Unipolar Politics: Realism and State Strategies After the Cold War, by Ethan B. Kapstein and Michael Mastanduno (eds.)

 

8. Realism(s) and Chinese Security Policy in the Post-Cold War Period

Alastair Iain Johnston

Why try once again to test realist theory, this time on the behavior of states in the post-Cold War period, when it has, apparently, been so bad at prediction and explanation in the past? ₁ One justification is that even if realist hypotheses do not do particularly well in explaining a range of cases, there is still some portion of these where these hypotheses do work. China would seem to be a good candidate. A a number of observers of Chinese foreign policy have remarked on the realpolitik nature of Chinese behavior, ₂ but there is still some controversy over how realpolitik, and over the roots of this realpolitik. Adding or subtracting the China case from the number of cases that realists can claim to explain well is important for advancing IR theory’s understanding of major-power behavior, which heretofore has rested largely on the European or American cases. Indeed, the question of Chinese balancing during the post-Cold War period ought to be as “most likely” a case for realism as any. With a rapid and massive shift in polarity after the collapse of the USSR, realism should be optimistic about confirming predictions about balancing from a major state with no prior alliance ties to the U.S. and a skeptical approach to international institutions such as China.

Some realists might object to this single-country empirical focus. Those who purport to rest their explanations entirely on system structure would argue this is an unfair test of their theory. They argue that their models work as if unit-level characteristics and variables are essentially irrelevant. They go on to suggest that even if these unit-level characteristics are different and have some effect on the behavior of particular states, in the aggregate these deviations cancel each other out. But when looking at specific cases, the structuralists would argue, one should expect some role for unit-level characteristics, since one is not looking at an aggregation of actors. So, they would conclude, don’t be surprised if their predictions don’t work on individual states such as China.

One could counter, however, that China is a major power, and it is major powers that should be more acutely sensitive to structural constraints (since they, not small states, are likely to be the target of rising hegemons). Thus major powers are more likely to such reflect the modal characteristics of the aggregation of the actors. This ought to make them amenable to tests of realist theory. ₃ I will take up the question of whether structural realism embodies expectations about the foreign policy behavior of individual states in a moment. But the point here is that while one single-country test of realist propositions is not sufficient to confirm or undermine realist claims, realists ought to be concerned about the cumulative implications of many single-country major-power tests. If the number of major-power cases in which behavior uniformly deviates from the direction expected by realism grows too large it is doubtful that realists’ claims about systemic outcomes will be right.

One of the problems with testing realist theory, of course, is deciding what it is. We often find ourselves bracketing or qualifying realism with adjectives such as “standard,” “mainstream,” “old fashioned,” “baseline,” “traditional,” as though it were obvious what was meant by these descriptors. More often, it seems, these adjectives are used to avoid acknowledging the problem of realism’s indeterminateness, particularly at the actor level, and to infuse materialist realist theory with far more coherence than really exists. As Wayman and Diehl argue, realism is not so much a theory as it is a “cluster of models, assumptions, hunches, hypotheses, and parameter estimates held together by their common focus on concepts including and related to national material capabilities, power, perceived power, major-power status, revisionist and status quo powers, coalition formation via the balance of power, resolve and commitment.” ₄

Instead of trying to distill an intersubjectively acceptable version of realist theory, I will test four versions of realism. Three of these would be considered the predominant strains of contemporary realism and are readily identified with prominent scholars who claim to be working in the realist tradition. Two versions of realism fall within the neorealist camp—that is, they derive their predictions from system-level variables: balance of power realism, identified with Waltz; and power maximization realism as exemplified by Mearsheimer. The other two versions fall outside of neorealism because they rely heavily on the explanatory role of nonsystemic variables: balance-of-threat realism, as explained by Walt; and what could be called identity realism, which draws on work in social-psychology and constructivist IR theory that suggests an ideational basis for realpolitik behavior.

The structure of the essay is essentially a simple interrupted time-series design, where the X treatment is “the end of the cold war” as defined by a change in the value of the key independent variable for each theory. One then observes trends in the dependent variable before and after the X treatment to see if there is a significant variation that is consistent with the expectations of the theory. ₅ To telegraph my argument, I find that Waltz’s balance of power and Walt’s balance-of-threat theories are not useful explanations of Chinese security behavior. Chinese security policy is not reacting to the end of the Cold War in ways these theories predict. Power maximization realism provides a relatively accurate description of Chinese security behavior. But it gets the explanation wrong, or at least it is incomplete. China’s power maximizing behavior in the post-Cold war period is not rooted in insecurity generated by structural changes in the international system. The drive to increase relative capabilities predates the shift of global military capabilities toward the U.S. direction. Moreover, although this drive persists under near-unipolarity it is not because the Chinese believe this structural condition is especially dangerous for China at the moment—indeed Chinese assessments describe the post-Cold War period as the most benign period in Chinese strategic security since at least 1949.

I argue here that identity realism provides a more nuanced understanding of the origins, timing, and content of Chinese security policy spanning the Cold War-Post Cold War transition. It does so by helping us understand two critical features of Chinese foreign policy in this transition: The first is why Chinese leaders believe that this relatively peaceful era provides a window of opportunity to increase relative capabilities relatively rapidly so as to enhance China’s international status, rather than security. The second is why Chinese leaders believe that the primary dangers in their environment come from threats to the achievement of higher status emanating from challenges to the cohesiveness of internal order. One consequence is that the Chinese leadership has tried to intensify popular identification with the regime and the nation-state, with the result that we have seen a somewhat starker “othering” of the external environment and a consequent hardening of realpolitik discourse and behavior.

 

Realisms and Their Predictions

Balance of Power Realism

Waltz’s balance of power realism makes a simple prediction about a major power’s behavior: it ought to balance against whichever state appears to be trying to establish a dominant or hegemonic position in the system. If ever one ought to see such balancing, it ought to be when there is such a dramatic change in the systemic distribution of power that it leaves another power in a commanding position. The end of the Cold War—essentially the disappearance of one of the poles in a bipolar system—was precisely such a moment of dramatic redistribution. ₆

Thus, if the end of the Cold War constitutes a change in an independent variable of some type, for balance of power theory it must signify a change in polarity if it is to have any observable implications for variation in the foreign policies of major powers. This raises the question of how to determine whether there has been a change in polarity. Waltz notes simply that power is “difficult to measure and compare”: How it is measured and distributed are “empirical” questions and “common sense can answer” them. ₇

One place to start is simply to use relative shares of world military expenditures as a measure of power. ₈ The key question, obviously, is what particular distribution of this measure would constitute different types of polarity? Neorealists provide no insights. For simplicity’s sake I use Modelski’s commonly used categories: a state that possesses 50 percent or more of system-wide military expenditures is the unipole; two states that, combined, possess 50 percent or more of system capabilities, with each having 25 percent or more, are the two bipoles; and any other distribution is multipolar. ₉

The question for balance of power theory, then, is: with the collapse of the Soviet Union did the distribution of system capabilities change enough to constitute a change in polarity? Depending on the source for military expenditures (ME) one uses, one arrives at somewhat different estimates. According to Arms Control and Disarmament Agency (ACDA) data in 1987 the U.S. and Russian shares of world expenditure were about the same at around 30 percent. This distribution met Modelski’s definition of bipolarity. By 1993 the U.S. share increased somewhat to 34 percent of world military expenditures, but the Russian share dropped dramatically from about a third to less than a sixth (see figure 8.1). These shares do not meet Modelski’s criteria for unipolarity, but neither does this fit his definition of multipolarity: Rather, this distribution slips between the definitional cracks because the U.S. has such a commanding lead in military capabilities over the next contender (Russia). According to IISS military expenditure figures, however, in 1992 the ME shares stood at 43.5%, 7.1% and 4.0% for the U.S., Russia and China respectively. ₁₀ This distribution doesn’t quite fit Modelski’s categorization of unipolarity either, since the U.S. has less than 50 percent and three or more states have more than 5 percent each, but neither is it multipolarity, given the clear lopsidedness. The U.S. would need only a few percentage points more for the system to be unipolar. Thompson recognizes this and calls this kind of distribution asymmetrical multipolarity or near unipolarity . ₁₁ Regardless of which data set one uses, however, both show a very clear redistribution of capabilities beginning in 1991 with the collapse of the USSR. In military power terms, the U.S. is the dominant actor. ₁₂ It is obvious that this differs from the distribution of power for most of the Cold War. Thus it is plausible to assume for the purposes of testing balance of power theory that the end of the Cold War (set at 1991), constituted a shift in polarity from a bipolar to a near-unipolar or highly asymmetrical multipolar system, where the U.S. is the one candidate hegemon.

Figure 8.1: U.S., Russian and Chinese shares of world military expenditure Source: ACDA 1995 World Military Expenditures

Balance of power theory therefore makes the following general prediction. H.A: In order to maximize security in an anarchical competitive environment where the use of force is the ultima ratio of state behavior, states balance against a rising state (or coalition of states) that is, or threaten(s) to emerge as, the dominant actor(s) in the system. China, therefore, ought to be balancing against the United States. More specifically: H.A1: We should expect that military expenditures and trends in acquisition and doctrine reflect the particular needs of militarily balancing against the candidate hegemon. This first hypothesis predicts that China ought to be increasing its military expenditures and adjusting its force posture to deal with the United States as the emergent post-Cold War hegemon. As figure 8.2 suggests, however, if China is balancing internally against the United States its military spending, adjusted for inflation, does not reveal much single-minded effort to concentrate resources on military capabilities after 1991. Or at least the statistics are inconsistent. The Pentagon’s East Asia Strategy Report of February 1995 provides one of the more alarmist estimates of real growth in Chinese military expenditures (ME)—40 percent from 1989 through 1994, or about 8 percent per year. ₁₃ The United States General Accounting Office, however, estimates only a 6.4% real growth between 1987 and 1993. ₁₄ ACDA figures for constant ME also suggest relatively slow but steady growth after 1989.

Figure 8.2: Chinese military expenditures 1981–1994 Sources: SIPRI

Military expenditures as a percentage of GDP and central government expenditures also do not indicate a “storming” approach to military power. That is, the economy is apparently not becoming increasingly militarized to deal with the large structural shift in power after 1991. Whether counting nominal or real military expenditures, these percentages have dropped over the 1980s and 1990s, showing no obvious reaction to the emergence of the United States as the dominant hegemon or as the one remaining superpower (see figure 8.3).

Figure 8.3: Chinese military expenditures as a percentage of GDP and Central Government Expenditures Source: SIPRI

Of course, one could argue that since the dominant hegemon’s own military expenditures have declined in real terms in the 1990s, the net relative change is quite significant. In the post Cold War period the United States share of world military expenditures increased, but the ratio of its spending to China’s decreased from about 6.6: 1 in 1988 to 5.3:1 in 1993 (using ACDA figures). The ratio of the Soviet/Russian spending to China’s dropped from 7.2:1 in 1988 to about 2:1 in 1993 (see figure 8.4). Thus China narrowed the power gap with both states in the post Cold War period. Still, this change is not so much a function of deliberate Chinese balancing against the U.S. as it is the unpredicted collapse of a major military spender, the Soviet Union.

Figure 8.4: Military Expenditure Ratios Source: ACDA

Internal balancing is not limited to overall increases in military expenditures. A balancing state could also be expected to develop operational capabilities designed to deal with potential military threats from the system hegemon. Chinese military acquisition patterns after the collapse of the USSR do not indicate a particularly alarmed balancing directed at the United States Rather the PLA is modernizing to deal with a range of contingencies from border conflicts with India, to naval conflicts over the South China Sea, to operations against Taiwan. There is some discussion in military writings on modern high-tech war to suggest that the U.S., among other high-tech militaries, is the standard that China should aim for. ₁₅ But these trends in doctrinal thinking were rooted in the “strategic decision” of 1985 to shift China’s force posture away from dealing with a Soviet blitzkrieg and toward the management of local, limited high-tech wars. This shift was reinforced with more vigor after the U.S. Gulf War, but its origins are in the mid-1980s, not the post-Cold War redistribution of power. This shift in doctrine has been reflected in the creation of rapid reaction forces (numbering variously between 200,000–500,000 or about 10–25 percent of the PLA). ₁₆ However, these, again, are multidirectional, multipurpose forces, and training is not preoccupied with a single superpower enemy the way Chinese military force posture, doctrine, and training were focused on the USSR in the 1970s and 1980s. ₁₇

As for quantitative indicators of military capabilities, there has in fact been a decline in the size of the PLA, from about 3 million in 1985 to about 2.5 million in 1990. These numbers held steady through 1995. This reflects a decline in the ground forces as the PLA slowly shifts from anti-Soviet missions to rapid reaction to local wars along its borders. Acquisitions and deployment of major weapons systems also show no particularly rapid pattern of arming after 1991 (see figure 8.5), and in qualitative terms most of the weapons modernization going on still leaves the PLA a generation or two behind Western militaries. PLA acquisition specialists admit that Chinese weapons are “now roughly 15–20 years behind those of advanced nations.” ₁₈

Figure 8.5: Weapon Deployment Trends Source: IISS, Military Balance 1985, 1990, 1995

Military exercises and combined arms exercises at the group army level have picked up in frequency after the collapse of the USSR, but U.S. army analysts do not detect any specific enemy or contingency in mind in these exercises. Rather, the focus is on rapid reaction to potential local wars along China’s periphery, and on an overall effort to increase the operational flexibility of a very backward force. ₁₉ Some smaller unit exercises have been directed at a potential conflict with Taiwan. ₂₀

Naval modernization has been only indirectly aimed at U.S. power in the Pacific. Rather, the predominant driver appears to be to secure control over resources in territorial waters claimed by China in the East and South China Seas. A secondary goal is to be able to control SLOCs that are used by the United States and Japan. Consonant with these goals, naval doctrine has shifted from the coastal defense of the mainland to active defense of maritime economic, resource, and strategic interests. This means developing an operational capability that should extend to the interdiction of enemy SLOCs and attacks on enemy bases within the first island chain around China (running roughly from Kamchatka Peninsula to east coast of Japan, east coast of Philippines and east of Indonesian archipelago), and to the protection of Chinese long-distance SLOCs. ₂₁ This is reflected in the somewhat more concentrated effort in the 1990s to build a blue-water surface navy. Naval acquisition is focusing on missile frigates, submarines, supply ships, V/STOL-capable aircraft carriers (there are apparently plans for between three and five of these), a fleet air arm, marine and rapid-deployment force, plus C3I, expanded berth, and operational command systems that are more mobile. ₂₂ Naval exercises and operations have extended progressively farther away from coastal waters beginning in 1980. In the words of one participant in a 1988 conference on military futures held at the PLA Air Force Command College in Beijing, in the next century China should try to exercise effective control over its territorial waters, recover occupied islands, and “move towards international waters, the core being to seek the rights we deserve in international waters.” ₂₃ But these goals, and the initiation of this kind of thinking, predate the emergence of the United States as the lone superpower. ₂₄

As for nuclear modernization there are internal discussions among strategists about whether China needs to develop a limited warfighting capability to deter conventional war, nuclear war, and intrawar escalation in future high-tech local wars. ₂₅ Possibly to this end, China is developing more accurate, mobile land-based ICBMs, and a second generation of SLBM. Obviously, the United States is a potential adversary in this conceptualization—along with Russia and a future nuclearized Japan. But the interest in limited nuclear war capabilities developed in the mid to late 1980s, prior to the collapse of the USSR; thus it cannot be attributed to the emergence of the United States as the dominant power in the system. While it takes time for new systems to come on line, thus far, according to the Natural Resources Defense Council, the numerical size of the Strategic Missile Forces has not increased much after the fall of the USSR. ₂₆ Again, this may change if China’s acquisition is guided by limited warfighting thinking (the relationship between those who think about doctrine and those who make R&D and acquisition decision is unknown), and if the United States proceeds with development of THAAD ballistic missile defense systems. But the primary impetus in nuclear modernization has not been the redistribution of power after 1991.

H.A2: A second hypothesis generated by structural balancing theory is that states will also balance against the existing or emerging hegemon(s) by joining, or constructing, loose security alliances aimed at containing the hegemon. At the same time the state should try to loosen or undermine any alliance structures that buttress the hegemon’s power. At a minimum the state should be signaling to other potential adversaries of the hegemon an interest in loose coordination of policies on the basis of the realist principle that an enemy of my enemy is my friend. This hypothesis suggests that Chinese diplomacy after the Cold War ought to be driven by a search for allies and partners with whom to coordinate in balancing against the United States

A simple question to ask first is whether China is, in fact, acting less cooperatively toward the United States after the end of the Cold War. This would be a reflection of China’s identification of the United States as the emerging hegemon, and the calculus that, at the very least, China should not be accommodating such a power. A quick and dirty look at quantitative events data suggest that, indeed, China’s behavior toward the United States in the last few years has been obviously less cooperative than in much of the 1980s. ₂₇ There is a moderate negative correlation between Chinese military expenditures in constant dollars and the intensity (r = -26) and proportion (r = -.22) of cooperative actions directed toward the United States from 1984 to 1994. ₂₈ Certainly this is consistent with the analysis of outside observers that Sino-U.S. relations in the 1990s have been troubled, to say the least, by disputes over trade, proliferation, human rights, and the Taiwan question. One does not see now the degree of strategic coordination that existed between the two countries during the 1980s, for instance.

However, as the percentage of cooperative and conflictual actions indicate (figure 8.6), the precipitous drop in overall cooperative behavior occurred in 1989, prior to the collapse of the USSR. This shift was triggered primarily by the June 4 crisis and the effect this had in souring both U.S. and Chinese perceptions of the intentions of the other in a range of other issues.

Figure 8.6: PRC actions directed at the United States: percentage cooperative and conflictual Source: KEDS/PANDA (counting only WEIS scores over 1, so as to reduce error from coding ambiguous actions)

Even if the initial trigger of this decline in Chinese cooperative signaling were the legacy of June 4, perhaps the emergence of the United States as the dominant state in the system in 1991 perpetuated or injected new life into China’s calculus? If so, one expectation might be that China should try, at a minimum, to improve relations with potential adversaries of the United States, and, at a maximum, to coordinate policies with them. The principle ought to be a soundly realpolitik one—the enemy of my enemy is my friend. The quantitative events data suggests a fair degree of ambiguity on this score. ₂₉ The KEDS/PANDA data show a reduction in the ratio of conflictual to cooperative actions in relations with Iran after 1991. The pattern is similar for reported Chinese actions toward Iraq. The data set shows no change in the pattern in relations with Cuba—the small N of Chinese actions all being coded cooperative.

Another expectation from balancing theory is that China ought to be trying to consolidate relations with existing allies and strategic partners, particularly those whose relations with the United States have declined. The exemplar, of course, is Pakistan. The N is too small to say much with confidence (the addition of only a couple of recorded actions could change the cooperative:conflictual proportions quickly), but the PANDA data suggest that China’s signaling has been somewhat more conflictual after 1991. Relations with Pakistan remain close as the Chinese continue to believe that arming Pakistan will help divert Indian military resources away from China, but in the November 1997 Sino-U.S. agreement to implement the 1985 nuclear cooperation agreement China essentially pledged to stop all assistance to Pakistan’s nuclear weapons program. As for PRC signaling toward its one formal ally, and sworn adversary of the United States, North Korea, the ratio of conflictual to cooperative actions did not change over the 1991 divide. Arguably, these data have not captured a deterioration in Beijing-Pyongyang relations as China has used behind-the-scenes pressure to convince the North Korean regime to adopt an economic reform program and to rejoin the NPT, pressure for which the U.S. government has openly thanked the PRC. Indeed, while China still has a military alliance with North Korea, Beijing has signaled that it does not consider this to be a particularly hard and fast commitment. ₃₀

Anecdotal evidence is consistent with this mixed picture. There was some speculation in the early 1990s about an evolving Iran-PRC axis, though the cooperation on nuclear power development, arms sales, and trade, began well before 1991. The Chinese motivation is not only to have at least some voice and presence in another region (as major powers must, in Beijing’s view), but also to ensure positive relations with a major energy supplier, and to maintain good relations with a radical Islamic state in order to understand better the more sensitive political-strategic issue of Islamic fundamentalism on and within Chinese borders. However, there is little evidence of a foreign policy axis emerging. Indeed, in order to preserve relatively constructive relations with the United States, in November 1997 China provided a written, private guarantee to the United States that it would avoid any new nuclear technology cooperation with Iran and would phase out existing commitments.

According to Waltz’s balancing theory, we might also expect China to be trying to undermine American relationships with other major players in the system, including American allies. However, most Chinese analysts and officials continue to indicate privately a preference for a continuing U.S. security relationship with Japan so as to restrain Japanese military power, at least for the time being. ₃₁ As for South Korea, China has moved quickly to consolidate friendly political and economic relations with Seoul—with formal normalization in 1992—even though the ROK is a formal American military ally. As yet there is no evidence that China is trying to undermine this alliance, despite perfunctory calls for a reduction in U.S. military forces in the peninsula.

Sino-Russian relations have improved since 1991, despite Chinese invective directed at Yeltsin in for undermining socialism in the Soviet Union. The two countries have negotiated settlements for most of the disputed border areas, and in 1996 they concluded, along with three central Asian republics, an unprecedented multilateral CBM treaty placing limits on peacetime military operations along their border. China has also made a couple of high profile purchases of weapons systems—SU27s have been delivered, as well as some transport aircraft and four batteries of SA-10 SAMs; and there are persistent reports of Russian weapons specialists working in Chinese military industries. However, none of this signals anti-American security coordination. Indeed, some Chinese elites are worried about Russian-American military collusion against China. The 1995 Clinton-Yelstin communique pledging joint development of TMD systems raised concerns in Beijing. Moreover, there are serious perceptual problems in the Sino-Russian relationship, not the least of which are Russian fears that China is deliberately flooding Siberia with workers and businesspeople. Russian Minister of Defense Grachev has termed this a “peaceful invasion.” Chinese exports to Russia as a percentage of Chinese exports to the United States stood at about 6 percent in 1994–95. Chinese trade with Russia accounted for about 2 percent of China’s total trade, as compared with 15 percent for trade with the United States

In general, there is little evidence of the kind of security coordination with American adversaries directed at the United States in the 1990s that China developed with the United States and its allies against the Soviet Union in the 1970s and 1980s. At that time U.S.-PRC security cooperation involved, among other shared interests, highly sensitive exchanges of intelligence, ₃₂ arms sales, the coordination of assistance to the Afghanistan resistance and the Nicaraguan contras, and American incorporation of Chinese concerns into U.S.-Soviet arms control agreements (such as the INF Treaty). China is not trying to manufacture an anti-American united front, nor is it obviously trying to undermine American security commitments to other states.

Finally, if China were balancing against the United States due to a post-1991 structural shift in power one might also expect China to try to reduce its economic dependence on the United States As Gowa argues, trade can create security externalities. ₃₃ If a state perceives that the relative gains of economic interaction may be used to develop the military power of potential adversaries, then it has an incentive to reduce its trade with these adversaries and to diversify its economic linkages with more strategically reliable partners. ₃₄ Although China has carried a trade surplus with the United States for a few years, and thus, materially, the relative security externalities are in China’s favor, Chinese leaders tend to discount these kinds of externalities and focus on how China’s dependence has made it more vulnerable to threats to cut off or limit trade. Indeed, they have even called for reducing China’s trade dependence on the United States This has not happened, however. China’s dependence has increased over time. In the mid 1980s about 10 percent of China’s trade was with the United States By 1993 the figure stood at about 14 percent, and for the year from August 1994 through August 1995, the figure is 15 percent. ₃₅ Exports to the United States as a share of all of China’s exports jumped even more dramatically after 1991 (See figure 8.7). Chinese military expenditures are positively correlated with the percent of China’s total trade that is with the United States (r = .67). In other words, if China is balancing against the United States, it is not trying very hard to reduce the negative security externalities of trade with its adversary.

Figure 8.7: China’s trade dependence on the United States ($billions) Source: Almanac of China’s Foreign Economic Relations and Trade, 1988–1995

H.A3: A final hypothesis is that a state’s policy processes should reflect the specific balancing calculus embodied in H.A1–2. ₃₆ That is, if, as balance of power theory suggests, realpolitik actors are those who choose to balance, then we should expect to find with some high probability that the decisionmaking discourse reflects a realpolitik worldview, and specifically, that these decisionmakers view the rising hegemon as the target of their balancing choices. Thus one hypothesis would be that with the emergence of hegemon, the discourse ought at least to stress the need to balance against, rather than accommodate, the hegemon.

Much of the more authoritative Chinese discourse on the nature of the international system accepts that the system is, or close to, unipolar, though this conclusion is not a consensus position. It emerged as the predominant line only after the Soviet coup (which suggests that treating 1991 as the end of the Cold War has some face validity). This unipolar assessment was signaled by a lead article in World Economics and Politics, the authoritative internal circulation journal of the Institute of World Economics and Politics, written by He Fang, a senior foreign policy specialist in the State Council’s Center for International Studies. He argued that “it must be acknowledged that the U.S. really does possess all-round superiority in political, economic and military affairs, and in science and technology.” ₃₇ The official line acknowledges, in a roundabout way, the unipolar characteristics of the system by arguing that the system is in a transitionary stage to multipolarity from unipolarity. ₃₈

However, the policy conclusions that He Fang and others have drawn from this transitional unipolarity are that China should accommodate the United States on certain issues, or at least try to avoid preventing any dramatic downturn in relations. As researchers from the China Institute of Contemporary International Relations—the rough equivalent to the analytic side of the CIA—implied in an article in World Economics and Politics in 1993, because of China’s economic ties, its technological and developmental goals, and its relatively weak “comprehensive national strength” (zong he guoli) balancing against the United States could be very costly. ₃₉ Glaser reported that by the end of 1992 the dominant position was that China could not counterbalance U.S. power but should try to improve relations with it. Economic reform, market access, and technology transfers all dictated accommodation, not competitive balancing. ₄₀ In a fascinating study of about 130 Chinese foreign policy, business and academic elites in 1991–1992, Wang Jianwei even found substantial support for, or at least understanding of, the United States as a world leader, despite China’s being subject to considerable bullying from the United States. Surprisingly, almost 40 percent of the respondents expressed varying degrees of “appreciation of or support for American leadership in world affairs” because the system needed a leader and the United States was less overbearing and dangerous than other contenders. ₄₁ Even after Taiwan President’s Lee Teng-hui’s visit to the United States in June 1995 pushed Sino-U.S. relations to a new low, the Chinese leadership has apparently been keen on avoiding any irreparable damage to the relationship. Thus, contrary to Layne’s expectations that eligible great powers should be striving to help create a multipolar system, ₄₂ China may be vaguely hoping for the emergence of one; but it is not expending much economic or political capital to hasten its arrival or to weaken American hegemony.

Power Maximization Realism

This realism is most clearly identified with John Mearsheimer, with certain elements shared by Morgenthau. Essentially the argument is that states seek to maximize relative power or capabilities. While Morgenthau’s starting point was greed in human nature, Mearsheimer derives his offensive realist propositions from the characteristics of structure. For Mearsheimer, states want to maximize relative power because the uncertainty and fear generated by anarchy puts a premium on being prepared for attack from other actors in the system. ₄₃ Since the probability of such an attack is a function of whether the defender can credibly deter or eliminate threats, this probability decreases as the defender’s relative capabilities improve. States therefore seek to maximize their instruments of control over other states. There is no a priori logical barrier in this realism to states preferring pre-emption, or at the very minimum superior capabilities if possible. “Since the desire to attain a maximum of power is universal, all nations must always be afraid that their own miscalculations and the power increases of other nations might add up to an inferiority for themselves which they must at all costs avoid.” ₄₄

This suggests that for power maximization realism, states must always be essentially dissatisfied with relative position, even if all they seek is to maximize security. They want to maximize relative capabilities since this is what makes them most secure in anarchy. This is where power maximizers have a beef with defensive positionalists. The latter argue that states seek mainly to preserve their relative position rather than to improve it. ₄₅ But to power maximizers this axiom does not really warn against trying to improve one’s own relative capabilities. By increasing one’s own relative share of capabilities, by definition one prevents any increase in the relative capabilities of the other, and more effectively so. Given the fear generated by uncertainty under anarchy, it makes little sense only to try to freeze the status quo in relative capabilities; one can never be sure the other side will be content with the status quo. This means that, ceteris paribus, states will prefer to maximize offensive, power projection capabilities.

What, then, is the critical structural independent variable for power maximization realism? One could argue that, in principle, there should be no obvious variation in the interest in improving improve relative capabilities. ₄₆ That is, it should be a constant feature of state policy. But the ability to realize this interest is affected by at least two exogenous material constraints: the drain on resources that this single-minded pursuit to improve relative capabilities brings; and the probability that these efforts might compel a counter-coalition to form. We should expect to see the pace of power maximization to pick up as these costs decline, which will happen when other states are unable or unwilling to race or coalesce. ₄₇ The state will then exploit this opportunity to rapidly increase relative capabilities. Thus one should expect that in periods where other states are reducing their military capabilities, the state should increase its efforts to improve its own. Alternatively, a change in the systemic power distribution toward unipolarity might provoke power maximizing behavior. Even though the costs of this behavior may be quite high (given a state is “racing” with a unipole), the costs of not balancing against the unipolar are potentially very high as well. ₄₈

Power maximization realism would therefore make the following general prediction: H.B. States act primarily to increase their relative military power. More specifically: H.B1: a state will increase its military spending and step up the development of power projection capabilities when the opportunity to improve relative capabilities is low cost (e.g., when others are standing still or reducing their relative capabilities). Power maximization realism arguably allows for enough intentionality in the behavior of units such that they will calculate strategically to exploit opportunities for improving relative capabilities. Gilpin, for one, argues that differential rates of growth in power have a dynamic effect on state behavior by altering the “cost of changing the international system, and therefore the incentives to changing the international system.” ₄₉ Did, then, the end of the Cold War constitute a change in this cost for China, and if so a change in which direction?

Arguably, the collapse of the Soviet Union reduced this cost to China quite dramatically. For one thing it reduced the Russian ability to race the Chinese and to preserve existing ratios in capabilities. Indirectly, as well, the Russian ability to maintain existing capabilities ratios was constrained by a series of bilateral arms control agreements with the United States in the late 1980s (e.g., INF, START). For another it marked the end of a period of relatively sizable increases in U.S. military spending, and it saw the retrenchment of the military industrial complex as U.S. political leaders searched for a peace dividend.

Finally, the collapse of the USSR constituted the end of bipolarity. While the United States at present constitutes the predominant military power, according to Layne other players ought to be working especially hard to prevent the consolidation of U.S. hegemony. Thus with end of the Russian and American competition, and with the limited success in U.S. Congressional efforts to consolidate U.S. hegemony by increasing military expenditures in an era of low threats, the Chinese have both a window of opportunity and an extra incentive to increase their share of world capabilities in a relatively short period. Both factors are products of changes in global distributions of power after 1991.

Hypothesis H.B1 therefore suggests that the recent upturn in Chinese military expenditures ought to be due primarily to the reduced costs of “racing” with Russia and the United States. Moreover, internal maximization strategies are more realistic now, given the payoffs from economic growth in the 1980s. The same data used to test H.A1 applies here. If the post-1991 period presents a window of opportunity to shift substantially more resources in the military’s direction, the ME/GDP and ME/CGE data suggest this is not being exploited with much fervor. Additionally, this increase in ME began prior to the collapse of the USSR—that is, prior to the opening of this window. The pace of these improvements has generally not picked up much with the collapse of the Soviet Union, as the figures on military procurement indicate. In those areas where acquisition has increased since 1991, it is not clear these will be sustained. For instance, one quick fix to close the relative capabilities gap would be to rely on arms imports. According to SIPRI, however (using a composite measure of cost, lethality, and technical sophistication), there was a sharp jump in the arms imports index in 1992, accounted for primarily by China’s acquisition of SU-27s and SAMs, but the figures for 1993 and 1994 were much lower, suggesting this may have been a one-time jump.

In short, as I noted in the discussion of H.A1, the motivation for the slow but steady modernization of Chinese military capabilities—particularly its power projection forces—is more consistent with the desire to improve relative power than it is with the goal of balancing the system hegemon. But the pace of military modernization seems slower than power maximization might expect.

H.B2: unilateral efforts to improve relative capabilities will be supplemented by efforts to build a coalition to help the state’s relative capabilities improve (e.g., by providing technology, military assistance, or support for cooperative arrangements that constrain the relative power of others). It will, however, be sensitive to the costs this coordination carries for the autonomy of the state. These coalitions will be loose, with low levels of mutual coordination.

The evidence for this hypothesis is essentially similar to that for H.A2. China is not obviously trying to construct an anti-American united front; it has improved relations with American allies in East Asia, and it would, for the most part, prefer to maintain stable political, military and economic relations with the United States One might be able to argue from a power maximization position that China ought to be trying to maintain relatively good relations with the United States so as not to provoke a militarized response that increases the cost to China of arming, or that shuts the Chinese off from to the economic and technological benefits of the U.S. market. But power maximization theory suggests states will guard their autonomy carefully, and thus might be reluctant to become to closely entwined economically with a potential adversary.

H.B3: There should be evidence in the policy process that is consistent with general calculus embodied in H.B1–2. Decisionmakers therefore ought to see or portray the international system as anarchical, in which unitary actors pursuing the maximization of power are locked into competitive relationships that lead most often to conflict. Decisionmakers also ought to be arguing that in such a world it is best to maximize relative capabilities.

Arguably many Chinese officials and analysts do indeed accept that the nature of world politics is essentially zero-sum and that competition is the primary feature of the system. As one recent article in the authoritative internal circulation journal, World Economics and Politics, put it, systemic anarchy renders all interdependence asymmetric, and asymmetric relationships are essentially ones involving zero-sum power political struggles. As long as gains are asymmetric, the basis for conflict exists. Interdependence can accentuate interstate conflict by trampling on the sovereignty of nation-states, by preventing them from controlling their economic, military and political resources, and by providing opportunities for states to interfere in the internal affairs of others. The basic characteristic of the international system, therefore, is that states pursue short-term benefits by adopting policies harmful to others. ₅₀

It is interesting to note that in Chinese-language analyses, particularly after the collapse of the USSR, one rarely sees the term “new world order” (shijie xin zhixu) to describe the Chinese leaders’ normative vision of world politics (see figure 8.8). ₅₁ Rather the preferred term is “new international order” (guoji xin zhixu). The former connotes a system in which a hegemon is the primary impetus behind this order. ₅₂ Moreover, a new world order connotes a system of interstate norms which undermines state sovereignty, ostensibly in the search for cooperative solutions to global problems. In a new international order, however, the sovereign state remains the unit of analysis, and the primary objective of this order is to preserve the sovereignty and independence of states. ₅₃

Figure 8.8: Average frequency of terms per article in Shijie jingji yu zhengzhi, 1985–1995

This does not mean that the system is imminently dangerous, only that states should be prepared for potential threats. Indeed, some analysts explicitly identify the relatively benign security environment of the 1990s as precisely the one in which China can make strides in narrowing the military gap with other major powers without having to divert resources away from economic development. As weapons experts in the military wrote, “The relatively peaceful international climate and our friendly relations with our neighbors are providing fine external conditions for PLA weapons development.” ₅₄ As an internal circulation 1993 book for military cadres stressed in reference to the navy,

At present our navy faces a tremendous opportunity for development that is has not had before: the international environment is moving toward detente, the national economy is in the midst of rapid development, the reform and opening policy has become a state policy, the people’s “ocean consciousness” is in the process of strengthening, the Central Military Commission has given a great deal of attention to the development and construction of the navy, and the development and expansion of the navy has received the understanding and support of knowledgeable people in all spheres of life. ₅₅

On balance, Chinese behavior appears to be more consistent with power maximization realism than with balance-of-power realism. In particular, China’s goal in its security policy is more consistent with the desire to maximize relative capabilities as insurance in a potentially dangerous world than it is with the intention to balance against a rising hegemon.

I am leery, however, of jumping to the conclusion that power maximization realism has been confirmed. The main problem has to do with the motivation behind Chinese power maximizing behavior. In this version of realism the driving force behind state behavior is ultimately fear of insecurity in an anarchical environment. Long-term interests and goals are exogenous (security in an anarchical world) while shorter term goals and interests (the means to these long-term goals) are endogenous, shaped by particular power distributions and the opportunities these present for power maximizing behavior. The problem is that in the Chinese case a critical driver of its long-term goals is not security per se but status. Indeed, elements of Chinese realpolitik are sometimes put almost as a conscious choice. Writing in 1988, one analyst at a military conference on East Asian strategy noted, “In the early 21st century we must utilize the Soviet-U.S. detente, and the advantageous opportunities presented by their partial retreat from the Asia-Pacific region, to speed up the development of our diplomacy, throw off the limitations of ideology, adopt flexible diplomacy and gradually expand our country’s influence, increase our status and effectiveness in the Asia-Pacific region.” ₅₆

The status arguments behind power maximization are explicit and implicit in internal discussions about why China needs nuclear weapons, why it needs aircraft carriers, and why it needs to control the South China Sea. In each of these cases a powerful argument is that these are signs that China has “stood up” in world affairs. Indeed, Chinese foreign policy discourse since 1949 has been replete with status-related language, ₅₇ and foreign policy behavior since 1949 has been aimed, in part, at protecting “national self-respect” (minzu zizun), at enhancing China’s “international position” (guoji diwei). ₅₈ The most visceral, hence deeply ingrained, emotional reaction to Western criticism of China’s record on human rights or its foreign policy behavior, is usually a heartfelt recitation of the 100 years of humiliation that China suffered at the hands of Western imperialism, including the loss of territory through “unequal treaties.” Thus when Chinese leaders and analysts talk about the post Cold War period being the most advantageous for China since 1949 (zui you li de guoji huanjing) this is not simply a reference to the relatively low level of security threats, but also to China’s influence in international institutions and in world economics and politics in general. ₅₉

This drive for status is not directly structurally determined. It may be a product of structure at time t, but its persistence at time t + 1 is independent of structural conditions (e.g., China’s drive for major-power status has been a persistent policy goal of leaders inherited from the legacy of weakness in the face of imperialism in the nineteenth century). It is this historically generated, ideationally propelled and politically institutionalized interpretation of China’s place in international politics that explains the persistence of power maximization behavior, independent of a structurally determined drive for security. The disconnect between structural threat and/or structural opportunity to increase relative capabilities is illustrated by Deng’s remark in 1984 that if China could quadruple its national wealth by the end of the century, it could devote more resources to military capabilities. ₆₀ A major policy study commissioned by the State Council in 1984, called “China 2000,” was explicit that in the 1990s more resources would be devoted to the PLA and to hinterland development. ₆₁ It is hard to say to what degree the study is still guiding the state’s development strategy. But the point of both examples is that regardless of broad structural transformations in the international system, Chinese leaders have taken it for granted that China needs to increase its military power in the not-too-distant future. ₆₂

Some Chinese strategists now also speak of the need for “survival space” (sheng cun kongjian) and for strategic frontiers that extend horizontally into the Indian Ocean, the South China Sea, the East China Sea, and vertically into space. But they have spoken of these “soft frontiers” at least since 1987, and not specifically in reaction either to U.S. hegemony or post-Cold War opportunities to increase relative capabilities for security reasons. ₆₃ Advocates of naval modernization and the development of the “near ocean defense strategy” (jin hai fangyu zhanlue) speak not only of protecting shipping lanes, fishing grounds, and SLOCs, but also of increasing China’s “national awesomeness” (guo wei) and expanding its “political influence” (zhengzhi yingxiang). ₆₄ As some proponents of naval modernization have put it: “If we are to be a world great power, we must not merely cast our eyes at the continental land, but must also look toward the oceans, and we must have a strong “ocean consciousness” [haiyang guan]. If we don’t have a strong and powerful ocean force, it is very difficult to enter the ranks of the world’s powerful states.” ₆₅

Status is not unrelated to security, of course. Status buys deference to one’s interests; it buys access to states and international institutions. But as Schweller points out, there are states that have revisionist tendencies not because the present structures make them insecure, but because the present structures do not allow them to appropriate other goods (e.g., status and deference). ₆₆ In China’s case, there is a fair degree of evidence that, at the moment, it is not security maximizer, but a “prestige maximizer.” ₆₇

How serious the damage, then, to power maximization theory if it gets the motivation for China’s power maximizing behavior wrong? I think it is quite serious since it means that the theory gets a basic prediction, H.B1, wrong: immediate security fears are not the the obvious driving force behind China’s military policies in the 1990s, and the opportunities to improve relative capabilities with the collapse of the USSR emerged after China began to increase its military expenditures and power projection forces.

Balance-of-Threat Realism

Balance-of-threat theory, as developed by Walt, predicts that states will balance against the primary threat as defined by aggregate power, geographic proximity, offensive power, and aggressive intentions (a component of which is ideology that may affect perceptions of enemy intentions). The primary threat may or may not be the dominant power in the system.

The theory is fairly indeterminate, however. For one thing, as Walt acknowledges, it does not provide any a priori aggregate weighting system for the four key variables. Each variable could lead to predictions about behavior that contradicted the other. ₆₈ This raises problems of falsifiability. For another, virtually any perception of threat drives the model, and the threat need not be to strategic/territorial interests, as is the focus of traditional realism. Threat can mean threat to the domestic politics and the internal legitimacy of a regime. ₆₉ Again this introduces a great deal of indeterminacy into the theory, since if one finds balancing behavior in the absence of a military threat to the territorial integrity or strategic interests of a state, one can always claim that the threat is in fact to the internal legitimacy of the state, an altogether more subjective, flexible, hence manipulable definition of threat. To make the theory testable, then, I think it has to define threat in more “traditional” realist terms—as a military threat to the territorial integrity or military security of the state. Threats to the legitimacy of the regime may be derivative of strategic threats, but they should not be included in an amorphous independent variable.

Finally, Walt’s theory is not clear about what happens when when one threat ends or declines. Is it necessarily succeeded by another? Do states require or expect that one threat will be followed by another? Balance-of-power theory suggests there is a constant shifting of the target of balancing: once one hegemon is balanced against, then the system balances against the initial balancing state or coalition. ₇₀ But Walt is unclear about this. If threats are constructed—that is if we assume, as I think Walt has to, that the crucial variable is threat perception—which interprets the meaning of military power, geographic proximity, and offensive doctrines—then the replacement of an old threat by an new one is not automatic. So once the threat disappears, balancing behavior should cease. ₇₁

H.C: The general prediction is that states balance against the primary threat. Thus with the collapse or end of that threat (in the absence of the appearance of another) we should see the end of balancing behavior until another threat emerges.

This first step in testing this this general hypothesis, and its derivatives, is to show that variation in Chinese internal and external balancing behavior is consistent with variation in perceptions of the primary threat (either the disappearance of the old threat and/or the appearance of a new one). There is little doubt that the collapse of the USSR signaled the end of a long-standing military threat to China, and the major impetus behind China’s force posture and alliance behavior from the late 1960s through the late 1980s. The question is whether Beijing sees newer and imminent threats at present or over the horizon. The evidence on this is ambiguous. Perhaps indicative of this, a major PLA conference consisting of a wide range of strategic and operational analysis institutions held in late 1993 to determine precisely this question came to a range of estimates. According to Hong Kong reports of the conference, about 60 percent of the attendees believed primary threat in the next century will be Japan, 40 percent believed it would be the United States, while 10 percent thought it would be a resurgent Russia. ₇₂

Nonmilitary analysts seem divided as well. During the sharp downturn in relations in mid 1995, some analysts apparently concluded that U.S. policy was determined to keep China divided and weakened, to prevent it from challenging U.S. hegemony. ₇₃ Others have observed, however, that American foreign policymakers are not engaged in a monolithic conspiracy to contain, divide, and weaken China, but are themselves being buffeted about by conflicting domestic interests, the division of powers, and other peculiar features of U.S. policy process. ₇₄ While some hardline analysts believe that the United States is trying to encircle China—vide U.S. efforts to control political and economic reform in Russian, U.S.-Vietnamese normalization, American promotion of multilateral security institutions in East Asia, improved U.S.-Indian relations, the promotion of Taiwanese efforts to increase their international status ₇₅ —even they are willing to concede that this strategy is largely political and economic in nature: “If one looks at it from a purely military security angle, [the United States] does not constitute a realistic threat.” ₇₆ The general assessment that China operates in the most benign strategic environment since 1949 still stands.

If the perceived threat is not clear, or as Chinese leaders proclaim, relatively low, then according to balance-of-threat theory we should not expect to see much internal or external balancing. On this score the evidence suggests this version of realism does not do a particularly good explanatory job.

H.C1: As the threat declines or disappears a state will slow down, stop, or reverse the rate of growth in its military expenditures and in its development of any power projection capabilities that could be used against the former threat. ₇₇ Arguably this hypothesis is disconfirmed by the continued, if moderate, real growth in Chinese military expenditures and military power projection capabilities in the 1990s. Of course, these power resources are not necessarily directed at Russia, but they could well be. They could be used against Russian capabilities in East Asia, Central Asia, or against Russia itself. More generally, however, military modernization is directed at multiple possible contingencies, one of which is considered imminent (anti-Taiwan operations).

H.C2: As the threat declines or disappears, a state will withdraw from, or cease seeking alliances that aggregate capabilities, and will draw down efforts to break apart existing or real coalitions that favor the former threat. Consistent with this proposition, given the collapse of the Soviet threat, China is no longer trying to build a united front directed at Russia, as it did in the 1970s and 1980s. Given that there is no consensus yet as to the degree to which the United States constitutes a clear strategic security threat, we should also not expect to see alliance building against it. And indeed, as I noted, despite problems in the Sino-U.S. relationship the Chinese are still interested in developing closer military ties with the United States, in part to acquire technology, training, information about command and control and operations through these exchanges.

The problem for the theory is that it is not clear that the end of the Cold War per se had much to do with this diffusion of potential targets of Chinese military power. On the one hand, though Soviet threat was, in the Chinese view, declining somewhat over the 1980s, this was debated inside China well into the Gorbachev period. His reforms were viewed as an effort to restore Russia’s ability to vie for hegemony and China was still a potential target. In this sense, the collapse of USSR in 1991 could be viewed as the critical change in the independent variable. Indeed, sometimes the Chinese discourse will speak of the end of the Cold War and the breakup of the Soviet Union in the same breath. On the other hand, Sino-Soviet relations were “normalized” in 1989. The Soviets unilaterally met China’s conditions for normalization, and in doing so reduced the operational military threat against China (e.g., by withdrawing forces from Mongolia). U.S.-Soviet arms control, particularly the INF agreement, also improved China’s security. All of these changes preceded 1991. So, as the theory predicts, the decline in the threat has led to a decline in external balancing efforts and a diffusion of internal balancing efforts. But it is not clear that the decline of this threat and the end of the Cold War are the same thing.

H.C3: As the threat declines or disappears, so does the assessment of threat in policy process. Since this theory relies heavily on subjective judgments about threat (given the potential role of ideology in defining threat), it is even more important for its validity that ideas and discursive practices are consistent with behavior. Indeed, in this theory these practices are at times causal, not ephiphenomal.

There is no doubt that Chinese threat assessments are consistent with an end to the Soviet threat. Chinese commentary on its relatively friendly external environment is not just public posturing: internal circulation papers written by top analysts in the intelligence and foreign policy communities generally endorse this assessment as well. ₇₈ The problem for the theory is that if Chinese leaders were balancing against threats, the discourse would be inconsistent with the behavior outline in H.C1. That is, even though China’s threat environment is relatively benign, its internal balancing behavior in particular is inconsistent with the behavior the theory predicts given this change in environment. If the behavior were consistent with the discourse, one might expect to see Chinese leaders arguing in favor of slowing down the military modernization program, now that threats have declined.

Identity Realism

This form of realism has heretofore been unexplicated in any detail, but it flows from work in social constructivism and social psychology on group identity formation and intergroup conflict. Much of this disparate literature essentially holds that the creation of in-group identities leads directly to the devaluation of out-groups. This in turn leads to competitive interpretations of the relationship with the out-group. The creation of and intensification of group identities, then, positively correlates with the degree of competitiveness with the out-group. ₇₉ Thus the initial motivation for self-help behavior is not the survival of the group (state) but its creation.

The literature is murky about the reasons for this out-group devaluation, however. Some explanations rest on individual material interest. ₈₀ But Kinder and Sears’s work on racial prejudice shows that the devaluation of the out-group is not obviously interest-driven among most members of the group. Put another way, the in-group does not resent or distrust the out-group because of some tangible threat to in-group’s interests (as neorealism/realism would posit, whether because the out-group is a predator or is structurally compelled to compete). Rather it is socialized and learned—whether or not a tangible threat exists—and is reinforced by pressures to conform. In symbolic racism the cognitive structure of this devaluation is based on dramatically differentiating the values of in-group and those of out-group: the other is seen as morally unalike (e.g., blacks are lazy, criminal, don’t work hard, are unpatriotic etc, among other stereotypes). ₈₁ The cognitive structure of nationalism and nation-state identity is probably akin to symbolic racism in its normative condemnation of the other. That is, identity creation must be based on stereotypes of self and other, not just on evidentiary-based arguments about conflicting material interests. The persistence of the former in the absence of the latter suggests identity construction is not epiphenomenal to the state’s need to mobilize resources in a given competitive anarchy (just as the persistence of symbolic racism even when no concrete interests are threatened by blacks suggests that racial conflict is a constructed, not an economic, fact). Rather the causal arrows run the other way: identity construction, and its intensity, determine anarchy and how much fear and competition results. Applied to international relations, then, the literature would suggest that changing intensities of in-group identity affect the degree of outwardly directed realpolitik behavior, regardless of changes in structural environment. ₈₂ Moreover, variation in in-group intensity will explain variation in the targets of this realpolitik. That is, weak or porous in-group identification provides greater cognitive space for identifying, at least in part, with other groups. To the extent that identity expands to these groups, they no longer serve as logical targets of realpolitik behavior. ₈₃

In sum, if there is an increase in the intensity of identity—or efforts to intensify identification—we should expect to see more competitive behavior directed at the out-group. As the boundaries of the in-group expand or contract we should expect a change in the degree of competitive behavior directed at the those actors comprising the old in-groups and out-groups at time t.

What might an identity explanation of realpolitik consider to be the critical independent variable explaining this initial variation in intensity of identity? Unfortunately there is no clear guidance from either the social psychology literature or constructivist work in IR theory. One obvious possibility that flows from the symbolic basis of identity is as follows: a threat to the legitimacy of the in-group’s composition, or to forms of legitimate organization, or to its cohesion and its values, rather than specifically to the material interests of the group, leads to efforts to shore up in-group identity. Thus threats to the in-group will vary depending on self-evaluations of the content of the primary valuations of the group. Put another way, identities have two elements: The first is the group specific values associated with the self-perceived role of the group—the content of identity. The second is the cross-group tendency to devalue out-groups. The specific content of the identity will tell the group what it ought to value, and thus how to define what and who threatens its integrity. The second element tells the group how it ought to react to these threats, namely in a realpolitik fashion. Thus competitive strategies will be adopted toward threats that are determined by the specific content or legitimacy of the (predominant) identity. The content of in-group identity is not given by position in the international system, as essentially structural realism would argue: indeed, empirically this content varies far too much across states similarly and differently positioned for it to be determined by material structures. Democratic identities shared by members of a European security community are provided neither by anarchy nor by different distributions of power under anarchy.

This does not mean that identity realism ignores material, structural variables. Identity realism acknowledges that structural variables—namely changes in relative capabilities—are important steps in predicting state behavior. But these are important because the realpolitik ideology that is a creation of in-group formation interprets changes in relative capabilities in particular ways. Identity creation makes a group sensitive to relative capabilities only to the extent that, given the prior content of the group’s identity, the group believes it is in a competitive relationship with an out-group. But the logic of the argument suggests that this sensitivity varies as intensity and scope of identity varies. So changes in relative capabilities matter between in-group and out-group, but they matter more as in-group intensity increases and the out-group is more starkly devalued. Conversely they matter less the less intense the in-group identity, and by definition, the less starkly the other is devalued, even though the groups exist, technically, in anarchical relations. They matter not at all between groups that believe they are part of a larger in-group.

This addresses a problem that other realisms have had difficulty with—an inability to account for variations in the intensity of realpolitik behavior and variation in the responses of different groups to similar changes in relative capabilities. Some realists have tried to account for the variation in competitiveness by invoking offense-defense balances and their effect on the intensity of the security dilemma. ₈₄ That is, the competitiveness of groups increases when offense dominates and the security dilemma is especially acute. This has nothing to do with in-group identity intensities—these are epiphenomenal—but with offensive-defense balances. But this still begs the question why one state would necessarily interpret another state’s offensive advantage as threatening if, say, the states are in a friendly relationship.

If self-help is a product of group identity creation, then it is in the interests of the group to preserve the high evaluation of the in-group and the devaluation of the out-group. Institutions are thus necessary to uphold, socialize, and transmit identity, and by doing so they perpetuate realpolitik interpretations of the out-group across time. It is in the group’s interest to do this, in part to prevent free riding. One would therefore expect to see the institutionalization of self-help axioms accompanied by norms that delegitimate the out-group and that punish those whose identities are loose enough to allow defection from the group. ₈₅ Hence the language used to intensify in-group identities, in effect, compels marginals, or those on the fringe of the group, to declare themselves for or against the group. The threat implicit in this language is that those who do not, or those who try to straddle identities are threats to the integrity of the group, and thus are beyond the moral pale, no different than the out-group. This stark othering deters members of the group who might defect. ₈₆

Thus realpolitik becomes an institutionalized ideology stressing the “linked fate” of the in-group, ₈₇ the dangers of the disorder in the environment beyond the group, and the dangers presented by marginal members. ₈₈ As Hobbes argued, domestic cohesion around the state-as-in-group—the Leviathan—requires this ideology. ₈₉ The group therefore needs to provide arguments about the competitiveness of the environment in order to reinforce in-group identification. This is what realpolitik as ideology does. Arguments about the inherently competitive and dangerous external world might be supplied, for example, through official interpretations of the group’s history, the accepted language of strategic debate, or policy process structures that marginalize certain voices and interests. ₉₀ Together, realpolitik ideologies and their institutions are designed to prevent groups from perceiving security dilemmas as the basis of insecurity. The logic of in-group creation compels a group to emphasize that insecurity comes from the predisposition of out-groups to threaten it. Thus from an identity realist perspective, the institutional struts of realpolitik should also be essential focii of research.

To summarize, identity realism would make the following general predictions: H.D: the greater the intensity of in-group identity, the greater the devaluation of the out-group, the more competitive the policies directed at the out-group. The root cause of this increasing intensity of identification is a challenge to the legitimacy of the content of a group’s identity. So, in the Chinese case there are three issues. First, to what extent has there been an increase in the intensity of in-group identification and a concomitant increase in the devaluation of out-groups? Second, to what extent is this change, if there has been one, a function of the end of the Cold War? Third, to what extent has this change correlated with any change in the competitiveness or coerciveness of Chinese realpolitik behavior? I will address this last question when I come to the specific hypotheses H.D1–3.

As to the first issue, I have no good indicator for intensity of identity. I am not alone on this score. Until very recently, the literature on identity is remarkably free of any effort to think about indicators of the scope and intensity of in-group identification. Some of the literature that measures patriotism and nationalism has used, in combination with other survey tools, 5-point Likert scales. One might also use Osgood semantic differential procedures, or a “meta-contrast” ratio. ₉₁ But these are applied to surveys, not to the content analysis of historical documents. When one has to rely on content analysis, applying these methods is difficult—one has to rely on groups of well-trained coders to judge where the author of an article or document might place her/himself on a scale with no reference to the ideas of the author other than what exists in the article. ₉₂

In lieu of rigorous measures of the intensity and exclusiveness of in-group identity, I rely for the moment on a mix of anecdotal indicators: the degree to which nationalist symbols have replaced Marxist-Leninist ones in the public and decisionmaking discourse on China and its role in international relations; the degree to which discourses about China’s relationship to the outside stresses the former’s uniqueness and differentness; ₉₃ and the language used to describe marginal or fringe actors (e.g., the discourse used to describe those whose identities are suspect, or potentially more diffuse: the more the group is interested in excluding fringe actors, or deterring their defection from the group, the shriller the discourse used to describe them).

The common, and probably accurate, observation from China specialists is that with the collapse of Marxist-Leninist ideology as a force for social cohesion and political legitimacy, the regime has turned with more urgency to a victimization discourse, ₉₄ and to a new amalgam of appeals to an ancient, glorious Confucian past, ₉₅ barely disguised racialist discourses on the greatness of the Chinese people, ₉₆ “Asian” ways toward human rights and political and economic development, ₉₇ and characterizations of conflict in the international system as residing increasingly in differences in culture and social structure, among other stark appeals to Chinese identity.

Much of this discourse has been embodied in the ubiquitous adjectival phrase “Chinese characteristics” (zhongguo tese) used to modify a dizzying array of activities: China is building socialism “with Chinese characteristics”; artists and musicians are urged to produce cultural products with “Chinese characteristics”; academics are tasked with developing social science theories “with Chinese characteristics.”There is, to be sure, a great deal of perfunctoriness in all of this, but it reflects a broader interest the regime has in wrapping valued behavior in allegedly unique cultural characteristics, and implicitly devaluing alternatives by labeling these foreign, different, and alien. I have not done any rigorous frequency count of the term, but my sense is that it became a shriller, more ubiquitous and more reactionary concept after June 1989.

The regime has also implicitly attacked those with marginal identities or who might try to hold multiple identities by once again delegitimatizing “cosmopolitanism.” The critique of cosmopolitanism (shijie zhuyi) picked up after June 1989, in part because the regime believed that some democracy activists saw themselves as part of a broader global historical ideological trend. Some scholars who worked on issues relating to cultural convergence (wenhua qutong) through internationalization of national economies had to drop their research focii. In some internal circulation materials the “open consciousness” (kaifang yishi) prior to June 4 was blamed for the proliferation of such unpatriotic concepts as “citizen of the world” (shijie gongmin), “global village consciousness” (diqiu cun yishi), “convergence theory” (qutong lun), as well as “foreign worshiping” behavior and the negation of the cultural legacies of the Chinese ethno-nation. ₉₈ The regime also critiqued the concept of the “new world order,” in part because the core unit of analysis in the term is the globe rather than the nation-state. In sum, there is important evidence that the regime has attempted, at least, to intensify identification with the Chinese state using symbolic cues with deep resonance in the socialization of the Chinese people.

It is clear the regime has thought fairly carefully about the politically instrumental purpose of these nationalist discourses. In what may have been the summary of an explicit research project on the uses of direct nationalism, an analyst with the State Council Information Office—the office charged with controlling information into and out of China—wrote in 1995 that nationalism could have a number of functions. Among the positive ones were unification of a people and nation, and preservation of the nation-state in the face of external aggression. ₉₉

The second question is when did this process begin? This has a bearing on the third question, namely, whether there is a correlation between this discourse and China’s security policy. One thing is fairly clear: the efforts to increase the intensity of identification with the Chinese state—while never absent in post 1949 history, of course—picked up before the collapse of the Soviet Union but after June 4, 1989. This should not be surprising. The protests against the regime and the aftermath of the crackdown in June 1989 constituted an unprecedented crisis in the regime’s legitimacy. The initial discourse was especially shrill in its accusations of foreign efforts to overthrow socialism in China through “peaceful evolution.”

In essence, June 4 provided information to the regime that “revealed” not only that the outsiders were trying to undermine the integrity of the in-group, but also that they held much different approaches to the inviolability of internal affairs. In other words, June 4 provided information to Chinese leaders that suggested more work needed to be done to intensify in-group identification (so as to maintain the legitimacy of the in-group political status quo) and also suggested that differences with the out-group were larger, more pressing, and involved higher and more competitive stakes than they had previously acknowledged.

The talk about peaceful evolution has been toned down in volume since 1991, but it is clear that the regime does not believe the crisis in internal legitimacy has passed. The crisis has been accentuated or at least perpetuated, in Chinese eyes, by American pressure on China on human rights, U.S. support for the internationalization of Taiwan’s foreign policy, and the condescending way that the United States has described China as a violator of international norms of behavior in a range of issues. The crisis of legitimacy has been accentuated as well by a deep concern about other processes of identity creation or consolidation within China, namely in Tibet, among Muslims in Xinjiang, and among island-born Taiwanese (hence the seriousness with which the Chinese leadership takes the crisis with the United States over Taiwan’s status). ₁₀₀

The end of the Cold War, in a sense, only indirectly explains this process of fostering more exclusive in-group identification. Chinese leaders argued that, as China’s strategic value to the United States has declined with the collapse of the USSR, the U.S. use of human rights to threaten China’s political stability reveals the true essence of U.S. hegemonism and power politics, thus accentuating the differences between the two social systems. The primary charge leveled at the United States has been that it wants to “divide and westernize” (fenhua xihua) China, and to prevent it from gaining its rightful status as a great power. The U.S. threat is not aimed directly at China’s territorial integrity or military security, but rather at the vision of China as a sovereign, independent, great power, both economically and militarily. The end of the Cold War, then, did not make this in-group–out-group differentiation inevitable. Rather it reinforced information about in-group–out-group differences that were made especially apparent with the events of June 4.

To reiterate, this intensified identity discourse is primarily a function of a regime of shaky legitimacy trying hard to convince its people that only the Communist Party can lead China to its place of greatness in the face of external adversaries. This discourse has, as I noted, sharpened after June 4, with the decline of even relatively benign interdependence themes of the 1980s. ₁₀₁ The discourse is not a product of the end of the Cold War, and the emergence of the United States as the lone superpower. It is therefore not an epiphenomenon of structurally induced balancing behavior. The discourse may be an ephiphenomon of the leadership’s quest for legitimacy, but it would not cue the desired identification in the population if there were not fertile ideational ground for such symbols.

Thus identity issues have had two effects: the crisis of legitimacy has compelled greater efforts to foster identification with the Chinese state as led by the CCP, and the symbolic content of the identity discourse is determined by a historically rooted identity as a victimized great power. Indeed, some political insiders in China have explicitly recognized the importance of cultivating nationalist sentiments and an “enemy psychology” (diqing xinli) to gird the population for competition with the United States and to prevent internal disorder along the lines of June 1989. The language should stress that the U.S. intention is to exterminate China (mie wang zhongguo), as stark a zero-sum image as one can imagine. ₁₀₂ The net hypothesized effect of the interplay of these two components of identity should be a more intense, or harder, realpolitik behavior toward out-groups, regardless of changes in material structures.

How well, then, does identity realism do in explaining change in Chinese security policy? At a very general level, of course, one could argue that it does a very good job establishing the necessary conditions for the realpolitik core of Chinese behavior. Like any most other states, ceteris paribus, the creation of Chinese state identity entails devaluing external actors and portraying the external environment as conflictual. This process creates relationships under anarchy and interprets them as conflictual: hence an overarching preference for realpolitik behavior. This preference ought to persist regardless of changes in material structures. The rhythm of realpolitik behavior, in other words, ought to be out of sync with changes in material power, but in tune with changes in the intensity of identification as required by the internal cohesiveness of the group. As for more specific predictions about security policy, one can deduce the following:

H.D1: Ceteris paribus the greater the intensity of in-group identity, the greater the devaluation of the out-group, and thus the greater the increase in military expenditures and arming directed at the out-group. Most importantly, for our purposes, the intensification of the identity discourse correlates temporally with the push to increase real military expenditures and improve military operational capabilities in the late 1980s. Moreover, the primary driver of this modernization program—achieving the status of a great power—is consistent with the content of Chinese identity as reflected in the preeminence of realpolitik ideology and institutions. Contrast this with the other realisms that predict security not status will be the primary goal, and that a change in military expenditures and acquisitions should be a function of the change in power distributions at end of the Cold War in 1991.

H.D2: the greater the intensity of in-group identity, the greater the devaluation of the out-group, the greater the effort to seek alliances that constrain or balance against the out-group. These alliances may be tighter than one might expect if there is a greater probability that allies will be identified as members of the in-group. If there is no such identification, then loose, temporary security coordination will be preferred. Therefore, China ought to support relatively tight alliances with states that are perceived to share similar in-group values. Otherwise it should prefer loose alliances against the devalued out-group. In many respects the post-June 1989 discourse on identity has stressed China’s uniqueness and its determination to preserve the sovereignty and autonomy of the state. Thus it has not sought or joined tight security coalitions aimed at devalued out-groups like the United States But neither has it sought or joined loose coalitions aimed at containing U.S. power. This is a problem for the argument if indeed the starkness of in-group–out-group differences has been accentuated.

China’s global human rights diplomacy is more consistent with the argument, however. Part of the discourse on human rights has stressed the common approaches to the relationship between the individual and the state shared by such Asian states as Singapore and Malaysia. As the theory would expect, this common discourse has been reflected in Chinese efforts to construct a coalition of states opposed to Western pressure on human rights, particularly human rights defined in terms of individual political and civil liberties.

H.D3: the greater the intensity of in-group identity, the greater the devaluation of the out-group, the starker and more denigrating the images the policy discourse will present of the out-group, the more the in-group will use these images to mobilize in-group identities. Discourse about strategic threats will follow from discourses about threats to the valuational integrity of the in-group (e.g., strategic competition will flow from perceived valuational competition, not the other way around).

As a first crack at measuring an increasing sensitivity to the independence and autonomy of the Chinese state (which, for most realists, ought not to vary), a randomly sampled set of articles from an internally circulated, authoritative journal on international relations, World Economics and Politics, was content analyzed. These were coded for the frequency of specific terms that reflected judgments about the conflictual nature of the international environment and about degree of congruence between “national interests” and the interests of other actors. ₁₀₃ Specifically, each article was coded for the frequency of concepts that tapped into two different poles of thinking about the international system and China’s relationship to it—a hard realpolitik pole in which the system is highly conflictual, and in which independence, sovereignty, and national interest are of primary importance, and an interdependence pole where cooperation, the predominance of global interests, and international institutions characterized the system. My assumption is that the more frequent the concepts that cluster around the former pole, the more likely that Chinese analysts stress China’s identity as an independent, autonomous actor with fewer common interests or values shared with other states. The more frequent the concepts that cluster around the latter pole, the more likely that Chinese analysts accept more porous identity boundaries with outside actors, and acknowledge some common interests and values.

This is, to be sure, a very rough indicator. Nonetheless, one does see a jump in the references that describe the international system as conflictual and that stress national interests and the Chinese version of a new international order (which is based on the enhancement of the independence and sovereignty of states) after 1988 but before 1991. As figure 8.9 shows, after 1988, not after 1991, the cumulative frequency of conflictual images takes off on a trajectory at a rate that exceeds the cumulative frequency of cooperative references. This is what one should expect if the legitimacy crisis of 1989 were driving the efforts to intensify in-group identity and realpolitik ideology.

Figure 8.9: Content analysis of Shijie jingji yu zhengzhi

 

Conclusion

Given the messiness of the realist theories I have tested here, I feel somewhat less concerned about the messiness of my conclusions. One thing is clear: the first three realisms do not come through with flying colors, even though I tried to create ideal test conditions for each of them. Balance-of-power realism is problematic for a number of reasons. First, the intensification of China’s military modernization efforts has preceded, rather than followed from, a substantial structural shift in power in favor of the United States Second, China is not determinedly internally or externally balancing against the United States, despite this structural shift. Third, despite the emergence of the United States as the predominant military power in the system, China is becoming more, not less, dependent on trade with the it. Finally, despite this shift, and despite the recognition in the security policy discourse that the present system is highly asymmetrical, if not unipolar, Chinese decisionmakers have not determined that the United States is China’s primary military adversary or the primary target of its balancing behavior. They continue to argue that from China’s perspective the post-Cold War period is highly competitive but for the moment relatively benign strategically.

At first glance, power maximization realism does a better job accounting for much of Chinese security policy. Chinese military modernization has proceeded as the costs of trying to maximize relative capabilities have declined with the collapse of the USSR. The Chinese policy discourse is also consistent with power maximization’s assumptions about the competitive nature of the environment and the need to exploit opportunities to increase relative power. The difficulty with the theory, however, is that, again, much of this behavior precedes the collapse of the USSR and the reduction in costs of power maximizing. Moreover, it is not clear that the motivation behind power maximization is security as much as it is status—a driver that is much harder to attribute to systemic anarchy and changing power distributions, as Schweller has noted. The fact that both balance-of-power and power-maximization versions of structural realism get what should be an easy, “most likely” case wrong is, or ought to be, profoundly troubling for those who believe that anarchy and polarity are critical variables in explaining the security policies of major powers. In other words, their disconfirmation in this particular case ought to be weighted quite heavily for defenders and critics of structural realism alike. If not in this case, then when does balancing occur for material structural reasons? ₁₀₄

Balance-of-threat realism is not a particularly good explanation of Chinese security policy either. In general, China’s military expenditures and modernization of operational capabilities are on a gradual upward trajectory even though the primary security threat has essentially disappeared. China is not balancing against any identifiable military threat at the present, and decisionmakers and strategists are divided about which state may constitute the next security challenge. Although the United States presents the most foreign policy headaches for the regime, even hardline analysts do not believe the United States presents an imminent military threat. Indeed, for the most part Chinese decisionmakers consider the post-Cold War security environment to be, on balance, the least threatening since the founding of the People’s Republic.

As for identity realism, the discussion here is a first cut at exploring a constructivist perspective on realpolitik behavior. To be sure, this version of realism has been the most difficult to operationalize, and hence to test convincingly. Two things can be said for it, however. First, the timing of China’s discourse about international relations and its role in the system is consistent with the acute accentuation of the regime’s legitimacy crisis in 1989. Identity realism would expect under this condition that the regime try to foster greater in-group identification with the Chinese state and CCP leadership. Second, the content of this discourse is also consistent with identity realism’s expectations. In order to intensify in-group identification the regime ought to stress the uniqueness of the in-group, the potential dangers of the external environment, and the valuational differences between in-group and out-group. Hence, the intensification of nationalist symbols, the invocation of traditional cultural values, the relative increase in conflictual images of international relations, the denunciation of Western interference in the internal affairs of sovereign states and the illegitimacy of U.S. “hegemonistic” foreign policy, among other manifestations. This discourse, by definition, is a realpolitik one, and the timing and content of behavior ought to be a consistent with it.

The difficulty for the identity argument, however, is that this intensification of realpolitik discourse and behavior is relatively minor in comparison with the long-time realpolitik nature of Chinese foreign policy. There are deeper roots to Chinese realpolitik than the effort to intensify in-group identities after 1989. As I have argued elsewhere, ₁₀₅ and as ought to be clear by the failure of the Chinese to respond in the 1990s in ways predicted by the two structural realisms, it is not at all obvious that structural anarchy explains these deeper realpolitik impulses. Much more work needs to be done on the interplay between deeply historical identity creation and state responses to the contemporary identity crises that they face.

Unit-level constructivists might argue that there are unit-level ideational structures or identities that can produce realpolitik behavior through specific interpretations of the intentions of out-groups, and these are sustained through the production of histories and symbols institutionalized in education systems and decisionmaking processes that stressed hostility, rivalry, threat, and competition. I would argue that this production is inherent in in-group identity construction, and should lead constructivists to a somewhat more pessimistic view of the persistence of realpolitik behavior. Systemic constructivists would have to argue that there are global ideational structures—interpretations of actors and their intentions—that are competitive, self-regarding ones but that are produced and reproduced independent of material structures. This may be part of the Chinese story if one were to argue that the Chinese have consciously adopted an idea of major-power behavior that was normatively acceptable and descriptively standard earlier in this century, and have perpetuated this model even as other major powers have moved on to definitions of major-powerhood based on soft power attributes. In what one might call a normative product cycle, Chinese leaders have inherited “sunset” concepts of sovereignty, independence, autonomy, and security because these appeared to have worked well in producing national power and status for major powers in the past.

In either case, a constructivist focus on social interaction—namely in-group identity construction—can provide important insights into the nature of realpolitik behavior, not just into the “deviant” cases on which constructivism heretofore has concentrated. I do not think the China case has necessarily resolved the debate between ideational and material explanations of realpolitik behavior. But I hope it starts one.

 

Acknowledgments

I would like to thank the following people for their input and feedback on this project: Tom Christensen, Dale Copeland, Jeff Frieden, Ethan Kapstein, Peter Katzenstein, Mike Mastanduno, Jack Snyder and the members of the project on realism in the post-Cold War period.

---

Endnotes

Note 1: Vasquez found that most of what he classified as realist hypothesis that have been tested in the quantitative literature have accounted for little of the variance in dependent variables of interest. See John A. Vasquez The Power of Power Politics: A Critique (New Brunswick: Rutgers University Press, 1993) and his “Realism as a Degenerative Paradigm” American Political Science Review (December 1997). See also Frank Wayman and Paul Diehl, eds., Reconstructing Realpolitik (Ann Arbor: University of Michigan Press 1994) for a series of quantitative tests of realist propositions that also find that the “theory” lacks predictive and explanatory power. In his qualitative analysis of alliance behavior in modern European history, Schroeder argues that structural realist predictions do not work particularly well in accounting for the wide number of reasons for states joining or leaving alliances. Paul Schroeder, “Historical Reality vs Neo-realist Theory” International Security 19 (1) (Summer 1994): 108–48. Back.

Note 2: Andrew Nathan and Robert S. Ross, The Great Wall and the Empty Fortress: China’s Search For Security (New York: Norton, 1997); Samuel S. Kim China In and Out of the Changing World Order (Princeton: Princeton University World Order Studies Program Occasional Paper No. 21, 1991); Peter Van Ness Revolution and Chinese Foreign Policy: China’s Support for Wars of National Liberation (Berkeley: University of California Press, 1971); Gerald Segal Defending China (New York: Oxford University Press, 1995); Thomas J. Christensen, “Chinese Realpolitik” Foreign Affairs 75 (5) (September/October 1996): 37–53. Back.

Note 3: As Elman points out, Waltz seems to be of two minds on whether great powers are more or less likely to behave consistently with the dictates of the theory. See Colin Elman “Neorealist Theories of Foreign Policy: Meaning, Objections and Implications” (Paper Presented to ISA Annual Meeting, Chicago, February 1995), 27–28; and Kenneth Waltz, Theory of International Politics (Reading MA: Addison Wesley, 1979), 198–99, 194, 195). Back.

Note 4: Wayman and Diehl, Reconstructing, 26; see also Brian M. Pollins, “Cannons and Capital: The Use of Coercive Diplomacy by Major Powers in the Twentieth Century” in ibid., p. 41. Back.

Note 5: There are a number of caveats, however. First, given that the Cold War has only recently ended the number of post-test observations is quite small. So if the observations suggest the X treatment has had little or no effect on the dependent variable, it may be simply that there has not been sufficient time for an effect to show up. Second, there is the problem of identifying in which class of independent variables the “end of the Cold War” belongs. The different theories might treat the “end of the Cold War” as a change in different variables. Balance of power might view it as a change in polarity; power maximization theory as change in the opportunity to maximize capabilities; and balance of threat as a change in threat. So the fairest thing to do is to interpret the end of the Cold War as a change in that independent variable on which each theory focuses. Finally there is the potentially arbitrary nature of the dependent variable. In this chapter I focus on the dependent variables of interest to realism: military expenditures and capabilities, and alliance behavior. If the theories don’t do well under these ideal conditions, then there is little reason to believe they’d do better in explaining change in other kinds of dependent variables (e.g., behavior in economic affairs, environment diplomacy, and international institutions). Back.

Note 6: One often hears the claim that Waltz’s balancing theory is a theory of systemic outcomes and not a theory of foreign policy of individual states. Thus it is pointless to test it to explain the behavior of any one particular actor. [e.g., William Wolforth, “Realism and the End of the Cold War” International Security 19 (3) (Winter 1994/5): 101; Christopher Layne, “The Unipolar Illusion: Why New Great Powers Will Rise” International Security 17 (4) (Spring 1993): 9]. This argument is not convincing. In a number of places Waltz implies that his theory leads to certain probabilistic expectations that states are likely to act to prevent other states or coalitions from dominating the system. I have difficulty seeing why this is not a prediction about how individual states will behave. Systemic balances are the consequences of the behavior of the units, but the behavior of the units that produced this outcome is a particular type of behavior, namely balancing behavior. It is hard to see how a balancing outcome is produced if most of the major actors in the system were to refuse to balance. Indeed, the theory cannot ignore intentionality, for as Waltz himself notes, in, say, a bipolar system secondary states “if they are free to choose, flock to the weaker side” (emphasis added) (Waltz, Theory p. 127). They choose to do so because they are realpolitik actors who calculate that to preserve and strengthen themselves they should act in ways that prevent other states from dominating the system (ibid., 117). See also Waltz’s most recent prediction about Chinese balancing against the United States in Kenneth Waltz, “Reply” American Political Science Review 91 (4) (December 1997): 916. For a sophisticated argument that Waltzian neorealism is indeed a theory of foreign policy see Elman, “Neorealist Theories,” especially pp. 31–32. Back.

Note 7: Waltz, Theory, 131. Back.

Note 8: This is not so arbitrary a decision as it sounds. Sullivan has found that the correlation coefficients among different indices, including COW’s six power indicators, Ray Cline’s formula for national strength, Clifford German’s complex measure of national power, and GNP, are quite high, sufficient to make these indices interchangeable for most uses. Interestingly, he finds a strong correlation between subjective perceptions of national power rankings among states and their actual rankings using GNP and military expenditure indices. This suggests that power indicators based on military expenditures or GNP considered alone, are probably sufficient as a first crack at measuring relative power and its distribution. See Michael P. Sullivan Power in Contemporary International Politics (Columbia: South Carolina University Press, 1990), 110–17. Back.

Note 9: George Modelski, World Power Concentrations: Typology, Data, Explanatory Framework (Princeton: Princeton University Press 1974). Modelski’s categories are somewhat arbitrary but they makes intuitive sense, and they correlate with the U.S. and Soviet capabilities shared during the Cold War, a period most people accept was a bipolar one. Back.

Note 10: This calculation uses an unofficial revision of the IISS figures for China’s official budget. As a rough rule of thumb the official budget may undercount by a factor of three. Back.

Note 11: William Thompson, On Global War (Columbia: South Carolina University Press, 1988), 209–10, emphasis added. Back.

Note 12: Layne’s subjective judgment is that the United States constitutes the unipole. See Layne, “Unipolar Illusion.” Back.

Note 13: DOD (Department of Defense) (1995) East Asia Strategy Report (Washington: USGPO, February). Back.

Note 14: GAO (General Accounting Office) (1995), “National Security: Impact of China’s Military Modernization in the Pacific Region” (Report to Congressional Committees, GAO/NSIAD-95–84, June), 3. Back.

Note 15: Guan Jixian, Gao jishu jubu zhanzheng zhanyi (Campaigns in high tech limited wars), (Beijing: National Defense University Press, 1993). Back.

Note 16: Dennis Blasko, et al, “Training Tomorrow’s PLA—A Mixed Bag of Tricks” (unpublished paper, Hong Kong, October 19, 1995). p. 19. Back.

Note 17: Shambaugh reports, however, that since at least 1991, the PLA has conducted war games at the Academy of Military Sciences where the the United States has been tagged as the simulated enemy. See David Shambaugh, “The Insecurity of Security: The PLA’s Evolving Doctrine and Threat Perceptions Towards the Year 2000” Journal of Northeast Asian Studies. 13 (1) (Spring 1994): 3–25. Back.

Note 18: Ma Fajing and Dan Yuntian, “Discussion of PLA Weapons Development Problems,” Xiandai bingqi (Modern ordnance) No. 6 (June 1993) JPRS-CAR-93–075 (October 12, 1993), 35. See also Ken Allen, et al. (1995) China’s Air Force Enters the Next Century (Santa Monica: The RAND Corporation). Back.

Note 19: Blasko, “Training,” p. 22. Back.

Note 20: Chongpin Lin “Beijing and Taipei: Dialectics in Post Tiananmen Interactions” China Quarterly no. 136 (December 1993): 770–804; Chongpin Lin, “The Power Projection Forces of the Peoples Liberation Army” (Paper presented to American Enterprise Institute 6th Annual PLA Conference, Coolfont, West Virginia, June 1995). Back.

Note 21: Liao Wen-chung, “China’s Blue Waters Strategy in the 21st Century: From the First Islands Chain to the Second Islands Chain” (Taipei, Chinese Council of Advanced Policy Studies, Occasional Paper, September, 1995). See also John Garver, “China’s Push Through the South China Sea: The Interaction of Bureaucratic and National Interests” The China Quarterly No. 132 (December 1992), 999–1028. Back.

Note 22: Liao, “China’s Blue Waters.” Back.

Note 23: Shen Zhiyan, ‘’21 shiji chu Sulian zai Ya-Tai diqu de junshi zhanlue dongxiang.” [Trends in Soviet military strategy in the Asia-Pacific region in the early 21st century] Ya-Tai de xuanwo [The Asia-Pacific Vortex] (Beijing, Academy of Military Sciences Press, 1989), 43, and Hua Zhongting and Tang Cheng, “Huan he, jing zheng, wu bei” [Detente, conflict, military preparations] in ibid, 12–24. Back.

Note 24: See, for instance, the discussion of off-shore, blue-water naval operations in an internal study materials published at the National Defense University in December 1987, in Liao, “China’s Blue Waters,” appendix 4. What the document calls “near ocean defense strategy” includes operations to defend the South China Sea, reunify Taiwan with the mainland, protect scientific activities, ocean shipping, fishing industry, etc.—a list that goes well beyond traditional coastal defense strategy. Back.

Note 25: Alastair Iain Johnston, “China’s New ‘Old Thinking’: The Concept of Limited Deterrence” International Security 20 (3) (Winter 1995/6): 5–43. Back.

Note 26: Robert S. Norris and Andrew S. Burrows and Richard W. Fieldhouse, British, French and Chinese Nuclear Weapons (Boulder, CO: Westview Press 1994) p. 359. Back.

Note 27: I used the Kansas Events Data System/Protocol for the Assessment of Nonviolent Direct Action (KEDS/PANDA) data base. These events are machine-coded reports from Reuters wire service. I culled only official government statements, actions directed at government actors/targets from 1984–1995. This reduced the N substantially from 11759 to 1410 over 10 years, with all states as targets. So this final list does not reflect the totality of interactions by all actors across the two states, but does capture, hopefully, the trends in cooperative and conflictual signaling from China. For a discussion of the caveats in using KEDS/PANDA, see Doug Bond and Joe Bond, “Protocol for the Assessment of Nonviolent Direct Action (PANDA) Codebook for the P24 Data Set” (Unpublished paper, Program on Nonviolent Sanctions and Cultural Survival, Harvard University, May 1995). On average the computer miscodes about 25% of the observations. I discovered this in the small N country cases as I read through the headlines and compared these with the coding. I excluded these observations in the small N data sets, but not on the China-U.S. set. Here I assumed that given the N (360) cooperative events coded as conflictual and conflictual events coded cooperative will cancel each other out. Weighted WEIS scores as developed by Goldstein are supplied as well. See Joshua S. Goldstein, “A Conflict-Cooperation Scale for WEIS Events Data” Journal of Conflict Resolution 36 (2) (June 1992): 369–85. Back.

Note 28: Intensity refers to the mean yearly WEIS scores for Chinese actions directed at the United States As ME moves up, the scores drop, meaning the actions are coded less cooperative. The proportion refers to the percentage of all actions that are cooperative, according to weighted WEIS scales. As ME increases, the proportion of cooperative actions decreases. I used ACDA figures in 1993 dollars. If one uses official ME the correlation coefficients increase to r = -.53 and r = -.42 respectively. Back.

Note 29: I don’t provide mean WEIS scores for these countries because the N of observations is too small. The means could change dramatically with the inclusion of a few additional events in the following country cases. Instead I coded all the headlines as cooperative or conflictual by reading them, and by looking at the KEDS/PANDA event category (categories above number 372 are essentially conflictual). Admittedly this misses the more fine-tuned information available using weighted WEIS scores, but as I noted, aggregate scores are potentially too volatile with small numbers of observations. Back.

Note 30: Korea Times, November 16, 1995 as reported in North East Asia Peace and Security Network Daily Report, November 16, 1995. Back.

Note 31: See Bonnie Glaser, “The Chinese Security Perspective on Soviet/Russian-Japanese Relations” in T. Haregawa et al eds., Russia and Japan (Berkeley: University of California Press, 1993), 223; Banning Garrett and Bonnie Glaser, “Multilateral Security in the Asia-Pacific Region and its Impact on Chinese Interests: Views from Beijing” Contemporary Southeast Asia 16 (1) (June 1994): 22–23; Christensen, “Chinese Realpolitik.” This does not mean that the Chinese support any and all forms of U.S.-Japanese security cooperation. The possibility of TMD deployments in Japan are viewed with alarm, for instance. American support for a more militarized Japan, in the context of Sino-U.S. conflict, is also viewed with concern. But this is an argument for preserving relatively good relations with the United States, not competitively balancing against it. Back.

Note 32: Indeed, there is no public evidence that the joint listening posts set up in Xinjiang province in the early 1980s have been closed down, even though Sino-U.S. relations have entered unstable times of late. Back.

Note 33: Joanne Gowa, Allies, Adversaries and International Trade (Princeton: Princeton University Press, 1994). Back.

Note 34: Edward D. Mansfield and Rachel Bronson, “Alliances, Preferential Trading Arrangements, and International Trade” (Paper presented to Olin Institute National Security Seminar, Harvard University, November 1995), 2–3. Back.

Note 35: China Monthly Statistics No. 9 (1995): 35–39. Chinese statistics on bilateral trade do not count exports reprocessed and exports from Hong Kong. U.S. data do include these exports, thus U.S. figures for China’s market dependence are even higher—roughly double that of the Chinese figures. Back.

Note 36: Waltz argues that a theory should not necessarily fall if its assumptions are empirically wrong or inaccurate (Theory, 8, 119). Consistent with the spirit of his positivism, however, if the theory stands we should be able to generate additional empirical predictions that are consistent with the theory. Even if the policy discourse is an ephiphenomenal artifact of the processes that the theory suggests are at work, it ought to be an important indirect indicator of balancing behavior. That is, it ought to be rare to find a deep discourse that does not stress balancing when a state is indeed balancing. Nonstructuralists would argue, of course, that the discourse reflects an intentionality that has a causal effect on state behavior and, in turn, on systemic outcomes. Either way, I would argue, H.A3 is an important hypothesis to test. Back.

Note 37: He Fang,”Shijie geju yu guoji xingshi” [Global system and the international situation] Shi jie jingji yu zhengzhi [World economics and politics] No. 11 (1991). As Ross has convincingly shown, this article signaled not only the “hegemony” of the unipolarity argument, but also a period of Chinese concessions to the United States on MTCR, trade reform, and prison labor. This is not entirely consistent with balancing against the dominant player in the international system. Robert S. Ross, “Two Level Games and Sino-U.S. Bargaining” (Unpublished paper, Harvard University, 1995). Back.

Note 38: Du Gong, ed., Zhuanhuan zhong de shijie geju [The world structure in transition] (Beijing: World Affairs Press 1992), 6–7. Back.

Note 39: Yan Xuetong et al., “Dangqian wo guo waijiao mianlin de tiaozhan he renwu” [The challenges and tasks that our country’s foreign policy currently faces] Shi jie jingji yu zhengzhi [World economics and politics] No. 24 [1993]). Back.

Note 40: Glaser, “Chinese Security Perspective,” 225–26. My impression from discussions with specialists in Beijing in the spring of 1996 is that there is still considerable debate in the Chinese leadership circles on the nature of American hegemony. Back.

Note 41: Wang Jianwei, United States-China Mutual Images in the Post-Tiananmen Era: A Regression or Sophistication? (Ann Arbor: University of Michigan, Political Science Department, PhD diss. 1994), 268–69. Wang’s intensive interviews with 127 members of Chinese foreign policy elite and foreign policy analysts, businessmen, and scholars from June 1991 to March 1992, before and after the collapse of the USSR, revealed a general agreement that the United States was among other descriptors, “all-out champion, first class military power, the sole superpower, number one in the world” (p. 263 note 14). Conservatives, of course, do see unipolarity as potentially threatening to China’s economic, political interests and internal cohesion. See Wang Chiming’s concern that the United States is moving to encircle China strategically in Asia, as exemplified by the U.S. interest in multilateral security institutions in East Asia that might “lasso” (tao zhu) China. Wang Chiming, “Shixi wo guo zhoubian huanjing zhong de ‘Meiguo yinsu’ ” [Preliminary analysis of the ‘American factor’ in the environment around our country’s periphery] Shi jie jingji yu zhengzhi [World economics and politics] No.1 (1994): 56–61. As of late 1997—there was still some debate in the leadership whether, even though the United States is the only superpower, it is on the decline (and therefore should be challenged) or is increasingly dominant (and therefore should be accommodated on some issues). The difficulty of accommodation on most of the major irritants in the relationship is due to domestic political constraints—the relatively weaker post-Deng leadership—not a consensus view that U.S. power is on the decline. Back.

Note 42: Layne, “Unipolar Illusion,” p. 3, fn.1; p. 9. Back.

Note 43: John Mearsheimer, “Back to the Future: Instability in Europe After the Cold War” International Security 15 (1) (Summer 1990): 12; John Mearsheimer, “The False Promise of International Institutions” International Security 19 (3) (Winter 1994/5): 11–12. Back.

Note 44: Morgenthau, cited in Bruce Bueno de Mesquita and David Lalman “Power Relationships, Democratic Constraints, and War” in Wayman and Diehl, Reconstructing p. 170. Back.

Note 45: Joseph Grieco, “Anarchy and the Limits of Cooperation: A Realist Critique of the Newest Liberal Institutionalism” International Organization 42 (3) (Summer 1988): 499. Back.

Note 46: The problem for power maximizing realism arises, as Glaser’s work implies, when defense has a decisive and long-term advantage over offense. If defense dominates to this extent, then no amount of effort to improve relative capabilities will lead to an offensive advantage. Moreover, if defense dominates, a power maximizing state ought still to feel secure enough that it need not drive for offensive advantage. Thus under conditions of nuclear deterrence, assured second strike deterrence, even power maximizing states should ease up in their effort to achieve decisive advantage. See Charles Glaser, “Realists as Optimists: Cooperation as Self-Help” International Security 19 (3) (Winter 1994/5): 50–90. If a state continues to maximize relative power even when defense dominates, this suggests there is some other, nonstructural basis for its dissatisfaction. We then need to term to the unit-level sources of dissatisfaction, as Schweller suggests we do. Back.

Note 47: It is not clear how power maximization realism would explain why these states would be unwilling, short of their collapse, if they too are power maximizers without making ad hoc assumptions about the characteristics of the units. Back.

Note 48: Layne, “Unipolar Illusion”. Back.

Note 49: Robert Gilpin, War and Change in World Politics (Cambridge: Cambridge University Press, 1991), 94–5. Back.

Note 50: Zhao Huaipu and Lu Yang, “Quanli zhengzhi yu xianghu yicun” [Power politics and interdependence] in Shijie jingji yu zhengzhi, [World economics and politics ] No.7 (1993): 36–41. See also He Xin’s description of the international economic system as one populated by developed country “wolves” and developing state “sheep” whereby policies that are advantageous for the former are not similarly advantageous for the latter. Evidently there are few opportunities for joint gains. See He Xin, “Wang Zhen tongzhi tanhua jiyao” [Record of key points in conversation with Comrade Wang Zhen” (1991.4.17) in He Xin (1993) He Xin Zhengzhi Jingji Lunwen Ji—Nei bu yanjiu baogao [Collected political and economic writings of He Xin—internal circulation research reports] (Harbin, Heilongjiang Education Press 1993) p. 325. See also the discussion of the zero-sum nature of competition for ocean resources,Yang Zhiqun, “Officer’s Forum—Brown Water, Blue Water—Thoughts on Naval Theory” Jianchuan Zhishi [Naval and Merchant Ships] (February 1994) JPRS-CAR-94–031 (May 13, 1994): 34. Back.

Note 51: This figure shows the average frequency of the terms “new world order” and “new international order” used in a randomly sampled selection of articles from World Economics and Politics. The drop in the former after 1991 is due to a conscious decision to critique the term after its prominence in the global discourse on international relations after the Gulf War. See below for a discussion of the coding procedures. Back.

Note 52: For a similar reason, China’s leaders do not like the term “new thinking” (xin siwei), as it is too closely identified with Gorbachev’s destruction of communist rule in the Soviet Union. Back.

Note 53: Li Shisheng, “Guanyu guoji xin zhixu ji ge wenti de tan tao.” [Discussion of several questions relating to the new international order] Shijie jingji yu zhengzhi. [World Economics and politics] No. 10. (1992): 43–44. Back.

Note 54: Ma and Dan, “Discussion of PLA” p. 34; see also Wang Pufeng et al, eds., Xiandai guofang lun [On modern national defense] (Chongqing: Chongqing Publishing House, 1993), 67. Back.

Note 55: Zhang Zhaozhong and Guo Xiangxing, Xiandai haijun qishi lu [A record of the inspiring modernization of the navy] (Beijing: People’s Liberation Army Press, 1993), 26–27. Back.

Note 56: Shen, “21 shiji chu” p. 43. Back.

Note 57: Mao noted in 1958 that if China didn’t have nuclear weapons, it would “count for nothing” (bu suan shu). Huang Cisheng and Wang Lincong, “Shilun Mao Zedong de he zhanlue sixiang” [Preliminary discussion of Mao Zedong’s thinking on nuclear strategy] in Quan jun Mao Zedong junshi sixiang xueshu taolun wen jing xuan [Selected essays from the all-Army academic meeting on Mao Zedong’s military thought] (Beijing: Academy of Military Sciences Press, 1992), 602. See also his comments just before the Great Leap Forward and the second Quemoy-Matzu crisis in 1958 that China had to increase industrial production and military power so that John Foster Dulles would sit up and take notice of China’s stature. See Thomas J. Christensen, Useful Adversaries (Princeton: Princeton University Press, 1996). Back.

Note 58: Gu Yan, “Duli zizhu shi Mao Zedong waijiao sixiang de linghun” [Independence and self-reliance is the spirit of Mao Zedong’s foreign policy thought] Shijie jingji yu zhengzhi. [World Economics and politics] (1994): 32. Position in this context is hierarchical, implying rank and status. Minzu is a hard-to-translate term that in its usage in the 1980s and 1990s blurs the division among nation, ethnicity, and race. Back.

Note 59: For a definition of this benign environment purely in security terms see Yan, “Dangqian” p. 22. For a definition that implies improvements in China’s status see Gu, “Duli Zizhu.” Back.

Note 60: Cited in Wan Ming, “Yuan Diplomacy: China’s Economic Statecraft since 1978” (Unpublished paper, Center for International Affairs, May 1994), 31 fn. 90. Back.

Note 61: Carol Hamrin, China and the Challenge of the Future: Changing Politics Patterns (Boulder: Westview Press, 1990) p. 124. Back.

Note 62: The argument here agrees substantially with Samuel Kim’s speculation that China’s military buildup has less to do with balancing against threats or hegemons and more to do with prestige and status. See his “China in the World in Theory and Practice.” in Samuel S. Kim ed., China and the World (Boulder, CO: Westview Press 1994), 13–14. Back.

Note 63: These arguments were first made most forcefully in an article in Jiefangjun Bao (Liberation Army Daily) in 1987. The purpose of the development of “three dimensional” military capabilities was to “protect China’s legitimate rights and interests”—no mention of security threats here. See Xu Guangyu “Pursuit of Equitable Three Dimensional Strategic Boundaries” Jiefangjun Bao (April 3, 1987) in JPRS-CAR-88–016 (March 29, 1988): 38. The status issue was also prominent in Lt. Cmdr. Yang Zhiqun’s praise of blue-water navies. “When a country has the humiliation of being chased out of an ocean, today when it has an extremely great oceanic interest, its navy will patrol the blue water.” His meaning—China’s navy needs to restore control over those blue water areas that rightfully belong to China. See Yang, “Officer’s Forum,” 34. Back.

Note 64: See “Curriculum Research Materials: General summary of discussions on building the strategic theory of the airforce and navy” (National Defense University, December 12, 1987) in Liao, “China’s Blue Waters,” 18. Back.

Note 65: Zhang and Guo, Xiandai haijun p. 27. Back.

Note 66: Randall Schweller, “Bandwagoning for Profit: Bringing the Revisionist State Back In” International Security 19 (1) (Summer 1994): 72–107; and Schweller “Neorealism’s Status-Quo Bias: What Security Dilemma?” Security Studies 5 (3) (Spring 1996): 225–58. Back.

Note 67: The term was suggested by Tom Christensen, personal correspondence, May 1996. Back.

Note 68: Stephen Walt,The Origins of Alliances (Ithaca: Cornell University Press 1987), 26. For instance, the theory would predict that, given a geographically proximate state with overwhelming military power and an offensive doctrine, Canada ought to be balancing against the United States On the other hand, Walt’s theory could also argue that this absence of balancing is not an anomoly because perceptions of intent suggest Canada need not worry. In this case the threat perception variable would clearly overwhelm the effects of the other three variables in Walt’s theory. In particular, this case would seem to be an instance of ideological solidarity (illustrated by the fact that the end of the Soviet threat has not led Canada to worry more about an American threat). Ideological solidarity, in this case, is a function of shared political and cultural traits, in other words shared democratic identities. Back.

Note 69: See Walt’s description of Nasser’s fears about the U.S. threat, in ibid., p. 69. Back.

Note 70: Though if this is so, then it suggests that the balancing coalition at time t is violating one of neorealism’s assumptions that states will not act to increase relative capabilities to the point where others form counterbalancing coalitions and thus threaten its security at time t + 1. If this axiom did hold for the balancer at time t, then it should have held as well for the state that sought hegemony at time t - 1. If it had held, then it should have been prudent enough not to seek hegemony at t - 1, and therefore it should not have provoked balancing at time t, which, in turn should not have provoked counterbalancing at t + 1. As Schweller (and Wendt) points out, the process of balancing and counterbalancing requires the presence of a predator state that seeks power regardless of the cost to security. See Schweller, “Bandwagoning”; Schweller, “Neorealism’s Status Quo Bias”; and Alexander Wendt, “Anarchy is What States Make of It: The Social Construction of Power Politics” International Organization 46. (1992): 391–425; Alexander Wendt, “Collective Identity Formation and the International State” American Political Science Review 88 (2) (June 1994): 384–96. Back.

Note 71: This assumes states are not power maximizers and do not take advantage of opportunities to maximize relative capabilities once they are secure. Walt is not all that explicit about whether states are “defensive positionalist” or power maximizers, though he appears to lean toward the former. See Walt, Origins, p. 9, where he notes that states balance when their security position is threatened. Back.

Note 72: Zong Lanhai, “Zhong gong yiding guoji tou hao diren” [The Chinese communists determine their main international adversary] Cheng Ming No.195 (January 1994): 16–18. Back.

Note 73: Wang Jianwei, “Coping with China as a Rising Power (New York: Council on Foreign Relations, Asian Project Working Paper, July 1995), 7; Wang Jisi, “U.S. Policy Toward Taiwan and Sino-U.S. Relations” Wen Wei Po (Hong Kong) (August 19, 1995) in FBIS-CHI-95–177 (September 13, 1995): 6–7; Wu Peng, “ ‘Zhongguo weixie’ lun pouxi” [An analysis of the “China threat” theory] Shijie jingji yu zhengzhi (World economics and politics) No. 10 (1993): 3. Back.

Note 74: Song Baoxian, “A Great Debate on U.S. Policy Toward China” Wen Wei Po (Hong Kong) (August 19, 1995) in FBIS-CHI-95–177 (September 13, 1995): 7–8; Kuan Wen-liang, “Pro-Taiwan U.S. Congressmen Abandon Elections One After Another—On Absence of Anti-China Foundation in U.S. Society” Wen Wei Po (Hong Kong) (September 9) in FBIS-CHI-95–181 (September 19, 1995): 15–17; Yuan Ming and Fan Shiming, “ ‘Leng zhan’ hou Meiguo dui Zhongguo (an quan) xingxiang de renshi” [China’s security role in post-Cold War American security perceptions] Meiguo yanjiu [American Studies] 9 (4) (Winter 1995): 7–29. Back.

Note 75: Wang, “Shixi wo guo,” 56–57. Back.

Note 76: Ibid., p. 60. Back.

Note 77: This is potentially a dodgy prediction. It is not clear in realist theories whether the relationship between threat and military expenditures ought to be linear or stepwise. On the one hand, one could argue that, given the role of uncertainty in producing self-help behavior, a decline in threat (measured objectively) should not produce a monotonic decline in balancing behavior. Rather the absence or virtual absence of threat may be necessary for a reduction in expenditures to begin. Measured subjectively, of course, a decline in threat is a decline in uncertainty about likelihood of attack, and thus should be linearly related to a decline in expenditure. Back.

Note 78: Yan, “Dangqian,” 22; Wang, “Shixi wo guo,” 56. Back.

Note 79: Marc Ross, The Culture of Conflict: Interpretations and Interests in Comparative Perspective (New Haven: Yale University Press 1993; Jonathan Mercer, “Anarchy and Identity.” International Organization 49 (2) (Spring 1995): 229–52; Daniel Druckman, “Nationalism, Patriotism, and Group Loyalty: A Social Psychological Perspective,” Mershon International Studies Review 38 (Supplement 1) (1994); Eugene Burnstein et al “How the Mind Preserves the Image of the Enemy: The Mnemonics of Soviet-American Relations” in William Zimmerman and Harold Jacobson eds., Behavior, Culture and Conflict in World Politics (Ann Arbor, University of Michigan Press, 1993); Chester A. Insko et al., “Individual-Group Discontinuity as a Function of Fear and Greed” Journal of Personality and Social Psychology 58 (1) (1990): 68–79. Back.

Note 80: It sometimes pays individually to call oneself a member of a group. But the attractiveness of group identification is not entirely self-interestedly material, nor are the sanctions against violating group identity entirely material. Experimental research has found that perception of shared identity leads to an “empathetic altruism” where the interests of others in the group are taken to be one’s own interests. (See John C. Turner, Rediscovering the Social Group: A Self-Categorization Theory (London and New York: Basil Blackwell, 1987), p. 65. And the source of shared identity is not necessarily interpersonal interest or interdependence, but perceived interpersonal similarity according to some socialized criteria of prototypicality. Nor are the symbolic cues that trigger identity-enforcing or -dissolving behavior uniformally material ones. Beyond this, however, there are a number of hypothesis about out-group devaluation. One possibility is that in-group identity creation by definition requires a devalued out-group, so as to reveal exactly why the in-group ought to exist. This opens the door to attribution effects and justifications for violence. Another possibility is that political entrepreneurs have an interest in accentuating in-group–out-group differences so as to consolidate their leadership. Still another hypothesis is that individuals have a psychological need to enhance their self-worth and status by belonging to a valued group. Finally, the devaluation of the out-group may increase the efficiency of the in-group in the production of benefits through coordination. Back.

Note 81: Donald R. Kinder and David O. Sears, “Prejudice and Politics: Symbolic Racism versus Racial Threats to the Good Life” Journal of Personality and Social Psychology 40 (3) (1981): 414–31. The mere anticipation of interacting with an out-group can lead to more competitive interpretations of the out-group’s likely behavior, even though the structural conditions of interaction with this group are not yet clear. See Willem Doise, Groups and Individuals (Cambridge: Cambridge University Press, 1978), 118. Back.

Note 82: An identity theory of realpolitik is important for at least three reasons. First, since the findings on the relationship between group formation and in-group–out-group competition in the experimental social psychology and cultural anthropology literature are so robust, it makes sense to see if these have implications for interstate relationships. Second, an identity theory of realpolitik introduces a third overarching category of realism by placing the origins of self-help behavior in the characteristics of group formation and membership, placing the theory between neorealism’s focus on the the causal attributes of system structure and classical realism’s stress on human nature. Finally, an identity realism is one potential route for a constructivist contribution to explaining realpolitik. If constructivists are to challenge materialist realism they will have to explain both the so-called deviant cases and the allegedly nondeviant ones. Back.

Note 83: This leads to two conclusions. First, as the intensity of identity varies so must the degree to which the other is devalued, alienized and thus feared. This means uncertainty varies. If uncertainty varies while technical anarchy persists, then the causal connection between structural anarchy and uncertainty/fear is severed. Thus the scope and intensity of identity of the in-group can expand or contract while anarchy remains constant. Second, it is possible for a decentralized community identity to form without the complete subsumption of all actors into a bigger group. This means that, technically, anarchy can exist among independent actors who share a group identity. Thus Western European states can operate under conditions of anarchy, while sharing identities as members of a democratic European community. These shared identities lead to the interpretation of changing relative power capabilities in ways that structural realism would be unable to predict. Again this suggests that anarchy and self-help behavior are causally disconnected. It is precisely because the boundaries of identity communities can move, that it becomes possible to create larger or smaller pockets of low-competition relationships. This means, in principle, that if in-group identity were global then no intergroup relationship would be affected by anarchy. Back.

Note 84: Glaser, “Realists as Optimists.” Back.

Note 85: Russell Hardin, One For All: The Logic of Group Conflict (Princeton: Princeton University Press, 1995), 74–75. Back.

Note 86: This raises the possibility that identity discourses are used instrumentally by political leaders to buttress their political power, and thus ought properly to be seen as elements in strategic “diversionary” behavior by elites. There is no doubt this is often the case. But leaders can chose only those discourses that find resonance among relevant constituencies, that is, in groups where these are deeply internalized. Race baiting may be used instrumentally by Republicans in elections, but as Kinder and Sears indicate, this works only among whites who have internalized “symbolic racist” values. Race baiting is also likely to be used by those who, ironically, devalue the importance of racism as a central feature of American society. Thus the decision to chose a particular political instrumentality is constrained both by the values of the political entrepreneur and by the values of the target population. Instrumentality does not, therefore, reduce the causal importance of identity. Back.

Note 87: Ross, The Culture of Conflict, 176. Back.

Note 88: This tendency of groups to institutionalize a realpolitik world view is reinforced by the cost of acquiring and processing information that suggests the world is not so competitive. Since information is costly, the most authoritative sources will be from within the group. People are much more easily convinced by arguments coming from those within the in-group, since there is a prior expectation that one is more likely to agree with their views (Turner, Rediscovering, 100, 154). This creates a vicious circle of interpretation of new data that reinforces realpolitik beliefs. Put in Bayesian updating terms, prior beliefs about the realpolitik nature of the environment are reinforced by the a realpolitik likelihood function. Thus new data—new events whose probability of occurring might suggest that a less competitive state of the world exists—is interpreted as confirming rather than challenging priors. Posterior beliefs do not change. Identity realism would argue that this realpolitik likelihood function is a product of group identity formation, not structural anarchy. On Bayesian updating and the parallels between rational choicers’ concept of the likelihood function and constructivists’ concept of ideology see Robert H. Bates and Barry R. Weingast, “A New Comparative Politics: Integrating Rational Choice and Interpretivist Perspectives” (Cambridge: Center for International Affairs, Working Paper No. 95–3, April 1995): 35–36. Back.

Note 89: See David Campbell,Writing Security: United States Foreign Policy and the Politics of Identity (Minneapolis: University of Minnesota Press, 1992), 61–68. On the role of images of external danger and disorder as mechanisms of control—specifically the delegitimization of non-realpolitik views of strategy in policy discourses—see also Robin Luckham, “Armament Culture,” Alternatives 10 (1) (1984), and Bradley S. Klein, “Hegemony and Strategic Culture: American Power Projection and Alliance Defence Politics,” Review of International Studies 14 (1988). On the relationship between identity creation, control, and disorder see Harrison White, Identity and Control: A Structural Theory of Social Action (Princeton: Princeton University Press, 1992). Back.

Note 90: One indicator, then, of the intensity of in-group identification would be the discourse used to describe marginals—the more intensive in-group identity is the less acceptable multiple identities are in the discourse. Back.

Note 91: Osgood semantic differentiation asks a respondent to place her/himself along a seven-point scale between two binary opposite terms used to describe potential attitudes toward a particular item. A meta-contrast ratio refers to the “ratio of the average difference perceived between in-group and out-group members over the average difference perceived between in-group members.” The smaller the denominator (e.g. the more closely in-group values converge), and the larger the numerator (e.g. the more divergences between group values) the greater the ratio, or the higher the measure of intensity of identification. See Margaret Wetherall, “Social Identity and Group Polarization” in Turner, Rediscovering, 155. Back.

Note 92: Glenn Chafetz has tried to measure the degree of identification between groups, essentially a measure of affect, by coding public statements about other states on a nine-point positive-negative scale. See Glenn Chafetz, “An Empirical Analysis of International Identity Change” (Paper presented to American Political Science Association Annual Meeting, Washington, D.C., August 28–31, 1997). It is not clear, however, that this measure captures deeply internalized, “taken-for-granted” assessments of other actors, or whether it simply measures the current state of amity-enmity in an interstate relationshup. Back.

Note 93: Social psychological work on group identity and intergroup conflict has shown that the preemptive devaluation of out-groups protects the sense of uniqueness of the in-group. Thus the intensity of group discourses that stress its uniqueness would also seem to be a logical indicator of in-group solidarity and out-group devaluation. Back.

Note 94: This works because of the intense socialization in particular interpretations of modern history. As one example, the first of a recently produced two-volume set on modern Chinese history written for Chinese youth is entitled, National Shame. It lists more than eighty instances of shame and insult to China, almost all of which were perpetrated by foreigners or Chinese working on behalf of foreign interests (the companion volume is entitled, National Prowess). See Che Jixin, ed., Guochi [National shame] (Jinan: Shandong Friendship Publishing House, 1992). The victimization has resonance even among Western-educated intellectuals. See for example, Wang, United States-China, p. 267. The regime also expertly exploited the U.S. Congressional opposition to China’s bid for the 2000 Olympics—the United States denied China a “chance” to show off its national spirit and energy—resulting in a surge of anti-americanism even among intellectual critics of the regime. See Jeremy Barme,”To Screw Foreigners is Patriotic: China’s Avante-Garde Nationalists” The China Journal No.34 (July 1995): 214, and Wang, “Coping with China,” 8. Back.

Note 95: See, for instance, the paean to traditional Chinese Confucian cultural values in a lengthy article on political education in the military written by Minister of Defense Chi Haotian. See “Build An Ideological Great Wall to Resist Corruption and Prevent Degeneration” Renmin Ribao (People’s Daily) (August 10, 1995) in FBIS-CHI-95–188 (September 28, 1995): 30–31. See also a recent book on China’s traditional ethics issued by the State Education Commission, Luo Guojie, ed., Zhongguo chuantong daode [China’s traditional morality] (Beijing, People’s University Press, 1995). Back.

Note 96: The racialist elements of the Chinese nationalist discourse is an extremely sensitive and difficult-to-handle issue, but too important to ignore. There is ambiguity as to which Chinese China’s leaders refer to when invoking nationalism—ethnic Chinese, citizens of the PRC? Who is it that must “stand up” and take their place in the sun? Conservative nationalists have referred to the threat from “white culture.” They will condemn those who are perceived to damage the national interest, like dissidents, as race traitors (han jian) (see for instance He Xin, “Weilai Zhong Mei guanxi” [Sino-U.S. relations in the future] (September 7, 1992) in He, He Xin zhengzhi jingji, 125). Conservative nationalists will distinguish between conflict based on socialist-capitalist ideology and issues like human rights on the one hand, and clashes of national interest on the other. However, the boundaries between racial-national interest and state interest are blurred. He Xin has characterized the conflict with the United States—namely American efforts to “peacefully evolve China”—as a conflict of “national-ethnic interests” (minzu li yi], not ideology (nor state interests— guojia li yi). See ibid., and He Xin, “Guanyu ‘zuo’ ji yishi xingtai gaige wenti de sikao” [Thoughts on the problem of the ‘left’ and ideological reform] (July 2, 1992) in ibid., p. 136. While He Xin is personally not so influential as he was in the late 1980s and early 1990s—he was close to Premier Li Peng for a time—his views are probably fairly representative of conservative nationalist voices in Beijing. On the racialism in modern Chinese nationalism see Frank Dikotter, The Discourse of Race in Modern China. (Stanford: Stanford University Press, 1992). Back.

Note 97: Andrew Nathan, “Human Rights in Chinese Foreign Policy” The China Quarterly No. 139 (September 1994): 622–43; Chen Jie, “Human Rights: ASEAN’s New Importance to China.” Pacific Review. 6 (3) (1993). Back.

Note 98: Wei Cizhu, “Zou chu aiguozhuyi jiaoyu ‘wuqu’ de duice silu” [Countermeasures for the road out of “error zones” in patriotic education], Neibu wengao [Internal manuscripts] No. 21 (1990): 8–9). See also Chen Xiushan, “Qutong lun pingxi” [A critique of convergence theory], Neibu wengao [Internal Manuscripts] No. 19 (1989): 5–8. Back.

Note 99: Hua Qing, “Jiekai minzuzhuyi shenmi de mianmao” [Open up the mysterious face of nationalism] Shijie jingji yu zhangzhi [World economics and politics] No. 7 (1995): 69–71). The more negatively described ones included the manufacturing of national hatreds and irrational ethno-national passions. There were implicitly distinguished from more progressive uses, as in the Chinese case. Back.

Note 100: On increase in the autonomous construction of ethnic identities inside China, the self-outing of people who previously saw themselves as Han Chinese, see Dru Gladney, “China’s Ethnic Awakening” Asia Pacific Issues (East West Center) No. 18 (January 1995); Yitzhak Shichor, “Separatism: Sino-Muslim Conflict in Xinjiang,” Pacifica Review 6 (2) (1994): 71–82. For a high-level expression of the dangers of Muslim separatism, fueled by Islamic revival worldwide, see the remarks by then-Vice President Wang Zhen in He Xin, “Wang Zhen tongzhi tanhua jiyao” [Record of key points in conversation with Comrade Wang Zhen” (April 17, 1991) in He, He Xin zhengzhi jingji. Back.

Note 101: Arguably one found in the mid 1980s, when the military challenge to China (from the USSR and Vietnam) was greater than the 1990s, and when China’s integration into economic institutions was lower, the discourse was less shrilly realpolitik. Contrast, for instance, Zhao Ziyang’s discussion of foreign policy in his report to the 13th Party Congress in 1987 with Jiang Zemin’s report at the 14th Party Congress of 1992. Zhao stressed disarmament and development themes, while Jiang’s was replete with references to preserving state sovereignty and independence. One was also more likely to find somewhat more sympathetic discussions of multilateral security institutions (e.g. summits, CBMs, crisis management centers, etc.) in the mid 1980s than one finds now. See for instance, Yu Shaoqiu, “Xi fang guoji chongtu yu weiji de kongzhi lilun pingshu” [Critique of Western theories on international conflict and the control of crises] in Shijie jingji yu zhengzhi. [World economics and politics] No. 7 (1986). Three of the realisms tested here would have a difficult time explaining the shift in discourse over the 1980s and 1990s. Identity realism’s prediction that in the face of a severe legitimacy crisis the state will make greater efforts to intensify in-group identities is consistent with this shift in discourse. Back.

Note 102: He, “Weilai Zhong Mei,” 126; He Xin, “Qianfu weiji yu qianjing yuce—dangqian Zhongguo nei wai xingshi de yanjiu baogao.” [Hidden crisis and prospective predictions—research report on the current internal and external situation] (September 25, 1989) in He, He Xin Zhengzhi jingji. Note this is similar to the rhetorical strategy of invoking Croatian fascism among Serbs: even though the probability of extermination is low, it is greater than zero, given modern history in China. Thus the effects of American strategy would be disastrous for China. This expected utility, manufactured through the invocation of historically rooted symbols, then rallies support for regime. Bates and Weingast, “A New Comparative Politics.” Back.

Note 103: Although this has been the official journal of Institute of World Economics and Politics at the Chinese Academy of Social Sciences, in practice it has been a forum for policy-relevant scholarship on IR topics from specialists from a wide range of institutes including the military, the Foreign Ministry, and the intelligence community. Thus, it has some degree of face validity as a source of authoritative thinking about IR and China’s role in the world. The procedures were kept simple. Ten articles were randomly selected from each of seven years spanning the 1989 and 1991 divides. All references to the eleven terms, including pronouns, were counted for each article (N = 1772). The number of references to each term in each article was then averaged, giving a yearly mean. The chart represents the cumulative sum of these yearly means. The terms were clustered into roughly opposing cooperative and conflictual concept categories. The cooperative terms were: cooperation (hezuo), interdependence (xianghu yicun/yilai); international interest (guoji liyi), new world order (shijie xin zhixu); and international organizations (guji zuzhi]. The terms signifying conflict were: power politics/hegemonism (qiangquan zhengzhi/baquanzhuyi); conflict/contradictions (douzheng, maodun); sovereignty (zhuquan); independence (duli); national interest (guojialiyi); and new international order (guojixinzhixu). I thank Robert Ross for lending me his copies of the journal, and Fu Jun and Qiu Dong for doing the coding work. We were, unfortunately, unable to find complete runs for 1987, 1990, 1993 and 1994. Back.

Note 104: In preciously few instances if one accepts in combination the very different work of Vasquez, Schweller, and Schroeder. Back.

Note 105: Alastair Iain Johnston, “Cultural Realism and Strategy in Maoist China” in Peter Katzenstein ed., The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996). Changing the last endnote: Back.

Unipolar Politics: Realism and State Strategies After the Cold War