3 DOMESTIC MODELS OF POLITICAL SURVIVAL
Why Some Do and Others Don't (Proliferate)
Etel Solingen
t
at the first meeting of the International Commission on Nuclear Nonprolif-eration and Disarmament in October 2008, the cochairman—former Australian Foreign Minister Gareth Evans—warned that nuclear weapons could be on the verge of spreading to many new states. "We are on the brink," he said, "of... an avalanche or a cascade of proliferation unless we are very, very careful indeed and find ways collectively to hold the line."1 At the same meeting, former U.S. Defense Secretary William Perry argued that "if we fail to deal effectively with [Iran and North Korea] I think we are facing a veritable cascade of nuclear proliferation." Statements of this sort have become common across the U.S. domestic and international political spectra. Indeed, a September 2008 bipartisan report maintains that "given historical instability in the region, the prospects of a nuclear Middle East—possibly Iran, Saudi Arabia, Egypt, and Turkey—are worrying enough, even before the proliferation cascade continues across North Africa and into Southern Europe."2
Increased concern with a proliferation wave, contagion, or epidemic compels a proper understanding of what exactly drives states to develop or acquire nuclear weapons. Various logics can explain the range of decisions associated with acquiring or abstaining from nuclear weapons.3 As an early study by Meyer suggested, it is quite likely that some assumptions from different logics are valid; the task is identifying when and why different logics apply.4 However, the same study reminded that all motives of nuclear behavior are, in the end, filtered through the domestic politics within which decisions are made. Yet, a systematic understanding of domestic effects has eluded most work in the area of nuclear proliferation in the last three decades. Country studies
38
DOMESTIC MODELS OF POLITICAL SURVIVAL 39
of nuclear aspirants often provide ad hoc depictions of the domestic scene, but these are often drawn without reference to an overarching comparative framework. Cleavages between "moderates" and "hard-liners" are often identified on the basis of inductive "who's who" analyses, lacking an underlying logic for what makes leaders, institutions, and ruling coalitions (or their opponents) "moderates" or "hard-liners."5
Efforts to understand the deep personal, philosophical, or normative sources for moderate or hard-line nuclear choices may sometimes be helpful— and even represent progressive moves relative to earlier fixations with structural neorealism—but also entail an open-ended and protracted enterprise that must confront severe methodological obstacles and holds unknown universal applicability.6 The pressing policy relevance of nuclear proliferation highlights the value of more discrete markers, shortcuts, or rules of thumb that might help identify the motivations of leaders, their ruling coalitions and opponents, institutions, and relevant constituencies. This chapter focuses on a particular conceptual framework that provides, among other things, one way of making the question of "Who is likely to be a moderate or a hardliner and why?" an integral part of the explanation. That is, those proclivities are not explained by some extraneous theory but are, instead, endogenous to the general argument that I develop in the first section. This argument is designed to address the broader issue of which states are more likely to pursue nuclear weapons and why. At its heart is the contention that different models of domestic political survival—how leaders seek to gain and maintain power— provide important information regarding nuclear decisions. The second section elaborates the scope and conditions under which the argument is expected to hold. The third section applies the general argument to the issue of proliferation chains and the conditions under which they might be more or less likely. This section thus builds on the proposed framework outlined in the preceding two sections to identify a coherent set of scenarios. The final section outlines how evolving models of political survival may portend continuity and change in nuclear trajectories.
THE ANALYTICAL ADVANTAGES OF DOMESTIC MODELS OF POLITICAL SURVIVAL
The argument that leaders and ruling coalitions adopt different domestic models of political survival that are consequential for nuclear decisions was designed to address second tier, or second nuclear age, nuclear aspirants, whose
40   ETEL SOLINGEN
decisions to launch or abandon quests for nuclear weapons were shaped under a particular global "world time" in place since the 1960s.7 Negotiations that crystallized in the Nuclear Non-Proliferation Treaty (NPT) in the 1960s and developments toward a globalizing political economy created this new "world time" under which decisions to consider, pursue, or abandon nuclear weapons were made.8 These decisions were inextricably linked to the models of political survival adopted by leaders and ruling coalitions to gain and maintain power. In particular, these models entailed different orientations to the global political economy and its associated economic, political, and security institutions and had different implications for nuclear choices.
On the one hand, leaders advocating economic growth through integration in the global economy ("internationalizing" models henceforth) had incentives to avoid the costs of embarking on nuclear weapons programs. Leaders vary in their tolerance for domestic, international, political, and economic (including opportunity) costs entailed by nuclear weapons. What specific aspects of models emphasizing economic growth and openness to the global economy as tools of political survival made leaders more receptive to denuclearization than others?9 The answer lies in a range of incentives and disincentives, including: (a) the need to appeal to foreign investors with an interest in domestic economic growth and stability; (b) the related need to reassure neighbors to preserve regional cooperation, stability, and attractiveness to international economic actors; (c) the requirement of securing access to international markets for exports, capital, technology, and raw materials; (d) the related aversion to risking reputational losses at home and abroad for uncertain nuclear gains; and (e) the costs of alienating domestic agents of internationalization— both within and outside state structures—that would be adversely affected by nuclear weapons development. Clearly, there are several causal pathways linking nuclear weapons' renunciation to models emphasizing economic growth through global integration. Nuclearization burdens efforts to enhance exports, economic competitiveness, macroeconomic and political stability, and global access—all objectives of internationalizing models—while strengthening state bureaucracies and industrial complexes opposed to economic transformation. Denuclearization thus often took place as part of a broader program of internationalization designed to strengthen market-oriented forces, leaders, and institutions—state and private—favoring export-led growth.
On the other hand, leaders relying on inward-looking bases of support had greater tolerance—and in some cases strong incentives—for developing
DOMESTIC MODELS OF POLITICAL SURVIVAL 41
nuclear weapons. Nuclearization entailed fewer costs for inward-looking leaders and coalitions whose political platforms were rooted in mistrust for international markets, investment, technology, and institutions. Such coalitions protected uncompetitive national industries, sprawling state enterprises, and ancillary military-industrial and nuclear complexes.10 Nuclearization entailed considerable domestic advantages for foes of internationalizing models in inward-looking, import-substituting regimes favoring extreme nationalism, religious radicalism, or other forms of autarky, such as North Korea's juche (national self-sufficiency). Such leaders and their political allies often relied on extreme language to compel and threaten regional adversaries, wielding potential nuclear and other weapons of mass destruction (WMDs) as means to coerce and intimidate. Statements such as North Korea's repeated threats to turn Seoul and Tokyo into a "sea of fire"; Saddam Hussein's threats to incinerate Iran, Saudi Arabia, and Israel; and similar Iranian threats to Israel have certainly been rarer in domestic political contexts driven by internationalizing objectives. Most inward-looking nuclear aspirants were NPT members who not only misled the IAEA or violated their nonproliferation commitments but were also more promiscuous regarding state-directed or state-endorsed exports of sensitive nuclear technologies, fueled largely by the very structure of the domestic models that sustained them in power. Militarily sensitive exports have generally been a source of income for those inward-looking nuclear aspirants that were ill suited to otherwise accumulate resources from competitive civilian exports to the rest of the world.
Thus, whereas inward-looking models might have regarded nuclear weapons programs as assets in the arsenal of building a regime's legitimacy and prestige, outward-oriented ones thwarted such latent utility." From this point of view, Middle East leaders faced lower barriers to, and stronger incentives for, the pursuit of nuclear weapons than did East Asian ones. As a region, the Middle East gravitated toward the inward-looking end of the spectrum for decades, accounting for most cases of nuclearization. Most, though not all, Middle East leaders relied on models of self-sufficiency and nationalism for their political survival, and they had stronger domestic incentives to seek nuclearization. Egypt, a leader of the Arab world, can be considered an anomaly, arguably suppressing nuclearization since President Anwar Sadat launched in-fitah, an opening to the global economy. In contrast to the Middle East, nuclearization has been much less attractive and far more costly for most East Asian leaders since the 1970s, except for North Korea, the autarky-seeking anomaly.
42   ETEL SOLINGEN
Heavy regional concentration of internationalizing models—economic growth via integration in the global economy—in East Asia reinforced each state's incentives to avoid nuclearization. Conversely, heavy regional concentration of nationalist, economically protectionist, and militarized models throughout the Middle East exacerbated mutual incentives to develop nuclear weapons.
Identifying core models of political survival underlying the domestic politics of nuclear aspirants provides a systematic tool, portable worldwide, with premises backed by important evidence.12 Both qualitative and quantitative studies validate the need to pay greater attention to the links between domestic models and nuclear choices. These models are not merely about "domestic politics" but, more broadly, about the way in which leaders define the very nature of their states' place in the global political economy and associated institutions. This very definition provides a filter through which leaders canvass external threats and opportunities, estimate the utility of international regimes such as the NPT, and formulate nuclear policies.
Building on preliminary findings in earlier articles13 and subsequent, more detailed case studies along these lines,14 a quantitative study found support for the propositions that "the process of economic liberalization is associ-ated with a reduced likelihood of exploring nuclear weapons"; that "economic openness has a statistically significant negative effect across all three levels of proliferation" (that is, exploring, pursuing, or acquiring nuclear weapons); and that "economic liberalization dampened the risk" of states deciding "to explore seriously the nuclear option."15 Another quantitative study found that economic liberalization had a positive and statistically significant effect on nuclear-weapons-free zone (NWFZ) treaty ratification.16 In-depth comparative study of nine cases in the Middle East and East Asia, in Nuclear Logics,17 finds the nuclear choices of all cases to be compatible with domestic survival models.18 A recent study refers to the connection between increasing trade openness and reduced incentives to develop nuclear weapons as a "general law."19 The connection between models of political survival and nuclear policies since the 1960s finds support in systematic observations across different regional security contexts, diverse associations with hegemonic powers, and over successive leaderships within the same state.
First, every known case of nuclear renunciation since the 1970s, where a weapons program was either entertained or launched, entailed a domestic evolution toward internationalization. Of all nuclear aspirants (under the world time stipulated earlier), not one endorsed denuclearization—fully and
DOMESTIC MODELS OF POLITICAL SURVIVAL 43
effectively—under domestic regimes that shunned integration in the global political economy. Only leaders and ruling coalitions advancing their political survival through export-led industrialization undertook effective commitments to denuclearize (Japan, Taiwan, South Korea, Egypt under Sadat, South Africa, Brazil and Argentina, Algeria, Libya since 2003). And nuclear decisions were invariably nested in a broader shift toward internationalization in economics and security.
Second, where internationalizing leaders and coalitions became politically stronger, as in Japan, South Korea, and Taiwan, the departure from nuclear aspirations was sustained even as their security context deteriorated (as in the Korean peninsula and the Taiwan Straits, intermittently). The relationship between politically stronger internationalizers and the timing of rolling back nuclear ambitions was also evident in Argentina under Carlos Menem, Brazil under Fernando H. Cardoso, Spain's accession to the NPT preceding European Union (EU) membership, and South Africa, among others.
Third, where leaders and coalitions favoring internationalization were weaker, as was the case historically in Argentina and Brazil until the early 1990s or in Iran until today, they were more politically constrained in curbing nuclear programs.
Fourth, most, albeit not all, defiant nuclear courses have been unmistakably embraced by autarkic or inward-oriented models from Juan D. Peron's Argentina to Getulio Vargas in Brazil, Gamal Abdel Nasser in Egypt, and leaders in North Korea, Iraq, Iran, and pre-2003 Libya. Indonesia's Sukarno and, more recently, the rulers of Syria and Myanmar—all suspected of coveting nuclear weapons—fit this model as well.
Fifth, even advocates of internationalizing models may have to contend with dangerous regions where neighboring leaders endorse alternative economic and nuclear policies. This problem is less intractable in East Asia, where export-led industrialization spread from Japan to most neighboring countries except North Korea, than in the Middle East. In other words, whether the regional center of gravity is internationalizing or inward looking matters for the individual calculations of states within that region.
As gleaned from the experience of an overwhelming number of cases, domestic models of political survival should be considered not merely as afterthought or residual factors, as has been the case thus far, but as more fundamental—indispensable—considerations in explaining nuclear choices. Given the limits of alternative understandings of nuclear behavior,20 the lack
44   ETEL SOLINGEN
of rigorous examination of political survival models as sources of nuclear postures is particularly puzzling. This omission has important implications. A "missing" or "omitted" variable may lead to an overestimation of other causal variables, granting them too large an effect on the outcome while rendering at least some of their effects spurious.21 Without taking into account political survival models, one may not properly understand nuclear behavior or estimate the actual effects of balance of power, international norms and institutions, or democracy. This is different from arguing that such models are the only relevant variable. Nor does introducing a previously omitted variable imply that other variables are rendered irrelevant, only that we are better able to understand their relative impact on nuclear choices.
Thus, political survival models help explain why security dilemmas are sometimes seen as more (or less) obdurate; why some states rank alliance higher than self-reliance while others do not; why nuclear weapons programs surfaced where there was arguably little need for them (Libya, the Southern Cone, and South Africa, among others); and why such programs were obviated where one might have expected them (Vietnam, Singapore, Jordan, and many others). Balance of power, norms, and institutions may be more (or less) significant than political survival in some cases than others; but, in the aggregate, complete explanations of nuclear behavior must include all relevant sources of nuclear behavior for any particular case. In sum, models of political survival and nuclear policies are not merely loosely associated but joined at the hip. Their omission as a significant independent variable may have led to an overestimation of other causal variables and to potential spurious effects. Their inclusion may improve our understanding of the actual effects of security dilemmas, international norms, and institutions when interacting with domestic models.
DOMESTIC MODELS: SCOPE CONDITIONS
The proposition that domestic orientations to the global economy and nuclear policy may be linked is probabilistic, bounded, and refutable. It is probabilistic, as are most arguments in the social sciences, because it does not suggest an inevitable or deterministic outcome.22 It is bounded in three ways.
First, resistance to the global economy may provide only necessary but not sufficient conditions for the development of nuclear weapons programs. For instance, there is no confirmed evidence that all Middle East leaders who rejected the global economy as a favored platform of domestic political survival have also tried to develop nuclear weapons. In previous work I had placed
DOMESTIC MODELS OF POLITICAL SURVIVAL 45
Syria in the category of an anomaly given its inward-looking, nationalist, protectionist, and militarized political-economy. I had done so because, despite long-standing suspicions that Syria was interested in developing nuclear weapons,23 the IAEA had never questioned Syria for possible violations of its NPT commitments. Yet recent evidence suggests that Syria, too—like Iraq, Iran, Algeria, and Libya—had evaded detection of its clandestine nuclear program despite allowing traditional IAEA inspections.24 Furthermore, as Albright and Scheel suggest, unwillingness to implement the Additional Protocol maybe an important indicator of increased risk of proliferation or of intentions to hide secret plutonium separation or enrichment efforts that cannot be detected through traditional safeguards.23 By this measure, there would be less confidence to completely exclude several inward-looking regimes that have not yet signed or ratified the Additional Protocol from the list of suspect states.26
Second, the proposition is also bounded by regional considerations. The extent to which a given region shares a congruent orientation toward the global political economy (either positive or negative) modifies domestic preferences on nuclear issues in each country. For instance, in recent decades East Asian— including Southeast Asian—leaders and ruling coalitions generated one of the most internationalized regions in the world. This fact reinforced the individual incentives of each East Asian leader to maintain a collective trend away from nuclearization for the sake of common regional stability, foreign investment, and domestic growth (even though China had already acquired nuclear weapons in earlier decades). North Korea's economic closure made it more impermeable to the positive regional synergies operating among other East Asian states. By contrast, in the Middle East, most rulers remained committed to relatively closed political economies for many decades.27 The disincentives to develop nuclear programs that often operate for more open political economies were thus weaker in this region as a whole. Individual leaders and their supportive coalitions who might have otherwise favored global integration— as in Jordan or Lebanon—faced an unwieldy neighborhood that actively discouraged it for many years, often through coercive measures.
Third, the argument emphasizing competing political-economy models may be bound by temporal sequences in the acquisition of nuclear weapons. For instance, the incentives of a globally integrated political economy may operate more forcefully on cases where nuclear programs have not yet yielded nuclear weapons, as seems to have been the case in Japan, South Korea, and Taiwan (as well as in Argentina, Brazil, Spain, and others). Such incentives
46   ETEL SOLINGEN
may have less impact once nuclear thresholds—often in the form of nuclear tests—have been crossed.28 China developed nuclear weapons in the early 1960s, many years prior to Deng Xiaopings decision to integrate China in the global economy.29 North Korean leaders have yet to take the China road to political survival through export-led economic growth. China and North Korea are the only declared nuclear weapons states in East Asia (unless one includes Russia in that region). The argument regarding China is also applicable to Israel, which arguably began efforts to develop nuclear weapons in the 1950s. In neither case was the subsequent internationalization of the economy accompanied by denuclearization, but China retained a minimal deterrent, and Israel kept its formal policy that "it will not be the first to introduce nuclear weapons into the region." In most cases, including first-tier nuclear states, it seems far more costly politically for leaders and ruling coalitions to eliminate their existing nuclear weapons entirely than it would be for those who have not yet acquired such weapons to abandon steps in that direction. This would be the case even for internationalizers and most particularly for those surrounded by inward-looking neighbors, such as India and Israel.
Prospect theory provides an explanation for this difference. Leaders value more what they already have ("endowment effect") than what they might get. Hence they are more averse to losing what they possess than foregoing potential future gains.30 Accordingly, leaders may be hypothesized to accept higher risks to retain existing nuclear weapons than to retain programs leading to their potential acquisition. Moreover, the disincentives stemming from an internationalizing model may be stronger at deliberative or incipient stages of nuclear weapons consideration than after they have been acquired. In other words, one may conjecture that, when nuclearization precedes the inception of internationalizing models, subsequent denuclearization may be much harder.
This expectation can be coupled with an argument related to audience costs, which may also play a role in the consideration to abandon nuclear weapons— as distinct from abandoning programs on the road to yield nuclear weapons. Audience costs, such as removal from office or no-confidence votes, are incurred by leaders when they renege on public commitments they have made.3' Because domestic audiences operating in democracies both possess the legal authority to remove leaders from office and face significantly lower hurdles in overcoming collective action problems inherent in removing an incumbent from office, democratic leaders are expected to incur audience costs at a higher rate than nondemocratic leaders.32 This argument could suggest that the costs
DOMESTIC MODELS OF POLITICAL SURVIVAL 47
of reneging on explicit, or even tacit, commitments to protect the country with a nuclear deterrent could be particularly high for democratic leaders, but they may also be high for authoritarian ones, as Weeks suggests.33 For instance, audience costs for Chinese leaders may be high due to domestic expectations that China's rising role in world politics must continue to be backed up by nuclear weapons.
Combining prospect theory and audience costs arguments yields four kinds of scenarios. The first leads to the argument that backing down from (even implicit) commitments to acquire full nuclear capabilities may be easier for autocratic leaders in countries that have not yet achieved weaponization. Audience costs in such cases are assumed to be lower for leaders who must back down from a program rather than a realized nuclear weapons capability. These conditions might enable would-be internationalizers to step down from a nuclear weapons program before it comes to fruition. Iran under an internationalizing leadership—a configuration that is quite hard but less impossible to envisage after the events of June 2009 in the longer run—would match these conditions. Libya may be an example of this scenario as well.
Second, and by contrast, leaders of a democracy that is already in possession of nuclear weapons (or assumed to be so) may perceive domestic audience costs to be too high for advancing denuclearization. Even internationalizing Indian and Israeli leaders may find themselves in such a position, given the regional domination of inward-looking models. Domestic audience costs may be further heightened in these two cases because both India and Israel are democracies surrounded by nondemocracies. Following arguments in democratic peace theory, this could suggest that their publics might be particularly distrustful of denuclearizing when facing nondemocratic regional adversaries.
Third, internationalizing leaders in possession of nuclear weapons in autocratic contexts may be as constrained as in the previous scenario. Chinese leaders, for instance, must compensate inward-looking constituencies (such as the military and other nationalist forces) that have been less than favorable to China's progressive internationalization. The audience costs that might be incurred by abandoning nuclear weapons—given an existing "endowment"— might thus be prohibitive.
The fourth scenario entails a democracy that is not yet in possession of nuclear weapons but has been in the process of developing them. Here, the presumably higher audience costs for a democracy might be offset by the fact that it is prepared to abandon only a program rather than actual weapons.34 Leaders
48   ETEL SOLINGEN
of internationalizing democracies may have stronger incentives to abandon such programs than leaders of inward-looking democracies. Historically, this has been the case with Sweden, Switzerland, Italy, Norway, and others.
Audience costs arguments can also lead to counterintuitive claims, as analyzed in Schultz, such as the "Nixon goes to China" phenomenon.35 Under some circumstances, "backing down" or reneging on commitments may be more likely to ensure a leader's political survival than "standing firm." Wolf, for instance, suggests that reneging does not always lead to the imposition of audience costs and that these costs vary depending on whether or not leaders are "hawks" or "doves."36 On the one hand, backing down may be interpreted by domestic "doves" as a sign of prudence rather than incompetence. On the other hand, because of their greater credibility in matters of national security, hawks may be able to back down without facing severe audience costs.
As the discussion in this section makes clear, the domestic political survival argument is only probabilistic. Internationalizing leaders may embrace nuclear weapons, and inward-oriented leaders may decide to abandon them,37 contingent on the relative incidence of one model or another throughout the region, the presence of a nuclear weapons program as opposed to actual nuclear weapons, and the nature of audience costs in democratic and autocratic regimes. Models of political survival may not capture all the correlates of nuclear preferences (no theory can) and are, after all, only ideal types or conceptual constructs. As such they need not fit every case or indeed any particularly case completely38 but rather provide a heuristic, a helpful shortcut, and a comparative, portable framework capable of reducing complex reality—and all cases—down to some fundamentals.
Models of political survival can explain a number of things: (a) why different actors within the same state vary in their approaches and preferences regarding nuclear policy; (b) why nuclear policies within states may vary over time as a function of the relative power of particular domestic forces; and (c) why different states vary in their commitments to increase information, transparency, and compliance with the nonproliferation regime even when their external landscapes remain unchanged. This heuristic also provides a different foundation for the design of positive and negative inducements to encourage denuclearization than those conceiving of states as unitary actors.
It is of course possible that, even if one finds this (or any other) approach reasonably persuasive in explaining the past, it does not necessarily follow that it will also apply in the future. Different dynamics at work could trigger
DOMESTIC MODELS OF POLITICAL SURVIVAL 49
conditions under which internationalizing models may no longer provide sufficient conditions for continued denuclearization. As Campbell and his coauthors suggested:
. . . there is widespread concern that the calculus of incentives and disincentives has shifted during the past decade, with incentives increasing and disincentives declining. New threats have arisen while the nuclear taboo has weakened. And it is not just a single factor in this new strategic landscape that gives pause. Rather, it is the accumulation of multiple factors and their interplay and mutual reinforcement that account for many of these new dangers.39
Nonetheless, the framework proposed here provides a roadmap for considering the conditions under which its expectations might be corroborated or refuted. This in itself is a significant advantage because various extant frameworks suffer from indeterminacy, tautologies, or post hoc-ism. Neorealism's concepts of self-help and relative power, for instance, can drive states to a wide array of nuclear choices, from straightforward acquisition of nuclear weapons to alliances to renunciation and myriad possibilities in between. Its tenets are thus indeterminate, rendering them hardly a reliable guide to anticipate what states might actually do. Anomalies for neorealism habitually require additional information unrelated to international power balances. As Betts argues, insecurity is not a sufficient condition for acquiring nuclear weapons; many insecure states have not.40 In other words, such theories are afflicted with the problem of multifinality, suggesting many outcomes consistent with a particular value of one variable.41 To this day, it is unclear what the precise underlying measures of relative power are that should lead to nuclearization or abstention.42 Yet such clarity is a sine qua non to avoid circularity and ex post facto rationalizations (such as, "state x went nuclear because of acute insecurity," whereby the acuteness threshold is detected by a nuclear test). Arguments about nuclear chains must be cast in falsifiable terms and enable more clearly defined and testable propositions.
Some of neorealism's deficiencies can be illustrated through the case of Egypt. By any (structural) neorealist account—where the domestic nature of states does not matter and only "relative power" between or among adversaries does—Egypt would have been a "most likely case" for going nuclear. First, it was geographically adjacent to a presumed nuclear weapons state, Israel. Second, over the last few decades it also faced other Middle Eastern states seeking nuclear weapons in its neighborhood, such as Libya in the immediate
50   ETEL SOLINGEN
vicinity as well as Iraq and Iran, and the prospects of additional ones. Third, it lacked conventional superiority over some of those potential adversaries. Fourth, it inhabited a multipolar environment that neorealist predictions have unequivocally associated with nuclearization. Fifth, it lacked a hegemon explicitly providing a nuclear umbrella.
Yet, changes in relative power do not provide coherent accounts of Egypt's denuclearization. Presumed threats from Israel and other regional rivals with nuclear ambitions arguably remained when Egypt rejected "reactive proliferation" since the 1970s. Egypt under Nasser considered nuclear weapons when its conventional gap with Israel was the narrowest and abandoned nuclear aspirations under Sadat, when the gap widened. Egypt considered nuclear weapons when it enjoyed stronger external (Soviet) security guarantees but abandoned them in their absence, as the United States never provided Egypt with equivalent guarantees. The fact that Egypt did not, thus far, acquire nuclear weapons deals a serious disconfirming blow to neorealism.
A test of a particular theory where there is a rather close fit between that theory and the actual outcome—but no corresponding fit between the expectations of other theories and that outcome—provides strong corroboration for that particular theory. Egypt seems to be such a case. Tracing Egypt's non-nuclear status to a domestic political survival strategy seems a risky prediction. Yet no other leading theory predicts that status. Indeed, most predict exactly the opposite. Instead, both the early efforts by Nasser to develop a nuclear weapons program and the subsequent shift by Sadat to renounce nuclear weapons were policies compatible with the respective models each leader relied on to maintain himself in power. Nuclear decisions thus followed changes in domestic political strategies that, in turn, had led to changes in Egypt's strategic alliances and regional policies. Whereas Nasser's Egypt thrived in an aura of inward-looking self-reliance, hypernationalism, and military-technical prowess, Sadat's emphasis on economic growth, foreign investment, exports, military conversion, and a new relationship with international markets and institutions did not leave much room for an expensive nuclear program. Egypt was able to retain its nonnuclear weapons policy despite a strong domestic revisionist current advocating nuclearization.43
Alleged signs of possible revisions in Egypt's policy have been strongest as the threat of Iran's nuclearization grew higher. The likelihood of this turn is, once again, hard to estimate given unclear thresholds for what kinds of changes in relative power yield decisions to acquire nuclear weapons. If indeed
DOMESTIC MODELS OF POLITICAL SURVIVAL 51
Egypt embarks on a weapons-related program in the future, neorealist perspectives would need to establish that: (1) Iran was a strategic threat to Egypt as a state rather than a political one to Egypt's leaders; (2) Iran was a more serious threat to Egypt's security than Israel has been (because dramatic steps would have been taken only after the rise of Iran as a threat and not vis-á-vis Israel for decades); and (3) that the causal mechanisms and consequences of changes in relative power are independent from domestic considerations, such as pressures for matching Ahmadinejad's defiance among Egypt's inward-looking constituencies. This would have far more to do with domestic strains in the existing model of political survival, and perhaps with concerns with Shi'a-Sunni cleavages, than with any fear that Iran might threaten Egypt's territorial integrity.
Scenarios
In the spirit of providing hypotheses applicable to the future that are cast in falsifiable terms and offer testable propositions, Table 3.1 suggests four possible scenarios for the application of models of political survival to 21st-century proliferation trends. Two of these scenarios are compatible with the premises of the framework analyzed so far. The other two scenarios falsify those premises. The horizontal axis in Table 3.1 refers to the two basic models, internationalizing and inward-looking. The vertical axis maps two basic trends, toward nuclearization and away from it. A more detailed elaboration of this schema could help assign more specific probabilities to each scenario.
Scenario 1 suggests a situation where leaders continue to steer internationalizing models in their respective countries and, at the same time, retain commitments to denuclearization. This joint outcome would be compatible with the framework's expectations. This scenario matches the reality of most of East Asia in the early 21st century and has a reasonable likelihood to persist, provided most enabling central features remain in place, including regional
Table 3.1. Models of political survival and nuclear outcomes: Four scenarios.
Nuclear outcomes	Model of regime survival	
	Internationalizing	Inward-looking
Denuclearization	Compatible	Anomaly
	(reasonably likely)	(reasonably unlikely)
Nuclearization	Anomaly	Compatible
	(reasonably unlikely)	(reasonably likely)
52   ETEL SOLINGEN
and global conditions—economic and political—propitious for these models' survival. This scenario is supported, among many other considerations, by the presence of some 28,000 Japanese companies employing over a million workers in China as of 2005, double the number merely a decade earlier, and of over a million Taiwanese entrepreneurs operating in the mainland. Former Japanese Ministry of Foreign Affairs official Kaneko Kumao draws attention to another requisite for the continuity of the postwar model of political survival incepted by Yoshida Shigeru:
Japan maintains cooperative nuclear agreements with six countries, the United States, Britain, France, Canada, Australia, and China. I personally negotiated ... most of these.... If Japan misuses its civilian nuclear program for military purposes, a set of stringent sanctions will be imposed on it, including the immediate return of all imported materials and equipment to the original exporting country. Should that ever happen, nuclear power plants in Japan [would] come to a grinding halt, crippling economic and industrial activities. It is simply unthinkable that the nation would be willing to make such a heavy sacrifice—unless it [was] really prepared to start a war. In this sense, the bilateral nuclear energy agreements provide a rather effective deterrent, certainly more effective than the NPT.44
Scenario 3 entails the continuity of internationalizing models accompanied by discontinuities in nuclear policies. In other words, internationaliz-ers go nuclear, which would constitute an anomaly for the main argument proposed in this chapter. This may be less likely under the current circumstances of a strongly internationalizing East Asia functioning as the locomotive of an expanding global economy. However, should some leaders backtrack on internationalizing models, such prospects could be higher. For instance, a Chinese leadership that does not cope appropriately with domestic challenges of economic and political transitions could be weakened or replaced by inward-looking opponents, with attending regional consequences, including heightened concerns among its neighbors.45 Furthermore, internationalizing leaders everywhere are not immune to miscalculations in overplaying nationalist cards or falling victims to "blowback" and entrapment by inward-looking constituencies more favorable to nuclearization46 The 2005 Chinese legislation codifying a declaration of war against Taiwan if the latter declares independence could provide an example of unintended effects of such miscalculations. In the Middle East, some have suggested that Turkey could, under
DOMESTIC MODELS OF POLITICAL SURVIVAL 53
certain circumstances, reconsider its nuclear status, for instance if it were to face a nuclear Iran.47 In the last two decades, Turkish leaders have appeared to have transcended the Middle East's modal inward-looking path, consolidating an internationalizing model and renouncing nuclear weapons. If this choice were reversed while Turkey sustains the current model, the political survival argument would be refuted. If, however, Turkey were to reverse its nuclear commitments in tandem with progressively more inward-looking domestic models—exacerbated by EU exclusion—the argument would be sustained. Such domestic changes could also unleash a deterioration in Turkey's relationship with the North Atlantic Treaty Organization (NATO), highlighting the importance of domestic considerations in shaping security policies.
Scenario 2 points to conditions where inward-looking models dominate but nonetheless embrace denuclearization. The past record of nuclear aspirants shows that this joint occurrence has been rare. This scenario would constitute another anomaly for the basic argument and could be illustrated by situations where inward-looking regimes in North Korea and Iran join and implement durable, transparent, and mutually and unconditionally verifiable agreements renouncing nuclear capabilities. The prospects for this outcome do not seem very likely in early 2010. However, if, for instance, Iran's and North Korea's nuclear policies change in tandem with domestic survival models, as they had in Egypt, Libya, and elsewhere, the argument that these models are deeply implicated in shaping nuclear policy would be corroborated. In East Asia, scenario 2 would involve the rise of inward-looking models in pivotal states that nonetheless retain NPT commitments and compliance. This outcome might be explained by path dependency or the increasing returns of a nearly four-decades-old legacy of shunning nuclear weapons (except for North Korea), a legacy that would have to overwhelm the incentives of inward-looking leaders to transcend it.
Scenario 4 suggests resilient inward-looking leaders resistant to internationalization, a defining characteristic of much of the Middle East for many decades, accompanied by intermittent efforts to acquire nuclear weapons. This scenario is compatible with the basic framework advanced here, and its permanence does not bode well for denuclearizing shifts in that region. The dramatic expansion in demand for nuclear technology in the Middle East could well present very real challenges to the nonproliferation regime. Much of it could take place in response to Iran's ability to flout nonproliferation commitments with relative impunity and an increased concern with Iran's hegemonic
54   ETEL SOLINGEN
assertion, particularly through support of Shi'a minorities within Gulf states. The outcome of the Iranian nuclear debacle, and of resolutions adopted by the U.N. Security Council, may be the single most important predictor of what happens in the region next, because other inward-looking regimes are bound to learn important lessons about the degree of robustness of the nonprolifera-tion regime.
In East Asia, the widespread replacement of internationalizing models of political survival is certainly plausible, although the early signs of recovery from the 2008-2009 financial crisis suggest otherwise. The outcome of the 1997 Asian crisis also signaled more resilience than anticipated and, despite some political turnovers, did not lead to significant departures from internationalizing strategies. Such turns remain nonetheless conceivable in conjunction with global recessions or other regional and domestic downward economic protectionist or nationalist spirals. Significant domestic evolutions away from internationalizing trajectories—from China to South Korea, Indonesia, and Japan—might encourage nuclear chains (see the discussion in the following section). This outcome, although unlikely under present conditions in the world's most economically dynamic region, would be compatible with predictions in Cell 4. Thus far, the leaders of China, Japan, and South Korea have responded to the most severe global economic crisis in recent decades with a first-ever joint summit meeting in December 2008, pledging to institutionalize regular three-way summits. As the Japan chapter in the companion volume to this one recounts, even highly conservative leaders such as Abe Shinzo and Taro Aso remained constrained by a significant domestic consensus to stay the internationalizing, nonnuclear course.
REGIONAL EFFECTS ON MODELS OF POLITICAL SURVIVAL AND OTHER DIFFUSION MECHANISMS
The scenarios in section II offer some guidance regarding continuity and change in nuclear trajectories on the basis of evolving models of political survival. This is not, however, a purely "domestic" garden-variety argument but one that concentrates attention on the connection between domestic and international politics by definition (ruling coalitions make choices about the kinds of links to the outside world that best serve their design to stay in power). Furthermore, these models are deeply interactive with regional circumstances, making the issue of chains and diffusion effects endogenous to the argument. In other words, the relative incidence of alternative models in neighboring
DOMESTIC MODELS OF POLITICAL SURVIVAL 55
states matters. One can think about the regional environment as an aggregate measure of the relative strength of internationalizing or inward-looking models. The extent to which regions share congruent orientations toward internationalization (either positive or negative) modifies domestic preferences on nuclear issues.
The collective evolution of East Asia toward internationalization reinforced individual incentives of leaders to avoid nuclearization to preserve regional stability, foreign investment, and domestic economic growth, despite China's 1964 tests. Converging internationalizing models thus appear collectively stable, creating an environment that exhibits stronger immunity to nuclearizing chain reactions, even if juche blocks these regional synergies from influencing North Korea. In the Middle East, many rulers retained relatively closed political economies, facing fewer domestic and international disincentives for nuclearization. A regional environment dominated by inward-looking models is also collectively stable, in the sense that it reinforces individual incentives by leaders and ruling coalitions to retain those models, which expand state entrepreneurship, military expenditures, and baroque weapons.48 Thus, such models feed on each other's existence, raising immunity against internationalizing models within the region and lowering immunity to nuclear chains. Neighboring leaders who might have otherwise favored internationalization face an unwieldy neighborhood that actively discourages it.
Beyond these contextual regional influences, deep local recessions, regional or global economic downturns, and other severe disruptions in the global political-economic system could provide important triggers capable of undermining internationalizing models and buttressing inward-looking ones. Writing in the immediate aftermath of the worst financial and economic crisis since the Great Depression indeed gives rise to concern. In a more globalized economy, the potential for economic crises to diffuse can arguably accelerate the ascendance of leaders logrolling antiglobalization inward-looking forces, in a coalitional domino of sorts. It is against such background that decisions to proliferate could become more feasible, a possibility entertained in Nuclear Logics, even under the more favorable conditions of 2006.49
The prospect for increased proliferation under such circumstances is quite different from arguing, as neorealism does, that the nuclearization of one state almost inexorably leads to the nuclearization of its neighbor(s). Nuclear decisions are not the automatic result of shifting international power balances that are very hard to measure accurately a priori but rather remain contingent on
56   ETEL SOLINGEN
the changing nature of domestic coalitions. The fact that automatic proliferation has not yet happened as systematically as one might think brings to relief the importance of understanding the conditions (political survival models) that mediate between triggers such as North Korea's nuclear tests and nuclear decisions by other East Asian states. East Asian leaders have adopted and diffused internationalizing models of engagement in the global economy. Some might argue that such models make individual states less secure, presumably because they thwart nuclear weapons programs. Yet many East Asian states lacking natural resources or nuclear weapons seem far less vulnerable or insecure than North Korea, Iran, or pre-2003 Iraq. Internationalizing models have turned them into engines of the 21st-century global economy, with much higher levels of domestic political stability, social and gender equity, human rights, expected life spans, employment, and educational endowments than unstable Middle East nuclear weapons aspirants, despite their rich natural and human resources.
The fate of the nonproliferation regime could also influence the domestic balance of power between internationalizing and inward-looking forces. A variety of challenges appears to have placed the regime in a tenuous and fragile predicament. Increased legal and illegal flows of sensitive technologies; the IAEA's inadequate access to information regarding illegal transfers (Iraq, Iran, Libya, North Korea, and Syria); NPT violations by North Korea, Iran, and Syria; and the failed 2005 NPT Review Conference remain important tests of the regime's effectiveness, increasing uncertainty about its ability to endure beyond the 2010 Review Conference.50 A collapse of the global nonproliferation regime could weaken internationalizers in their efforts to stem pressures for nuclearization from inward-looking (domestic and foreign) adversaries. At the same time, such collapse should not be taken for granted. It is precisely in times of crisis, as Dunn points out, that institutions may develop stronger foundations.51
In sum, as this chapter's overview suggests, domestic models of political survival are filters through which other considerations operate, tilting the balance of incentives and disincentives to acquire nuclear weapons in one direction or another. This point is often missed by reductionist interpretations of the argument, which falsely assume that taking domestic political survival models into account implies lack of attention to external dimensions, including threats. One might consider these models as something akin to a lens that refracts, enhances, or redirects external stimuli according to the (dis)incentives
DOMESTIC MODELS OF POLITICAL SURVIVAL 57
embedded in each model. Without a proper understanding of the operation of this variable, much of the utility of the argument can inevitably be lost. Properly interpreting these models as filters for a wide-ranging set of domestic, regional, and global opportunities and constraints also requires tolerance for complexity. A state's evolution toward or away from nuclear weapons takes place neither in the soul of a single person nor as an autopilot response to system-level changes in relative power. As Philip Tetlock's masterful treatise on expert political judgment and prediction suggests, parsimony can be the enemy of accuracy, a substantial liability in real-world forecasting.52