4.

Walter Russell Mead, “The Return of Geopolitics: The Revenge of the Revisionist Powers,” Foreign Affairs, Vol. 93, No. 3 (May/June 2014), http://www.foreignaffairs.com/articles/141211/walter-russell-mead/the-return-of-geopolitics. See also Andrew F. Krepinevich Jr., “How to Deter China: The Case for Archipelagic Defense,” Foreign Affairs, Vol. 94, No. 2 (March/April 2015), http://www.foreignaffairs.com/articles/143031/andrew-f-krepinevich-jr/how-to-deter-china.

5.

G. John Ikenberry, “The Illusion of Geopolitics: The Enduring Power of the Liberal Order,” Foreign Affairs, Vol. 93, No. 3 (May/June 2014), http://www.foreignaffairs.com/articles/141212/g-john-ikenberry/the-illusion-of-geopolitics.

6.

Adam P. Liff and Andrew S. Erickson, “Demystifying China's Defence Spending: Less Mysterious in the Aggregate,” China Quarterly, Vol. 216 (December 2013), pp. 824, 827.

7.

Ikenberry, “The Illusion of Geopolitics”; and Andrew S. Erickson and Adam P. Liff, “Not-So-Empty Talk: The Danger of China's ‘New Type of Great-Power Relations’ Slogan,” Foreign Affairs, October 9, 2014, http://www.foreignaffairs.com/articles/142178/andrew-s-erickson-and-adam-p-liff/not-so-empty-talk.

8.

M. Taylor Fravel, “Regime Insecurity and International Cooperation: Explaining China's Compromises in Territorial Disputes,” International Security, Vol. 30, No. 2 (Fall 2005), p. 83.

9.

Andrew Kydd, “Sheep in Sheep's Clothing: Why Security Seekers Do Not Fight Each Other,” Security Studies, Vol. 7, No. 1 (Autumn 1997), pp. 114–155.

10.

Glaser, “Political Consequences of Military Strategy,” pp. 497–538; Thomas J. Christensen, “The Contemporary Security Dilemma: Deterring a Taiwan Conflict,” Washington Quarterly, Vol. 25, No. 4 (Autumn 2002), pp. 7–21; and John J. Mearsheimer, “Reckless States and Realism,” International Relations, Vol. 23, No. 2 (June 2009), pp. 252–253. For a challenge to claims that security dilemmas can exist if one of the states involved is a revisionist power, see Shiping Tang, “The Security Dilemma: A Conceptual Analysis,” Security Studies, Vol. 18, No. 3 (July 2009), pp. 594, 607–608.

11.

Keir A. Lieber, “The New History of World War I and What It Means for International Relations Theory,” International Security, Vol. 32, No. 2 (Fall 2007), pp. 155–191; and Robert Jervis, “Was the Cold War a Security Dilemma?” Journal of Cold War Studies, Vol. 3, No. 1 (Winter 2001), pp. 36–60.

12.

See, for instance, Robert Jervis, “Cooperation under the Security Dilemma,” World Politics, Vol. 30, No. 2 (January 1978), pp. 170, 185; Evan Braden Montgomery, “Breaking Out of the Security Dilemma: Realism, Reassurance, and the Problem of Uncertainty,” International Security, Vol. 31, No. 2 (Fall 2006), pp. 168–173; and Joseph M. Parent and Joshua M. Baron, “Elder Abuse: How the Moderns Mistreat Classical Realism,” International Studies Review, Vol. 13, No. 2 (June 2011), p. 207.

13.

Charles L. Glaser, “When Are Arms Races Dangerous? Rational versus Suboptimal Arming,” International Security, Vol. 28, No. 4 (Spring 2004), pp. 61–64.

14.

Liff and Ikenberry also observe that “investments [in military power], even if not initially motivated by external threats, can create or exacerbate a security dilemma” (p. 60).

15.

Proponents of security dilemma theory have tended to focus more on Anglo-German antagonism than on Britain's relationships with other European states in the years preceding World War I. Revisionist historical accounts, however, also highlight British concerns with other major powers. Concerning this point and the revisionist versus post-revisionist debate, see Jon Tetsuro Sumida, In Defence of Naval Supremacy: Finance, Technology, and British Naval Policy, 1889–1914 (1993; repr. London: Routledge, 2014); and Matthew S. Seligmann, The Royal Navy and the German Threat 1901–1914: Admiralty Plans to Protect British Trade in a War against Germany (Oxford: Oxford University Press, 2012). Liff and Ikenberry examine multiple states in the Asia Pacific but do not focus on the influence of some other major powers, such as India. See Evan Braden Montgomery, “Competitive Strategies against Continental Powers: The Geopolitics of Sino-Indian-American Relations,” Journal of Strategic Studies, Vol. 36, No. 1 (February 2013), pp. 76–100; and Jarrod Hayes, Constructing National Security: U.S. Relations with China and India (Cambridge: Cambridge University Press, 2013).

16.

Nicholas A. Lambert, Sir John Fisher's Naval Revolution (Columbia: University of South Carolina Press, 2002); and Sumida, In Defence of Naval Supremacy. Political and economic objectives would continue to influence Britain's strategic policies in the decades after World War I. See David James Gill, Britain and the Bomb: Nuclear Diplomacy, 1964–1970 (Stanford, Calif.: Stanford University Press, 2014).

1.

Adam P. Liff and G. John Ikenberry, “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” International Security, Vol. 39, No. 2 (Fall 2014), pp. 52–91. Further references to this article appear parenthetically in the text.

2.

For his classic elaboration of the security dilemma, see Robert Jervis, “Cooperation under the Security Dilemma,” World Politics, Vol. 30, No. 2 (January 1978), p. 169.

3.

Charles L. Glaser, “The Security Dilemma Revisited,” World Politics, Vol. 50, No. 1 (October 1997), p. 175.

4.

Jervis, “Cooperation under the Security Dilemma,” pp. 174–175.

5.

Ibid., p. 191.

6.

Glaser, “The Security Dilemma Revisited,” p. 200.

7.

Alastair Iain Johnston, “Stability and Instability in Sino-U.S. Relations: A Response to Yan Xuetong's Superficial Friendship Theory,” Chinese Journal of International Politics, Vol. 4, No. 1 (March 2011), pp. 18–19; and Shiping Tang, “The Security Dilemma: A Conceptual Analysis,” Security Studies, Vol. 18, No. 3 (September 2009), pp. 587–623.

8.

Glaser, “The Security Dilemma Revisited,” p. 177; and Jervis, “Cooperation under the Security Dilemma,” p. 187.

9.

The contested nature of the status quo in Asia today is particularly evident in recent disputes over sovereignty in the South China Sea. See, for example, Bill Hayton, The South China Sea: The Struggle for Power in Asia (New Haven, Conn.: Yale University Press, 2014).

10.

Jervis, “Cooperation under the Security Dilemma,” p. 188.

11.

Glaser, “The Security Dilemma Revisited,” p. 174.

12.

Ibid., p. 186. Distinguishing between offense and defense may be much more difficult in a maritime context. See Charles L. Glaser, “How Oil Influences U.S. National Security,” International Security, Vol. 38, No. 2 (Fall 2013), pp. 112–146.

13.

Glaser, “The Security Dilemma Revisited,” p. 176.

14.

Ibid., p. 174.

15.

The author is grateful to Charles Glaser for raising this point.

16.

Military exchanges between the United States and China are a particularly good example of ongoing cooperation in this arena. See Shirley A. Kan, “U.S.-China Military Contacts: Issues for Congress” (Washington, D.C.: Congressional Research Service, July 29, 2014).

17.

Malaysia only recently began attempting to substantially upgrade its forces, and it remains unclear whether its leadership will be able to complete this project, given domestic political restraints. See Andrew Chuter, “Cash-Strapped Malaysia Looks to Lease Fighters,” Defense News, February 15, 2015; and Trefor Moss, “Malaysia's Military Set to Go Shopping,” Wall Street Journal, April 16, 2014. Although the Philippines is increasing its defense spending, its budget is still “modest by regional standards.” See Trefor Moss, “Philippines Military Goes on Shopping Spree,” Wall Street Journal, July 17, 2014. Taiwan's defense spending has “languished,” and it continues to decrease the size of its armed forces. See Bonnie Glaser and Anastasia Mark, “Taiwan's Defense Spending: The Security Consequences of Choosing Butter over Guns” (Washington, D.C.: Center for Strategic and International Studies, March 18, 2015). On downsizing, see, for example, Jason Pan, “Legislators Slam Military Downsizing,” Taipei Times, March 16, 2015.

18.

Johnston, “Stability and Instability in Sino-U.S. Relations.”

1.

Adam P. Liff and G. John Ikenberry, “Racing toward Tragedy? China's Rise, Military Competition in the Asia Pacific, and the Security Dilemma,” International Security, Vol. 39, No. 2 (Fall 2014), pp. 52–91. Further references to this article appear parenthetically in the text.

2.

See also G. John Ikenberry, “Between the Eagle and the Dragon: America, China, and Middle State Strategies in East Asia,” Political Science Quarterly, forthcoming.

3.

Original approach and metrics are drawn from Adam P. Liff, “Whither the Balancers? Reconsidering Methods and Metrics in Contemporary Security Studies and Military Responses to China's Rise,” Security Studies, forthcoming.

4.

Ibid., especially pp. 15–17.

5.

Adam P. Liff, “From ‘Bottle Cork’ to ‘Achilles’ Heel'? China's Deepening Cynicism about the U.S.-Led Alliance System's Role in Asia-Pacific Security,” working paper, Indiana University, 2015.

6.

Susan Shirk, “What China's Lack of Transparency Means for U.S. Policy,” ChinaFile, May 28, 2015, http://www.chinafile.com/reporting-opinion/two-way-street/what-chinas-lack-transparency-means-us-policy.

7.

Adam P. Liff and Andrew S. Erickson, “Demystifying China's Defence Spending: Less Mysterious in the Aggregate,” China Quarterly, December 2013, pp. 805–830.