Abstract

When does a nuclear-armed state's provision of security guarantees to a militarily threatened ally inhibit the ally's nuclear weapons ambitions? Although the established security model of nuclear proliferation posits that clients will prefer to depend on a patron's extended nuclear deterrent, this proposition overlooks how military threats and doubts about the patron's intentions encourage clients to seek nuclear weapons of their own. To resolve this indeterminacy in the security model's explanation of nuclear restraint, it is necessary to account for the patron's use of alliance coercion, a strategy consisting of conditional threats of military abandonment to obtain compliance with the patron's demands. This strategy succeeds when the client is militarily dependent on the patron and when the patron provides assurances that threats of abandonment are conditional on the client's nuclear choices. Historical evidence from West Germany's nuclear decisionmaking provides a test of this logic. Contrary to the common belief among nonproliferation scholars, German leaders persistently doubted the credibility and durability of U.S. security guarantees and sought to acquire an independent nuclear deterrent. Rather than preferring to renounce nuclear armament, Germany was compelled to do so by U.S. threats of military abandonment, contradicting the established logic of the security model and affirming the logic of alliance coercion.

Introduction

When does a nuclear-armed state's provision of security guarantees to a militarily threatened ally inhibit the ally's nuclear weapons ambitions?1 The security model of nuclear proliferation argues that by providing extended deterrence to an insecure client, a patron obviates the client's need for indigenous nuclear weapons capabilities.2 Underpinning this argument is the logic of “nuclear dependence,” which posits that clients will prefer to delegate their security to a patron to avoid the risks associated with nuclear acquisition. For clients facing military threats, however, uncertainty about the durability and credibility of a patron's protection means that security guarantees are imperfect substitutes for independent capabilities. As a result, clients are compelled to minimize risks to their survival by pursuing both security guarantees and nuclear weapons of their own.

Critics of the security model contend that this indeterminacy in its logic indicates that security guarantees cannot explain nuclear restraint, and that competing causal variables are necessary to account for client states' nuclear choices. In particular, past scholarship has largely turned to domestic factors to explain the choice of external reliance versus nuclear weapons acquisition, including the politico-economic preferences of a country's ruling coalition, the popularity of nonproliferation norms, the psychology of political leaders, and the prevalence of domestic actors with a veto over nuclear decisions.3 These accounts, however, have mistakenly downplayed the security model's explanatory power by neglecting to consider a critical feature of alliance politics: a patron's use of coercion.

In this article, I propose an original model of security guarantees based on the logic of alliance coercion—that is, a strategy consisting of a patron's use of conditional threats of military abandonment to obtain a client's compliance with its demands. This account builds on the argument that security guarantees form a basis for nuclear restraint, while jettisoning the assumption that militarily threatened states will always prefer to rely on a patron's extended nuclear deterrent. Instead, I argue that by threatening to withdraw its security guarantees, a patron can compel its client to forgo development of an indigenous nuclear arsenal.

The effectiveness of alliance coercion as a nonproliferation tool depends on two factors. First, the client must be militarily dependent on the patron. If a client can independently deter its adversary, the prospect of military abandonment will have little effect on the client's security, diluting the patron's coercive leverage. Although acquisition of a credible nuclear deterrent reduces the client's dependence, nuclear arms take time to develop, creating a window of opportunity during which the patron can coerce its client into repudiating its nuclear ambitions. Once a client acquires a credible deterrent, however, its vulnerability to alliance coercion declines precipitously.

Second, the patron must provide assurances that threats of abandonment are conditional on the client's nuclear choices. If the client expects military protection to be withdrawn regardless, then it has little incentive to comply with the patron's demands. Consequently, the patron must signal its commitment to the client if it disavows nuclear armament, including by deploying forces on the client's territory and by consulting with the client on military strategy affecting its security. This article therefore builds on existing analyses demonstrating that commitment signals augment the proliferation-inhibiting effect of security guarantees.4 But whereas past research suggests that such signals alone will inhibit the nuclear ambitions of threatened clients, my account argues that without coercion, even clients with credible security guarantees will be driven to acquire nuclear weapons.

To test these competing logics of security guarantees, the article traces the process of nuclear decisionmaking in the Federal Republic of Germany (FRG) from 1954 to 1969.5 As a militarily threatened recipient of U.S. security guarantees, West Germany is representative of other U.S. clients, enabling scholars to make inferences from its nuclear history. Additionally, the country is an especially easy case for the logic of nuclear dependence, as the U.S. imperative to contain Soviet expansion should have led West Germany to perceive U.S. security guarantees as highly credible. By contrast, although Germany's acute dependence arguably makes this an easy case for the logic of alliance coercion, existing theories of alliance politics suggest that Washington's unique interest in West Germany's defense should have undermined the credibility of U.S. threats of abandonment, making the case a hard one for my argument.6 Finally, because U.S. alliance policy toward Germany varied over time, scholars can employ within-case variation to test the distinct predictions of the two competing logics of security guarantees.

By examining West German nuclear history, this article helps to correct the mistaken belief—still common among political scientists studying nuclear proliferation—that German nuclear ambitions were fleeting or nonexistent. For example, two widely cited quantitative analyses of nuclear proliferation code West Germany as having never pursued nuclear weapons.7 More recent research contends that the country only explored the weapons option from 1957 to 1958.8 With notable exceptions, political scientists have concluded that there was never “any serious rethinking of the German acquiescence to renunciation of a truly independent, national nuclear capability.”9 As I argue below, these assertions overlook evidence that Germany sought to acquire an independent nuclear deterrent from 1956 to 1963, and that it fought to retain its weapons option from 1964 to 1969.10 Rather than voluntarily forswearing acquisition of a nuclear capability, the FRG was compelled to do so by explicit U.S. threats of military abandonment. By failing to incorporate evidence of West German nuclear ambitions and U.S. abandonment threats into past analyses of nuclear decisionmaking, nonproliferation scholars have not only misunderstood a critical case of nuclear reversal, but have also overstated the explanatory power of the logic of nuclear dependence.

By describing and testing the logic of alliance coercion, this article advances the study of nuclear proliferation and improves upon existing theories of alliance politics. Most important, my argument contributes to a growing literature that highlights the centrality of U.S. nonproliferation policy for containing the spread of nuclear weapons.11 Nonproliferation scholars had previously accepted Kenneth Waltz's assertion that “in the past half-century, no country has been able to prevent other countries from going nuclear if they were determined to do so.”12 By identifying the conditions when security guarantees serve as the basis for coercive leverage, this article tempers Waltz's skepticism and demarcates the contours of U.S. influence. In addition, it challenges the argument that nuclear decisions can be understood only by examining domestic political variables.13 Instead, by demonstrating how interactions between a client and its patron shape the client's incentives for nuclear acquisition or restraint, it returns the focus to external security imperatives for explaining nuclear choices. Finally, by developing the logic of alliance coercion and demonstrating how it shaped the nuclear trajectory of a critical U.S. client, this article provides an important amendment to standard theories of alliance politics, which argue that threats of abandonment will not be credible against strategically valuable allies.14

These theoretical insights provide lessons for the design of U.S. non-proliferation policy. Policymakers assume that they can inhibit the nuclear ambitions of militarily threatened states by providing credible security guarantees, consistent with the logic of nuclear dependence.15 Likewise, to keep current clients such as Japan or South Korea from acquiring nuclear weapons, the logic of nuclear dependence suggests that the United States should strengthen its security guarantees by offering improved commitment signals. But if technically capable clients nevertheless seek to acquire nuclear arms in response to a worsening military environment, the effect of security guarantees would also depend on whether the United States is willing to employ alliance coercion and on the degree that clients are vulnerable to its use.

This article proceeds as follows. First, it explains the logic of nuclear dependence and details its conceptual and empirical flaws. Second, it describes a new logic of security guarantees, based on the mechanism of alliance coercion. Third, it tests the utility of this logic against evidence from the West German case. Fourth, it responds to potential objections and competing interpretations of West Germany's nuclear behavior. Finally, it analyzes the implications of alliance coercion for international relations theory and U.S. nonproliferation policy.

The Flawed Logic of Nuclear Dependence

The security model of nuclear proliferation posits that states seek to acquire nuclear weapons in response to external military threats. Insofar as a threatened state exhibits restraint, “it does so largely as a function of [a] countervailing deterrent capability” provided by a nuclear-armed patron.16 Underpinning this claim is the idea that acquisition of nuclear weapons heightens a state's insecurity, either by increasing its odds of becoming a target for nuclear first strikes or by fueling a needless arms race. States will thus prefer to forgo nuclear armament and rely exclusively on their patron's nuclear capabilities. The key assumption of this logic is that a patron's guarantees “obviate the minor partner's need to develop independent nuclear forces.”17 In other words, “A security guarantee serves as a substitute for a state obtaining nuclear weapons.”18

Notably, the logic of nuclear dependence does not exclude the prospect that clients will “hedge” by investing in civilian nuclear infrastructure that they can convert to military use.19 According to proponents of this logic, however, hedging will be the maximum extent of clients' nuclear activities if their patron provides credible security guarantees. As T.V. Paul contends, the “credibility of the ally's commitment is critical in dissuading a technologically capable state in a zone of high conflict from ‘going nuclear.’”20 And although Nuno Monteiro and Alexandre Debs acknowledge the possibility that a patron can employ threats of abandonment to discourage a client's nuclear ambitions, they nevertheless argue that credible guarantees are sufficient to produce voluntary external reliance. As they predict, “Should a protégé expect its ally to remain a reliable guarantor of its security, it [will] lack the willingness to acquire the bomb.”21

The logic of nuclear dependence, however, overlooks how military insecurity and uncertainty about a patron's intentions create powerful incentives for states to seek both security guarantees and nuclear weapons. As Waltz argues, states facing military threats can either balance “internally” by acquiring armaments or “externally” by seeking allies.22 The choice between arms and allies need not be exclusive, and states pursuing nuclear weapons may still seek security guarantees to minimize risks to their survival.23 Unlike acquisition of indigenous capabilities, however, dependence on a patron leaves militarily threatened clients vulnerable to the catastrophic risk of abandonment, because patrons can withdraw their protection in response to changing strategic imperatives or domestic political shifts.24 And although acquisition of nuclear weapons can aggravate security competition, for states already facing a threatening adversary, the dangers of military abandonment exceed the risk of heightened regional hostility. Given the existential consequences of being left alone to face military threats, states in an anarchical international system “cannot depend on others for their own security.”25 Consequently, even fleeting signs that a patron lacks commitment will aggravate underlying anxieties about the durability of security guarantees, driving insecure clients to pursue nuclear weapons.

By disproportionately and unambiguously raising the costs of conquest, nuclear weapons introduce a qualitative change in the military balance, enabling an otherwise weak state to preserve the territorial status quo against an adversary with superior conventional military power.26 Because possession of an independent retaliatory capability virtually guarantees a country's survival in the face of existential threats, militarily insecure countries will be particularly interested in acquiring nuclear weapons as an insurance policy against abandonment.27 By contrast, when a patron's territory is vulnerable to nuclear retaliation by the client's adversary, doubts about the patron's willingness to risk nuclear devastation for the sake of the client will produce anxiety about the credibility of extended deterrence guarantees. By acquiring an independent deterrent, the client can circumvent those concerns, bolstering the client's incentive for nuclear weapons acquisition.

Proponents of the logic of nuclear dependence may object that a client's nuclear arsenal is unlikely to be as effective or survivable as its patron's, creating incentives for external reliance.28 Contrary to this claim, however, a number of second-tier nuclear powers have achieved survivable second-strike forces, including France, which faced the daunting challenge of a potential Soviet first strike.29 Moreover, an independent arsenal has substantial deterrent utility even for states that cannot acquire a survivable force; by predelegating tactical and strategic nuclear weapons to frontline commanders, states can deter conventionally superior adversaries by threatening to escalate asymmetrically in response to military assaults.30 A relatively small nuclear force also has the potential to catalyze the patron's use of nuclear and conventional weapons by threatening an uncontrollable escalatory spiral with the adversary.31 Thus, by gaining a “finger on the nuclear trigger,” a client facing a nuclear-armed adversary can ensure that its patron will not be deterred from activating its guarantees in a crisis. Together with uncertainty about the durability and credibility of security guarantees, the unique deterrent and catalytic effects of nuclear weapons undermine a threatened client's preference for relying exclusively on a patron's protection. Table 1 illustrates the divergent expectations of the logic of nuclear dependence and the logic of self-help regarding a militarily threatened client's hierarchy of preferences.

Table 1.

Divergent Hierarchies of Preference for Militarily Threatened States

Logic of Nuclear DependenceLogic of Nuclear Self-Help
1. security guarantee, no nuclear deterrent 1. security guarantee and nuclear deterrent 
2. security guarantee and nuclear deterrent 2. nuclear deterrent, no security guarantee 
3. nuclear deterrent, no security guarantee 3. security guarantee, no nuclear deterrent 
4. neither 4. neither 
Logic of Nuclear DependenceLogic of Nuclear Self-Help
1. security guarantee, no nuclear deterrent 1. security guarantee and nuclear deterrent 
2. security guarantee and nuclear deterrent 2. nuclear deterrent, no security guarantee 
3. nuclear deterrent, no security guarantee 3. security guarantee, no nuclear deterrent 
4. neither 4. neither 

The imperative for recipients of security guarantees to acquire nuclear weapons is greatest when they face military threats.32 For clients that do not face a high risk of military conflict, the downsides of nuclear acquisition— especially the risk that armament will exacerbate security competition with neighboring states or trigger the imposition of economic sanctions—are likely to outweigh the smaller benefits of an independent arsenal. Such states are likely to rely on security guarantees while concurrently hedging by maintaining a dual-use civilian nuclear infrastructure.33 Likewise, some states will opt for external reliance because they lack the technical wherewithal or financial resources to construct nuclear weapons. A high risk of military conflict and the material capacity to acquire a deterrent are scope conditions for the pursuit of nuclear weapons—in their absence, states are more likely to prefer reliance on a patron.

The number of states that received U.S. security guarantees but which nevertheless exhibited nuclear weapons ambitions underscores the flaws in the logic of nuclear dependence. Of the twenty-seven states that signed formal defense pacts with the United States during the Cold War and reached a minimum threshold of nuclear technical capability, twelve—or 44 percent— pursued nuclear weapons or explored the weapons option.34 These were Argentina, Australia, Brazil, France, Iran, Italy, Japan, Norway, South Korea, Taiwan, the United Kingdom, and West Germany. Of those, only France, Norway, and the United Kingdom were not pressured by the United States to give up their weapons ambitions, while the remainder faced U.S. demands for nuclear renunciation.35 And although a recent study by Philipp Bleek and Eric Lorber reveals a negative association between security guarantees and nuclear pursuit and acquisition, it does not test whether countries opted for nuclear restraint voluntarily or whether their ultimate external reliance was the product of coercive pressure.36 By distinguishing these logics, this article enables examination of the causal mechanism responsible for the effect of security guarantees.

Critics of the security model have interpreted the flaws in the logic of nuclear dependence as evidence that a patron's provision of security guarantees is secondary to clients' nuclear decisions. Most scholars, however, have failed to consider how military protection can serve as a basis for coercion, deflating the causal impact of security guarantees and inflating that of competing variables. Identifying the conditions under which alliance coercion is possible avoids the flaws in the logic of nuclear dependence, and it explains variation in the inhibiting effect of security guarantees on nuclear proliferation.

The Logic of Alliance Coercion

Unlike the logic of nuclear dependence, the logic of alliance coercion posits that for technologically advanced states facing military threats, a patron's promises of protection alone, however credible, are unlikely to inhibit a client's nuclear ambitions. Although security guarantees ameliorate a client state's imperative to acquire nuclear weapons, military threats—together with uncertainty about the durability and credibility of the patron's promises—will drive clients to seek both a nuclear deterrent and external protection. Alliance coercion precludes this possibility. By threatening to withdraw protection if a client attempts to acquire nuclear weapons, a patron forces a trade-off between bomb ambitions and security guarantees. This strategy succeeds under two conditions: first, the client must be militarily dependent on the patron; and second, the client cannot expect the patron to withdraw its guarantee regardless of the client's nuclear choices.37

Theories of alliance politics identify four factors determining the degree of a country's external military dependence.38 First, the greater the disparity in military capabilities between a state and its potential adversary, the more the state requires assistance from allies. Second, the greater the odds that the state will engage in military conflict with a potential adversary, the more the military imbalance with that adversary will affect the state's security. Third, dependence varies with a potential ally's ability to compensate for military imbalances. By definition, however, a superpower has preponderant military resources, obviating the salience of this factor in the analysis of superpower security guarantees.39 Finally, dependence varies inversely with the state's alignment alternatives—the fewer allies it can choose among, the greater its dependence.

Threats of military abandonment linked to nonproliferation demands succeed because a credible nuclear deterrent takes time to develop. Once the client constructs a nuclear arsenal and adopts either a posture of assured retaliation or asymmetric escalation,40 it can deter threats without external intervention, reducing its dependence and, by extension, the patron's coercive leverage.41 As a result, clients anticipating coercive pressure from their patrons have an incentive to quickly and secretly acquire nuclear weapons. But because a patron can withdraw security guarantees before the client can procure indigenous nuclear capabilities, the patron has a temporary period during which it can compel militarily dependent clients to constrain their ambitions by renouncing nuclear weapons and accepting international safeguards. Facing the choice between external reliance and potential destruction, a militarily dependent client is compelled to accept the former, despite its preference for acquiring nuclear weapons.

Critics of this logic might object that insofar as a client's security is a strategic interest of the patron, coercive threats of abandonment will lack credibility. According to Glenn Snyder, this was the case during the Cold War: “The superpowers [were] solidly committed by their own interests to defend their allies, hence their de-alignment [was] irrational.”42 If abandoning the client requires the patron to act against its interests, then the patron not only lacks coercive leverage, but should have little need for it in the first place, because valuable clients will have confidence in their patron's commitment, enabling them to voluntarily forswear nuclear acquisition.

Yet, notwithstanding a superpower's interest in coming to its client's defense, the client can never be perfectly certain about the patron's reliability. Even if the client believes that its patron is likely to remain committed, the catastrophic consequences of abandonment mean that the slightest possibility that the patron will renege on its promises has disproportionate effects on the client's calculations. This irreducible uncertainty about the patron's reliability not only motivates the client's pursuit of an indigenous nuclear deterrent in the first place, as argued above, but also provides the patron with a credible threat to withdraw its protection. By manipulating the client's underlying fear of abandonment, the patron can employ what Thomas Schelling calls the “threat that leaves something to chance” to compel the client into compliance with the patron's demands.43

Three additional factors can exacerbate the client's baseline anxiety about its patron's level of commitment, bolstering the patron's coercive credibility. First, if the patron's leaders face domestic political pressure for military retrenchment, they can more credibly threaten to withdraw their provision of security guarantees. Even if the client believes that its patron has a strategic interest in providing security guarantees, isolationist domestic constituencies magnify the client's underlying uncertainty about the patron's intentions and enable the patron to signal coercive resolve by drawing attention to the domestic benefits of withdrawing its protection.

Second, the lower the strategic benefits to the patron of maintaining its security guarantees, the more the client will fear that its patron may forgo those benefits by opting for military abandonment. A nuclear-armed superpower in a bipolar or unipolar international system derives the fewest benefits from alignment with other states because, as Waltz argues, “[t]he gross inequality between the two superpowers and the members of their respective alliances makes any realignment of the latter fairly insignificant.”44 Owing to its enormous conventional and nuclear capabilities, a superpower patron can survive the loss of any one client, providing the patron with substantial leeway in determining its level of commitment, and consequently increasing the client's uncertainty about its patron's intentions. Moreover, because the impact of domestic politics on a state's foreign policy is greatest when that state faces weak systemic constraints, the patron's military independence magnifies the effect of any isolationist domestic constituencies on the client's expectations about the patron's reliability.45 Thus, even if a client expects the superpower patron to suffer costs from terminating their alliance, a superpower has the capacity to absorb the loss of a client with moderate effects on its security. That capacity reduces the client's certainty that its patron will always behave consistently with its own strategic interests, reinforcing the credibility of the patron's coercive threats.

Third, the greater the cost to the patron of intervening militarily in defense of its client, the greater the client's fear of abandonment. In particular, if the client's adversary develops military capabilities that directly threaten the patron's territory, the patron's motivation to risk its own security for the sake of its client declines. Even when a client believes that its patron has an interest in maintaining promises of military protection, the adversary's ability to punish the patron for acting on those promises creates a countervailing pressure that exacerbates the client's underlying anxiety. Such anxieties explain why during the Cold War, it required only “five per cent credibility of American retaliation to deter the Russians, but ninety-five per cent credibility to reassure the Europeans.”46 It is precisely because of the client's fear that its patron will be deterred from defending its security that the client seeks signals of commitment such as troop deployments. But because the client depends on those signals to assuage its doubts, the patron can exploit them for coercive effect, threatening to withdraw or weaken its commitments if the client defies the patron's demands.

The client's anxiety about the patron's reliability enables the patron's credible threat of abandonment. That anxiety, however, impedes the patron's corresponding assurance that abandonment will not occur regardless of the client's nuclear choices. The patron must therefore balance threats to withdraw security guarantees if the client chooses to pursue nuclear weapons with promises to maintain protection if it does not. As Schelling argues, “[A]ny coercive threat requires corresponding assurances; the object of a threat is to give somebody a choice.”47 Coercion cannot succeed without conditionality—if the patron threatens to withdraw military support irrespective of the client's nuclear choices, the client has no incentive to change its behavior. To successfully implement alliance coercion, therefore, the patron must maintain—or provide— signals of commitment.48 One such signal is to deploy its military forces on the client's territory, creating a trip wire that ensures the patron's military engagement in the event of aggression by the client's adversary. A second signal is to implement regular consultations with the client regarding military plans that affect the client's security. Such institutionalized military consultation allows for greater information sharing about the patron's intentions, while giving the client influence over strategic planning.

Conversely, a patron's actions that cause the client to believe that the patron intends to reduce its military commitments decrease the patron's leverage by severing the linkage between the client's nuclear choices and the patron's security guarantees. To maximize its coercive leverage, the patron must avoid unconditional military drawdowns, given that these effectively punish the client irrespective of its compliance with nonproliferation demands. Likewise, the patron must abstain from linking its signals of commitment to ancillary, nonnuclear disputes, because this action threatens to weaken the client's expectation of protection even if the client halts its nuclear ambitions.

Critics may raise a second objection to the logic of alliance coercion— namely, that threats of abandonment will themselves undermine the client's willingness to rely on the patron's security guarantees. Whereas the previous objection contends that the patron's threats of abandonment will lack credibility, this argument suggests that the patron's promises of protection will lack credibility. This claim, however, mistakenly assumes that alliance coercion operates on the basis of trust, rather than dependence. Although alliance coercion requires that the patron continue to provide protection if the client does not pursue nuclear weapons, the client need not have perfect confidence in the patron's guarantee to forgo armament. Because a dependent client cannot achieve its security without the patron's protection—and because the patron can withdraw that protection before the client develops nuclear weapons— the client has no choice but to rely on its patron.

Alliance Politics and West German Nuclear Ambitions

In the analysis below, I test the preceding claims against evidence from the case of West Germany. It is useful to briefly summarize these claims. The logic of nuclear dependence posits that a patron's explicit, formal security guarantee backed by force deployments and military integration will dissuade clients from pursuing nuclear weapons. By contrast, the logic of alliance coercion posits that for a militarily threatened client, a patron's provision of security guarantees alone will not bring about nuclear renunciation. Instead, a client will constrain its nuclear ambitions only after its patron issues conditional threats of military abandonment linked to demands for nuclear restraint. To achieve this result, however, the patron must assure the client of the conditionality of its coercive threats by abstaining from unilateral reductions in its commitments, and by assuring the client that it will maintain its security guarantees if the client complies with its nonproliferation demands.

WEST GERMAN MILITARY DEPENDENCE, 1954–69

In light of its military inferiority relative to the Soviet Union, the persistent threat of armed conflict on its territory, and its lack of alternative alignment options, West Germany was highly dependent on U.S. military assistance. By 1967 the German army could field roughly 325,000 soldiers against the Soviet Union's approximately 2 million soldiers, who were backed by Moscow's nuclear arsenal.49 As a result, Germany relied on U.S. military forces to deter a Soviet invasion, a dependence that only worsened as Soviet nuclear capabilities expanded. West German leaders were especially worried that as the superpower nuclear balance narrowed throughout the 1950s and 1960s, the Soviet Union could employ limited military probes meant to impose faits accomplis without triggering nuclear retaliation.50 Accordingly, the FRG persistently sought U.S. force deployments, not only to preclude limited Soviet military advances but also to signal Washington's commitment. Successive German leaders saw U.S. troops as “the best protection and guarantee for German security,”51 and understood that without U.S. military assistance—and especially without American nuclear forces—German troops would simply be “cattle for slaughter.”52

West Germany also faced a high risk of military conflict on its territory. In addition to being ground zero for the superpower confrontation, West Germany had bilateral disputes with the Soviet Union over the status of West Berlin, the presence of Western forces in the Federal Republic, and Bonn's access to nuclear weapons. Repeated threats of force by Soviet leader Nikita Khrushchev over these and other issues led Germany's first postwar chancellor, Konrad Adenauer, to conclude that “Moscow's policies were aimed at absorbing additional European territory.”53 Although tensions abated somewhat after tentative steps toward détente in 1963, the prospect of conflict with the Soviets persisted throughout the 1960s.54 West Germany and the Soviet Union repeatedly failed to come to agreement on a nonaggression pact prior to 1970 and, as a result, German threat assessment focused on growing Soviet military capabilities.55 Citing Soviet medium-range ballistic missile deployments, Adenauer's successor, Ludwig Erhard, declared that Germany was “threatened perhaps more than any other country in Europe,” and sought “to be defended with the same weapons as threaten us.”56 And as Erhard's successor Georg-Kurt Kiesinger noted in a letter to President Lyndon Johnson, “[T]he Soviet forces amassed in the other part of Germany … give rise to great concern among the German people.”57

Moreover, in a bipolar international system where the Soviet Union was its primary adversary, the Federal Republic had no hope for realignment, a constraint that successive West German leaders understood. As Adenauer observed, “[T]here are only two great powers, the Soviet Union and the United States.”58 And responding to a suggestion in 1961 that he openly criticize the United States over its West Berlin policy, Adenauer replied: “We cannot allow ourselves to upset the Americans. We need them desperately.”59 Because alignment with Moscow was incompatible with Germany's democratic political system, the Federal Republic had no choice but alliance with the United States.

The Federal Republic's military dependence had seemingly paradoxical effects. On the one hand, it made the country's leaders highly sensitive to changes in the credibility and durability of U.S. security guarantees, notwithstanding the U.S. strategic interest in West German security. Insofar as West German leaders believed that those guarantees could weaken, they sought to escape dependence by acquiring an indigenous nuclear deterrent. Dependence, however, also made Germany susceptible to U.S. pressure by enabling the coercive threat of military abandonment. This dynamic—of German aversion to military dependence and U.S. exploitation of it—would be the primary force shaping West German nuclear policy from 1954 to 1969.

GERMAN FEARS OF ABANDONMENT AND NUCLEAR AMBITIONS, 1954–60

During Dwight Eisenhower's administration, anxieties about U.S. protection triggered West German nuclear ambitions, but—contrary to the logic of nuclear dependence—these fears persisted despite strong U.S. signals of recommitment. As part of negotiations in 1954 over the Federal Republic's entry into NATO, the country had pledged not to produce nuclear weapons on its territory. Far from being voluntary, however, this promise itself originated with Germany's military dependence. Adenauer understood that without the non-production pledge, France would not permit German entry into NATO, but he held back from making it until the last possible moment and hoped that the restriction would eventually atrophy. In the interim, because West Germany could not indigenously construct nuclear weapons, the country remained dependent on collaboration with its allies to satisfy its nuclear ambitions.60

These ambitions initially emerged as a consequence of doubts regarding the durability of U.S. security guarantees. President Eisenhower believed that European security “cannot always and completely depend on the U.S.,” and he saw American deployments as a temporary measure meant to bridge the force imbalance with the Soviets while the Europeans rebuilt their militaries.61 Eisenhower never intended to abandon Europe to Soviet conquest, and he opposed a precipitous withdrawal of U.S. troops, but the expression of his long-term aim of redeploying forces triggered German fears of abandonment.62 Because West Germans saw U.S. deployments as a signal of the United States' security commitment, the July 1956 leak of the Radford plan—a proposal for redeploying U.S. conventional forces from Europe and relying instead on nuclear deterrence—led Adenauer to conclude that the “United States does not feel itself to be strong enough to keep pace with the Soviet Union.”63 No longer confident in the durability of U.S. commitments, Adenauer concluded, “NATO is finished,” and expressed his interest in acquiring an independent nuclear deterrent.64

At a cabinet meeting in September 1956, Adenauer declared: “Germany cannot remain a nuclear protectorate” of the United States, and he argued that by joining EURATOM—the European Atomic Energy Community—Germany could one day “get at nuclear weapons in the normal way,” meaning via domestic production.65 Although he continuously adjusted his tactics—for example, by publicly denying German nuclear ambitions—Adenauer's broader nuclear aims changed little over the course of his tenure in office. For Adenauer, it was “intolerable” that only the United States and the Soviet Union had substantial nuclear forces. Germany had to have them, too.66

Despite the continuing presence of U.S. troops in Europe, West German fears deepened as a result of growing Soviet missile capabilities, which exacerbated doubts about the credibility of U.S. extended deterrence. In particular, the launch of the Sputnik 1 satellite in 1957 produced fears throughout Europe that American nuclear superiority was declining and that the United States would not risk the destruction of Washington or New York to protect Paris or Bonn.67 These fears led French leaders to initiate talks with West Germany and Italy in November 1957 on joint production of nuclear weapons, with the aim of enabling a European catalytic strategy. As French Defense Minister Jacques Chaban-Delmas noted, by acquiring nuclear weapons, Europeans could ensure that “an attack on Germany or France would mean immediate and total war for the whole of the Alliance.”68 Adenauer supported this aim, saying of nuclear weapons: “We must produce them.”69

Eisenhower and Dulles—who were informed of these negotiations— responded to European fears by attempting to strengthen trust in NATO's credibility.70 Accordingly, in addition to indefinitely placing on hold plans for conventional redeployments, Eisenhower proposed the NATO nuclear stockpile plan in December 1957. As part of this arrangement, Germany acquired direct access to American nuclear weapons, with approximately 500 bombs under “dual-key” European control by the end of the decade. In fact, American custody was “illusory”—allied aircraft armed with U.S. nuclear weapons were often guarded by a single American sentry, and allied military officers often held both keys to U.S.-armed missiles.71 In the event of a Soviet attack, there was virtually nothing to stop European countries from launching a nuclear riposte without U.S. authorization. Eisenhower favored this arrangement and believed it was important that the United States “not deny to our allies what our enemies … already have.”72 By giving U.S. allies direct control over nuclear weapons, the United States would give them a voice in their use, bolstering the credibility of extended nuclear deterrence.

According to the logic of nuclear dependence, these unprecedented recommitment mechanisms should have produced confidence in U.S. security guarantees and inhibited West German nuclear ambitions. Historical evidence refutes this prediction. In April 1958, several months after the introduction of the NATO stockpile plan, Adenauer signed the completed France-Italy-Germany agreement. Although its stated goal was nuclear energy cooperation, the actual purpose was “an exchange of … German money and expertise for access to the nuclear warheads eventually developed under French leadership.”73 Germany's nuclear behavior thus contradicts the expectations of the logic of nuclear dependence. Despite having a formal security guarantee backed by force deployments and devolution of nuclear authority, Adenauer could not be confident that NATO would persist indefinitely or that the United States would always act on its threats. Given the existential threat from the Soviet Union, Adenauer sought an independent deterrent as insurance.

Despite West Germany's hopes, joint production of nuclear weapons with France never took place. After the collapse of the Fourth Republic, President Charles de Gaulle canceled the secret collaboration and prolonged German dependence on the United States. Continued dependence prompted Adenauer to press the Eisenhower administration for additional nuclear sharing as a second-best alternative to an independent arsenal. Unfortunately for Adenauer, however, Eisenhower's liberal approach to nuclear sharing encountered opposition in Congress and the State Department, both of which feared that a European nuclear force without a U.S. veto over launch authority would devolve into a German national nuclear capability.74 With the election of John F. Kennedy, this concern would soon be mirrored in the White House.

U.S. CONTROL AND FRANCO-GERMAN NUCLEAR COOPERATION, 1961–63

Soon after coming to office, Kennedy reneged on the Eisenhower administration's liberal nuclear-sharing plans, replacing them with a policy of centralized U.S. authority over NATO nuclear decisionmaking. To ensure that American nuclear weapons in Europe could not be launched without his authorization, Kennedy equipped them with Permissive Action Links and demurred when Adenauer pressed him to fulfill Eisenhower's nuclear-sharing pledges.75 Kennedy also sought to retain a veto over nuclear use by a mix-manned NATO nuclear fleet. The idea for a multilateral force (MLF) had emerged in 1960 to assuage congressional fears about nuclear sharing, but whereas Eisenhower hoped the MLF could eventually operate independently, Kennedy privately opposed conceding U.S. control.76 As a result, Adenauer correctly came to believe that the MLF was an American ploy, intended to placate European demands for nuclear weapons without surrendering operational authority over their use.77

The administration's emphasis on centralized nuclear authority originated in part with U.S. concern that national nuclear forces would give allies the ability to trigger a nuclear war independent of the United States, a prospect Secretary of State Dean Rusk described as “frightening.”78 In addition, Kennedy and his advisers hoped to avert the broader risk of war by settling outstanding disputes with the Soviet Union.79 Administration officials understood that the Soviets felt a deep-seated anxiety about autonomous West German control over nuclear weapons, so negotiating a political settlement required German leaders “to swallow a lot of things that they had hitherto maintained were entirely unacceptable to them.”80 Chief among these were restrictions on the FRG's nuclear armament, and Kennedy informed the Soviets of his willingness to concede this point if Khrushchev recognized Western rights in Berlin.81

Adenauer had not given up on the hope of a nuclear deterrent under West German national control, so Kennedy's negotiating position generated indignation. Adenauer saw the proposal as “disastrous, even impossible,” and argued it would “represent a constant invitation to the Soviet Union to advance further into Western Europe.”82 In meetings with Kennedy, Adenauer disingenuously denied seeking nuclear weapons production, but reaffirmed his opposition to nuclear renunciation, arguing that the 1954 pledge did not bar the FRG from possessing nuclear weapons, only producing them.83

For Adenauer, renouncing the nuclear option meant that West Germany would be permanently dependent on the United States, an unnerving prospect given his fears about the credibility of the U.S. extended nuclear deterrent. As he later told Rusk, “[T]he possibility might arise where [the U.S. president] would be incapable of taking a decision on the use of nuclear weapons”—for West Germany, influence over this decision was “a matter of survival.”84 Adenauer was thus unsettled by Kennedy's demands for exclusive U.S. nuclear authority and by his concurrent opposition to German control of nuclear weapons. These concerns were exacerbated when Kennedy gave an interview to the Soviet newspaper Izvestiia in which he extolled NATO supremacy over West Germany's armed forces and openly expressed his objections to the FRG's acquisition of a national deterrent.85 In March 1962, Kennedy again conveyed his willingness to concede German nuclear rights in a “principles paper” submitted to Soviet Foreign Minister Andrei Gromyko.86

After hearing of the paper, Adenauer decided to reduce German dependence on the United States by establishing closer relations with France, a policy that culminated in the signing of the Élysée Treaty of friendship in January 1963. As Adenauer told his cabinet, “[W]e must be prepared to live for a number of years in a state of tension with the Americans.”87 Both de Gaulle and Adenauer were upset at U.S. détente overtures toward the Soviet Union and at U.S. nuclear policy.88 Both leaders also believed that Kennedy's threats to reduce conventional deployments in Europe unless Germany agreed to offset U.S. military spending portended the eventual breakup of NATO.89

Closer relations between France and Germany carried implications for the FRG's nuclear status. In a private conversation in 1960, de Gaulle had assured Adenauer that the Federal Republic could not remain nonnuclear in perpetuity, and that “the day would come when there would be no discrimination as regards nuclear weapons for the two countries.”90 And at a public press conference in January 1963, de Gaulle rejected the MLF and openly stated that Germany had the right to decide its own nuclear status, an argument he repeated over the following months to senior U.S. officials. De Gaulle also told Adenauer that he understood Germany's nuclear ambitions, anticipated that its nonnuclear status would eventually end, and would not oppose these developments.91 This support was not merely rhetorical. U.S. documents from the time indicate that American and British officials were aware of evidence that Germany and France had discussed nuclear cooperation throughout 1962, including German financing of French uranium enrichment facilities.92 And in a cabinet discussion one day after de Gaulle's press conference, Adenauer expressed the treaty's nuclear subtext directly, asking his ministers, “Should I tell [de Gaulle] that we are prepared to construct a nuclear weapon together with him?”93

While Adenauer had previously toed a pro-American line, Kennedy's demands for West German nuclear concessions encouraged the chancellor to establish closer relations with France, with the aim of eventually purchasing French-built nuclear weapons.94 Adenauer was aware of West Germany's vulnerability to U.S. pressure, but he believed that by describing his policy merely as a historic rapprochement between France and Germany, he could conceal his true aims and avoid an open break with Kennedy, while expanding West Germany's nuclear options. This explains why Adenauer secretly encouraged de Gaulle to speed development of France's nuclear arsenal while openly agreeing to the U.S.-proposed MLF.95 The sooner France achieved a retaliatory capability, the sooner Germany could reduce its dependence on the U.S. nuclear umbrella by purchasing French nuclear bombs for an independent German deterrent.96

Unfortunately for Adenauer, the White House was aware of the hidden purpose of closer Franco-German relations. Speaking with the British defense minister in 1962, Kennedy warned of his suspicions of French-German nuclear cooperation, which he saw as a direct challenge to U.S. leadership.97 As he told French Cultural Affairs Minister André Malreaux, a Europe “in which we should have to bear the burden of defense without the power to affect events” was unacceptable. Kennedy then suggested several times that if it was no longer welcome in Europe, “the United States would like nothing better than to leave.”98 In a meeting with Adenauer later that year, National Security Adviser McGeorge Bundy conveyed a similar message, reminding the chancellor that Europe needed to rely exclusively on U.S. extended deterrence.99 Adenauer continued to doubt U.S. commitments, however, and remained angry over U.S. willingness to entrench German dependence in an agreement with the Soviets. These positions notwithstanding, U.S. alliance coercion would soon force Adenauer to accept Germany's nonnuclear status.

U.S. ALLIANCE COERCION, 1963

The Kennedy administration opted to employ alliance coercion in response to de Gaulle's public rejection of the MLF and his encouragement of West German nuclear weapons ambitions. Kennedy understood that the basis for U.S. influence was its “military defense of Europe,” and expressed his intention to use this last remaining sanction to force German rejection of French leadership. Given France's acquisition of a nuclear arsenal, the United States had little leverage over de Gaulle, but German dependence meant that the United States could still effectively pressure the FRG into repudiating cooperation with France and affirming its continued subordination to NATO.100 According to Kennedy, Germany needed to understand that it “faced a choice between working with the French or working with us”—Adenauer's double game would no longer be acceptable.101

To make the Germans feel “nervous about [U.S.] relations with them,”102 Kennedy exploited domestic political pressure for retrenchment as a means of bolstering his coercive leverage, informing Adenauer that American public opinion could force the president to end the U.S. military presence in Europe if the Federal Republic continued its cooperation with France.103 Rusk conveyed the same message to the West German foreign minister and reminded Germany's ambassador that the weakness of the French nuclear arsenal and Germany's vulnerable geography left the militarily dependent country little choice other than to accept U.S. demands.104 Kennedy and his advisers believed that forcing the Germans to reject the Élysée Treaty outright would be difficult, so they sought the more limited goal of a preamble affirming the supremacy of NATO and Germany's alliance with the United States.105 By ratifying an amended treaty, the Germans would reject nuclear cooperation with France while affirming their reliance on the U.S. security guarantee.

For these reasons, Adenauer opposed a preamble, but in addition to U.S. blackmail, the chancellor had to contend with Kennedy's domestic meddling.106 Kennedy encouraged the West German finance minister, Ludwig Erhard, to publicly challenge Adenauer's French policy and, more broadly, the chancellor's leadership of the Christian Democratic Union (CDU). Erhard opposed the Élysée Treaty precisely because it risked rupturing relations with the United States, and he accused Adenauer of trying to substitute “dependence on the French for dependence on the Americans”—for Erhard, it was clear “under which dependence we would feel more secure.”107 Emboldened by Kennedy's support, Erhard mobilized the chancellor's anti-Gaullist critics in the CDU to demand a preamble as part of the ratification process. When it became clear to Adenauer that the treaty would not pass without a preamble, he consented, concluding: “Better a treaty with a preamble than no treaty at all.”108

Adenauer's behavior contradicts the logic of nuclear dependence while corroborating the logic of alliance coercion. This is clearest in West Germany's agreement to the MLF despite Adenauer's belief that it was a “mirage” intended to keep the Europeans from acquiring independent control over nuclear weapons.109 Adenauer wanted a continental force that was not subject to a U.S. veto, but in light of U.S. coercion, he had little choice but to accept the MLF as a consolation prize. Even with this concession, however, Adenauer was unable to repair his relationship with Kennedy. Without Kennedy's support, Adenauer could not retain his standing in the CDU, leading to his replacement with Erhard in April 1963, though Adenauer would stay in office as a lame duck until later in the year. By exerting pressure on West Germany, Kennedy was not only able to contain the country's nuclear ambitions, but was also able to bring to power leaders whose preferences on NATO nuclear questions more closely matched his own.

Having prevented a German turn to France, Kennedy next pressed the Federal Republic into signing the Limited Test Ban Treaty. Senior policymakers in the United States, Great Britain, France, and the Soviet Union saw a test ban as a means of forestalling further nuclear proliferation.110 As Kennedy told Khrushchev in 1961, while “a nuclear test ban would not of itself lessen the number of nuclear weapons possessed” by the existing nuclear powers, it “would make development of nuclear weapons by other countries less likely.”111 A test ban thus permitted the superpowers to come to an implicit bargain based on the March 1962 principles paper, in which Soviet acceptance of the status quo in Berlin was exchanged for the United States' agreement to control Germany's nuclear armament.112 Negotiated in secret and announced in July 1963, the treaty surprised Adenauer, who opposed signing it, attacking it in letters to Kennedy and in public, where he described it as “senseless and pointless.”113 Adenauer understood that the treaty would reinforce Germany's nuclear dependence, but Kennedy's earlier success at forcing West German repudiation of nuclear cooperation with France made it clear to Adenauer that he had little hope of not signing the test ban. As Bundy informed the new foreign minister, Gerhard Schröder, it was “unthinkable” that the FRG would not sign, and Adenauer's close parliamentary ally Heinrich Krone acknowledged the impossibility of resisting U.S. pressure, lamenting that Germany “was the victim of America's détente policy.”114

In return for repudiating French nuclear cooperation and signing the test ban, Germany received pledges of continued American protection and a commitment to nuclear consultation. The former was embodied in the U.S. promise to maintain deployment of six divisions in Europe, while the latter took the form of German participation in the MLF.115 By underscoring these commitments, the United States created a linkage between its security guarantee and Germany's nonnuclear status. So long as the United States promised to protect the FRG—while threatening to leave if Bonn tried to acquire nuclear weapons—Germany would reluctantly forgo a nuclear deterrent under national control. But if West German leaders came to believe that protection would be withdrawn irrespective of Germany's nuclear choices, their interest in retaining the nuclear option would persist.

U.S. COERCIVE CONDITIONALITY AND GERMANY'S NUCLEAR OPTION, 1963–66

Throughout the Johnson years, renewed anxieties that the United States would unilaterally reduce its commitment to West Germany undermined the country's reluctant acquiescence to U.S. nonproliferation demands. These anxieties originated with international opposition to the MLF and with U.S. threats to withdraw its conventional forces unless Germany offset their cost. By tying the clearest signal of U.S. protection to an ancillary financial dispute, Johnson weakened the linkage between U.S. security guarantees and Germany's nuclear restraint. As the logic of alliance coercion expects, West Germany responded by assiduously seeking to preserve its weapons option.

From its inception, the MLF faced opposition from Germany's allies and adversaries, undermining the viability of the planned force and German confidence in U.S. extended deterrence.116 Kennedy himself saw the MLF as a “facade”—intended to placate German demands for participation in nuclear strategy without conferring control over nuclear use—but he nevertheless felt that abandoning the plan would undermine Germany's nonnuclear status.117 Rather than cancel the project, Kennedy chose to delay it, but pressure from the State Department led Johnson to revive the idea in April 1964 as a way of satisfying German demands for nuclear participation while keeping the Federal Republic “on a leash.”118 Johnson's decision raised German hopes of signing an MLF charter by the end of the year, but British resistance led Bundy to advise that the United States “let the MLF sink out of sight,” which Johnson did in December.119

Johnson's abandonment of the MLF undermined Germany's faith in the credibility of U.S. nuclear retaliation.120 Erhard was angered and embarrassed by Johnson's decision, and the resulting firestorm in bilateral relations led U.S. Ambassador George McGhee to write that “the atmosphere [in Bonn] is the worst I have encountered” in two years.121 By weakening this signal of its deterrent credibility, the United States rekindled fears in the Federal Republic that the NATO security guarantee was hollow, undermining its willingness to remain nonnuclear. As Erhard made clear, without a solution to nuclear sharing, Germany would seek to preserve the option of acquiring nuclear weapons. In a January 1966 meeting with Henry Kissinger—then a White House consultant—Erhard argued that his support for a NATO nuclear force was based on his desire for an agreement on formal military consultation. By securing U.S. consultation commitments, Erhard hoped to reduce German anxieties that the United States would remove its nuclear weapons from Europe. More provocatively, Erhard hinted that a “hardware solution” in the form of a NATO nuclear force could offer a back door to a “nuclear option for Germany,” if the United States were ever to leave Europe.122

In addition to its demands for nuclear consultation, Bonn also vociferously opposed a nonproliferation treaty.123 The two issues were closely linked: the German ambassador told U.S. officials that without agreement on the “nuclear problems of [the] alliance,” Germany could not acquiesce to the nascent NPT.124 Absent guarantees that Germany would have influence over NATO's decision to use nuclear weapons, West German leaders were not prepared to permanently renounce their weapons option. Opposition to nonproliferation ran even deeper than this conditional linkage suggests. As Erhard bluntly informed Johnson in 1965, it was “impossible to assume that Germany will go forever without a nuclear deterrent.”125

In parallel to the problem of West Germany's nuclear role, a dispute about the continued presence of American conventional forces in Germany threatened to weaken NATO and undermine the FRG's nonnuclear status. In 1961 Germany had agreed to offset the costs of U.S. military deployments in Europe, but by 1963, this agreement had threatened to unravel. Kennedy successfully lobbied German leaders to commit to a linkage between offset payments and the continued presence of U.S. personnel, but by 1965, Bonn had fallen behind in its payments, leading Defense Secretary Robert McNamara to threaten that anything less than a full offset would lead to a complete withdrawal of U.S. troops.126 Erhard was under pressure from an economic slow-down to reduce payments, and in a July 1966 letter to Johnson, he warned that Germany would not be able to meet its offset obligations.127 While McNamara pushed Johnson to take a hard line, Deputy National Security Adviser Francis Bator and Ambassador McGhee warned of the trade-off between pressing the Germans to renounce their nuclear option and demanding a full offset.128

By linking threats of military abandonment to the offset dispute, McNamara and Johnson undercut the linkage between U.S. protection and West Germany's nonnuclear status. West German Ambassador Heinrich Knappstein warned of a growing “neo-isolationist trend” in U.S. politics, and Foreign Minister Schröder complained that rumors of American withdrawal plans were endangering U.S. relations with Germany.129 When Johnson responded to Erhard's letter by proposing trilateral negotiations with the United Kingdom to resolve both the offset and nuclear issues, Erhard balked, concluding that the United States was looking to enlist British help to gang up on Germany130 Undersecretary of State George Ball recognized the dangers of this contentious atmosphere: the Germans harbored a “dark suspicion that the U.S. no longer regards the defense of Europe as vital,” and there was “growing conviction that de Gaulle is right that the U.S. cannot be trusted.”131

Together with U.S. withdrawal plans and pressure for an NPT, the lack of a consultation solution was undermining Germany's faith in the U.S. security guarantee and, as a result, its nonnuclear status. As Erhard told Johnson at the September 1966 summit meeting, “[T]he German people wanted to know where exactly they stand at this point in terms of security, [and] how they would be protected in the event of a non-proliferation treaty.” Referring to the MLF, Erhard acknowledged that “nobody was expecting a ‘hardware solution’ any longer,” but the FRG still wanted a “voice” and a “common solution under NATO.”132 Erhard informed the U.S. ambassador to the trilateral talks, John McCloy, that “questions were being asked as to just when the U.S. intended to apply the nuclear riposte and how much of Germany had to be overrun” before the U.S. struck back.133 So long as the United States continued to subvert the assurances it had provided in 1963, Erhard had incentives to preserve Germany's future options by opposing any nonproliferation accord.

NATO PROTECTION, U.S. COERCION, AND FRG NUCLEAR RENUNCIATION, 1966–69

West Germany's ultimate acquiescence to U.S. nonproliferation demands followed both U.S. pressure and the resolution of ancillary disputes over American security commitments. Contrary to the logic of nuclear dependence, however, Federal Republic leaders continued to harbor doubts about the long-term durability of NATO, and they were forced to accept nuclear renunciation by explicit U.S. threats of military abandonment, consistent with the logic of alliance coercion.

To satisfy West German demands for nuclear consultation, Johnson organized German participation in the Nuclear Planning Group. Control over the decision to launch a nuclear strike remained in U.S. hands, but by agreeing to consult with the Federal Republic on nuclear deployments, targeting, and crisis management, the United States increased German confidence in the American deterrent, and it mitigated the most acute concerns that Washington would fail to act on its nuclear guarantee or withdraw it altogether. After the first meetings of the group, Germany agreed to give up the MLF in exchange for a permanent seat on the new consultative body.134 But even though the September conference resolved the question of nuclear consultation, Johnson's continued pressure on Erhard for a full financial offset prolonged the crisis, causing German decisionmakers to worry that imminent U.S. and British force withdrawals would lead to a “lasting weakening of the Atlantic Treaty” and expose Germany to Soviet military pressure.135 Johnson hoped to placate domestic pressure from Senate Majority Leader Mike Mansfield to drastically reduce U.S. overseas deployments, but his strategy backfired— Erhard's coalition collapsed, and negotiations were delayed for months as Kiesinger formed his new government.136

Germany had previously agreed to forestall its nuclear ambitions in return for U.S. security guarantees, but Johnson was undermining that agreement by linking force deployments to ancillary disputes over money. In 1966, West Germans were still anxious that intra-alliance spats were a prelude to NATO's imminent collapse, and negotiations to resolve the disagreements had stalled. Moreover, the Johnson administration was divided on how to proceed: McNamara wanted to withdraw two divisions if Germany refused to fully offset U.S. spending, but McCloy warned that any withdrawals would have disastrous effects: “No single factor is more directly the symbol of U.S. intent and guarantee than our troops in Europe.”137

Given West Germany's acute anxiety about the durability of U.S. protection, its leaders balked at the prospect of permanently renouncing their nuclear option. As such, when the United States and the Soviet Union revealed their first draft of the NPT in late 1966, the new German government was outraged. Kiesinger—by then the chancellor in a grand coalition with Willy Brandt's Social Democratic Party—declared that the treaty constituted “atomic complicity” between the superpowers at Germany's expense.138 In addition, he told de Gaulle in April 1967 that the Federal Republic intended to refuse to sign the treaty and hoped for a common French-German effort against it. As Kiesinger informed Vice President Hubert Humphrey, the treaty constituted “the most difficult problem that has emerged in a long time” between the United States and Germany.139

Given U.S. opposition to even moderate nuclear-sharing arrangements, German leaders understood that the likelihood of acquiring control over nuclear weapons was very low. As U.S. Ambassador McGhee said, “We could not permit it—i.e., we would withdraw our forces and support for Germany first.”140 But given German anxieties about the durability of NATO, Kiesinger and Brandt were not yet ready to permanently relinquish their country's nuclear option. Speaking with McCloy, Kiesinger warned of a rising sense in Germany “that NATO was in danger of developing into a sort of ‘shell with no real spirit left in it.’”141 Kiesinger suggested that the United States and Germany could “come to agreement” on the treaty, but only if NATO persisted. As he told Johnson, “In the solution of this problem [of the NPT] … my government bears in mind the continued existence and the durability of our alliance.”142 Johnson—ultimately recognizing the incompatibility of U.S. offset demands with its nonproliferation goals—gave in, accepting a meager German offset in return for a token redeployment of U.S. forces.

Johnson's decision eased fears of NATO's imminent collapse, but Germany's anxiety about the durability of U.S. security guarantees persisted. As a State Department paper argued, Germans were still worried that the United States would “progressively reduce its forces in Europe,” exposing the FRG to Soviet coercion.143 These anxieties had a deleterious effect on U.S.-German NPT consultations, which had begun in January 1967.144 German leaders worried that an NPT of unlimited duration would outlast NATO, exposing their country to Soviet “nuclear blackmail.”145 These concerns were significant enough for the Germans to request that the United States make a separate commitment to the nuclear defense of the Federal Republic in case of NATO's collapse.146 Despite being told that West Germany could leave the treaty if that occurred, Kiesinger and Brandt's attitude to the NPT was at best one of “grudging acceptance.”147

To ensure the Federal Republic's signature of the treaty, U.S. officials once more brandished the threat of military abandonment. In a conversation with CDU Parliamentary leader Rainer Barzel, National Security Adviser Walt Rostow raised a “simple fact”—that the FRG depended on the United States and that a decision to acquire nuclear weapons would “tear apart the alliance” and expose Germany to “a very difficult period during which you might well be destroyed.”148 Together with the resolution of intra-alliance disputes, these coercive threats created unsustainable pressure on West German leaders to sign the treaty. Even the treaty's opponents on the German right acknowledged that an “absolute no” on the NPT would be as difficult as a “yes.”149

Thus, even prior to the election of Willy Brandt, U.S. pressure had ensured that German signature was only a matter of time. Brandt's election, combined with Soviet messages that signature would speed negotiations on a nonaggression treaty, accelerated the process, leading Germany to sign the treaty in November 1969.150 Although the Federal Republic would not formally accept International Atomic Energy Agency safeguards until it ratified the NPT in 1975, by finally signing, West Germany codified a tacit bargain with the United States in which it renounced nuclear weapons in exchange for lasting promises of U.S. protection.151 If West Germany remained nonnuclear, the United States would continue to provide security; any movement away from nuclear restraint, however, would result in the rapid withdrawal of U.S. guarantees.

Response to Potential Objections

This section responds to four potential objections to the preceding analysis: (1) that absent U.S. signals of unreliability, West Germany would not have sought nuclear weapons; (2) that U.S. threats of abandonment could not have been credible given the United States' interest in West Germany's security; (3) that shifts in West German coalitional politics drove Bonn's nuclear decisions; (4) and that competing international dynamics, especially an improving threat environment and emerging nonproliferation norms, were responsible for the country's nuclear renunciation.

First, proponents of the logic of nuclear dependence may argue that the Federal Republic's pursuit of a nuclear deterrent did not result from an aversion to military dependence per se, but instead from U.S. signals of unreliability, especially public expressions of the United States' intention to draw down its conventional forces. According to this argument, if the United States had not signaled its lack of commitment, West Germany would have voluntarily relied on U.S. guarantees. This interpretation is inconsistent, however, with Germany's continued pursuit of nuclear armament after the United States strengthened its security guarantees in 1957. By the end of that year, the United States had not only canceled earlier plans to redeploy its forces from Europe, but had also delegated authority over U.S. nuclear weapons to NATO commanders to bolster the credibility of its extended deterrent. Yet despite these powerful signals of commitment, Adenauer continued to seek joint development of nuclear arms with France and Italy, and he privately stressed that indigenous weapons production was superior to reliance on the U.S. arsenal. As such, even if U.S. actions initially triggered German concern about NATO's longevity, West Germany's preference for nuclear armament even when U.S. guarantees were at their strongest indicates that a broader anxiety about military dependence motivated its decisions.

Second, critics might contend that alliance coercion could not have brought about West Germany's nuclear renunciation, as U.S. threats of abandonment would not have been credible against a vital Cold War ally. This argument assumes that the U.S. interest in containing the Soviet Union should have made Germany confident that the United States would not carry out its withdrawal threats. The preceding historical evidence demonstrates, however, that— notwithstanding the U.S. interest in deterring Soviet conquest—German leaders harbored persistent and severe doubts about the durability of U.S. security commitments, uncertainties that only increased as domestic voices for U.S. military retrenchment grew louder. These doubts bolstered West Germany's interest in expanding its nuclear options, but also enabled U.S. alliance coercion, with President Kennedy explicitly highlighting public opposition to U.S. deployments as a way of making Germany “nervous” about bilateral relations. Both Kennedy and Johnson understood that West German military dependence gave Washington substantial coercive leverage over its client, and they repeatedly reminded German leaders of their country's military vulnerability. Rather than doubting that the United States would progressively reduce its commitments, German leaders feared that it would do so irrespective of West German choices. As a result, U.S. threats of abandonment were highly credible throughout the period in question, enabling Washington to obtain Bonn's begrudging renunciation of nuclear armament.

A third objection suggests that West German domestic politics were responsible for Bonn's nuclear decisions, as evidenced by the signature of the test ban and the NPT following leadership changes in 1963 and 1969, respectively. Evidence indicates, however, that rather than independently driving the decisions to terminate nuclear cooperation with France and accede to the test ban, the 1963 replacement of Adenauer with Erhard was the product of U.S. pressure, Kennedy's mobilization of Adenauer's political opponents, and Erhard's fear of rupturing relations with the United States. Hence, instead of supporting domestic models of proliferation, which argue that coalitional or institutional variables filter the effect of international stimuli, the German case demonstrates that international stimuli drove both domestic political shifts and nuclear decisions.152 Moreover, that the putatively antinuclear Erhard voiced strong reservations about Germany's nonnuclear status when the United States weakened its conditionality further underscores that German decisions followed international imperatives rather than domestic politics. Likewise, although the election of Brandt in 1969 sped signature of the NPT, U.S. threats of abandonment had already led the otherwise pronuclear CDU to concede the impossibility of refusing West Germany's signature. This demonstrates that although domestic changes can encourage nuclear renunciation, they are neither necessary nor—given Brandt's insistence on U.S. protection as a condition for NPT signature—sufficient.

Fourth, critics might object that West Germany's signature on the NPT was the product of alternative international factors, especially nonproliferation norms and an improving security environment. There is little evidence, however, that norms drove West Germany's decisions; to the contrary, Adenauer secretly sought nuclear weapons despite large public protests in 1957 against nuclear armament,153 and Kiesinger's grudging acceptance of the NPT followed both private and public denunciations of the treaty. Similarly, West German politicians conceded the impossibility of refusing West Germany's signature despite publicly noting the persistent military threat from the Soviet Union, with even Brandt referencing the risk Germany faced from Soviet “blackmail.”154 Although the Soviet offer to speed negotiations on a nonaggression pact doubtless accelerated the country's NPT signature, West Germany still faced a menacing threat environment in 1968–69, and it was primarily U.S. alliance coercion that drove the country's reluctant decision to renounce its nuclear arsenal.

Conclusion

This article provides a new way of understanding the inhibiting effect of security guarantees on the spread of nuclear weapons. In contrast to the established logic of the security model, which posits that client states will voluntarily opt for reliance on patrons' security guarantees backed by costly signals of credible commitment, this article argues that military threats and anxieties about the inherent risks of military dependence will spur clients to seek nuclear weapons. To keep anxious clients from acquiring the bomb, a nuclear-armed patron will need not only to provide promises of protection, but also to employ coercive threats of military abandonment linked to demands for nuclear restraint. By doing so, the patron not only lowers the benefits of nuclear acquisition, but also prohibitively raises its costs.

To verify this claim, this article examined the nuclear decisions of the Federal Republic of Germany. Contrary to the idea that West Germany was a model case of voluntary nuclear restraint, this analysis showed that successive German leaders were anxious about the durability and credibility of U.S. security guarantees and sought control over an independent deterrent. Given the U.S. interest in containing the Soviet Union and in light of the costly signals that Washington deployed to reassure its ally, persistent West German apprehension about U.S. protection is puzzling for the logic of nuclear dependence. The Federal Republic's behavior makes sense, however, when considered as a response to the risks of military dependence, which drove Bonn to seek nuclear weapons. Unfortunately for West German nuclear aspirations, the country's military dependence enabled the United States to effectively employ alliance coercion. By threatening to reduce its military commitments if the Federal Republic acquired nuclear weapons while promising to maintain its security guarantees if it did not, the Kennedy and Johnson administrations exploited German dependence to obtain compliance with their nonproliferation demands.

Although the empirical analysis in this article concentrates on West Germany, the article's logic extends to any militarily threatened state with the material capacity to produce nuclear arms, and it helps to explain historical variation in the effect of U.S. security guarantees on nuclear proliferation. For example, scholars have wondered why U.S. extended deterrence failed to keep France and the United Kingdom from acquiring nuclear weapons while seemingly inhibiting the ambitions of countries such as South Korea and Taiwan. The established view of security guarantees cannot explain this variation, because the United States was at least as committed to the defense of Western Europe as it was to East Asia. Resolving this puzzle requires investigating differences in the use of alliance coercion. Specifically, by the time the United States initiated efforts to keep its allies from acquiring nuclear weapons, both France and the United Kingdom had acquired a credible deterrent and could resist potential threats of abandonment. By contrast, in response to South Korean and Taiwanese ambitions, the United States threatened to withdraw its military commitments if either country sought nuclear weapons. Like West Germany, both countries were dependent on U.S. security guarantees, resulting in their compliance with U.S. nonproliferation demands despite lingering anxieties about the durability and credibility of U.S. protection.155

The idea that the strategy of alliance coercion forms a potential basis for nuclear restraint has critical implications for U.S. nonproliferation policy. Unipolarity has increased U.S. structural leeway in determining the United States' degree of overseas commitment, so coercive threats of abandonment will be equally or more credible today than they were during the Cold War.156 Consequently, the success of U.S. efforts to keep militarily threatened clients from acquiring nuclear weapons will hinge on U.S. willingness to employ alliance coercion and on clients' degree of military dependence. In East Asia, North Korea's growing nuclear stockpile and recent military tensions with China have prompted discussions in South Korea and Japan of one day acquiring nuclear weapons.157 Yet despite growing Japanese and South Korean military capabilities, both countries will only become more dependent on U.S. security guarantees if Chinese power continues its rapid expansion, providing Washington with leverage it can use to block its clients' nuclear ambitions. In the case of Japan, the country's putative ability to rapidly construct nuclear weapons might narrow the window of opportunity for alliance coercion, meaning U.S. success will depend on the rapid detection of any Japanese moves toward a nuclear breakout.

In the Middle East, Saudi Arabia has threatened that it could acquire nuclear weapons if Western powers fail to keep Iran from doing so.158 In contrast to U.S. clients in East Asia, Saudi Arabia faces a lower probability of engaging in direct military conflict with its rival. As a result, Riyadh's imperative to acquire nuclear weapons is comparatively weak, and the country is more likely to substitute U.S. extended deterrence for indigenous acquisition. This preference may change, however, if the country's likelihood of military confrontation with Iran rises, leading it to seek nuclear weapons of its own. In that case, U.S. leverage may be limited by Saudi Arabia's reported option of procuring nuclear weapons from Pakistan and by Riyadh's ability to purchase advanced conventional armaments from suppliers other than the United States and its partners.

It is important to emphasize that coercion is not the sole determinant of nuclear nonproliferation, and policymakers concerned with containing the spread of nuclear weapons can turn to other tools when security leverage is low. These include mediating disputes to reduce military tensions, encouraging voices for restraint in countries contemplating nuclear acquisition, and strengthening international controls on the diversion of nuclear technology. If these efforts are successful, nuclear restraint can prevail even in the absence of coercive pressure. The history of nuclear proliferation suggests, however, that for countries facing military threats, domestic opposition to nuclear armament is likely to progressively weaken, while supply-side controls will impose only moderate constraints on technical progress.159 In that case, this article suggests that U.S. coercive leverage will be a critical variable in determining the incidence of further nuclear proliferation. Insofar as the United States reduces its overseas commitments or refrains from employing coercion, militarily threatened allies will be more likely to acquire nuclear weapons of their own.160 But by maintaining its security guarantees while exploiting its clients' military dependence for coercive effect, the United States will continue to underwrite the global nonproliferation order, precluding the grimmest predictions of its demise.

Acknowledgments

For their helpful suggestions and critiques, the author would like to thank Ahsan Butt, Hilary Fruitman, Francis Gavin, Charles Glaser, Colin Kahl, Chad Levinson, Jonathan Markowitz, John Mearsheimer, Nicholas Miller, Miranda Priebe, Joshua Itzkowitz Shifrinson, Dan Slater, Paul Staniland, Rachel Whitlark, participants at the Belfer Center's seminar on international security, and the anonymous reviewers. He is especially grateful to the Stanton Foundation, the Belfer Center, and the Security Studies Program at the Massachusetts Institute of Technology for supporting this research.

Notes

1. 

By “security guarantee,” I mean an explicit promise by one state (a “patron”) to intervene with force in response to military threats against the territory of its ally (the “client”). According to Jeffrey W. Knopf, these commitments are a subset of “security assurances”—promises of military protection or nonaggression. See Knopf, “Security Assurances: Initial Hypotheses,” in Knopf, ed., Security Assurances and Nuclear Nonproliferation (Redwood City, Calif.: Stanford University Press, 2012), pp. 16–17.

2. 

On the security model, see Scott D. Sagan, “Why Do States Build Nuclear Weapons? Three Models in Search of a Bomb,” International Security, Vol. 21, No. 3 (Winter 1996/97), pp. 54–86.

3. 

Etel Solingen, Nuclear Logics: Contrasting Paths in East Asia and the Middle East (Princeton, N.J.: Princeton University Press, 2007); Maria Rost Rublee, Nonproliferation Norms: Why States Choose Nuclear Restraint (Athens: University of Georgia Press, 2009); Jacques E.C. Hymans, The Psychology of Nuclear Proliferation: Identity, Emotions, and Foreign Policy (Cambridge: Cambridge University Press, 2006); and Jacques E.C. Hymans, “Veto Players, Nuclear Energy, and Nonproliferation: Domestic Institutional Barriers to a Japanese Bomb,” International Security, Vol. 36, No. 2 (Fall 2011), pp. 154–189.

4. 

For research that highlights the importance of commitment signals for security guarantees, see the contributions in Knopf, Security Assurances and Nuclear Nonproliferation. See also Clark A. Murdock et al., Exploring the Nuclear Posture Implications of Extended Deterrence and Assurance (Washington, D.C.: Center for Strategic and International Studies, 2009).

5. 

Process tracing uses evidence of timing, sequence, and motivation to test whether purported causes and observed outcomes are linked by expected mechanisms of action in a particular case. See Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge, Mass.: MIT Press, 2005).

6. 

A “hard test” is one in which countervailing factors reduce the expected effect of a theory's variables. “Easy tests” are those in which a theory suggests that a particular outcome would be especially likely. West Germany is also a “double-decisive” test, as the competing arguments of security guarantees make predictions that are both unique and certain. See Stephen Van Evera, Guide to Methods for Students of Political Science (Ithaca, N.Y.: Cornell University Press, 1997), pp. 30–32.

7. 

Sonali Singh and Christopher R. Way, “The Correlates of Nuclear Proliferation: A Quantitative Test,” Journal of Conflict Resolution, Vol. 48, No. 6 (December 2004), pp. 859–885; and Dong-Joon Jo and Erik Gartzke, “Determinants of Nuclear Weapons Proliferation,” Journal of Conflict Resolution, Vol. 51, No. 1 (February 2007), pp. 176–194.

8. 

Philipp C. Bleek, “Why Do States Proliferate? Quantitative Analysis of the Exploration, Pursuit, and Acquisition of Nuclear Weapons” in William C. Potter and Gaukhar Mukhatzhanova, eds., Forecasting Nuclear Proliferation in the 21st Century, Vol. 1: The Role of Theory (Redwood City, Calif.: Stanford University Press, 2010), pp. 159–192; and the data coding in Philipp C. Bleek and Eric B. Lorber, “Security Guarantees and Allied Nuclear Proliferation,” Journal of Conflict Resolution, Vol. 58, No. 3 (April 2014), pp. 429–454. One analysis that acknowledges Germany's “nuclear weapons activities” throughout this period is Harald Müller and Andreas Schmidt, “The Little Known Story of De-Proliferation: Why States Give Up Nuclear Weapon Activities,” in Potter and Mukhatzhanova, Forecasting Nuclear Proliferation, Vol. 1, pp. 124–158.

9. 

Jenifer Mackby and Walter B. Slocombe, “Germany: The Model Case, A Historical Imperative,” in Kurt M. Campbell, Robert J. Einhorn, and Mitchell B. Reiss, eds., The Nuclear Tipping Point: Why States Reconsider Their Nuclear Choices (Washington, D.C.: Brookings Institution Press, 2004), p. 196. For a similar claim, see T.V. Paul, Power versus Prudence: Why Nations Forgo Nuclear Weapons (Montreal: McGill-Queen's University Press, 2000). Both accounts rely on evidence from Catherine McArdle Kelleher, Germany and the Politics of Nuclear Weapons (New York: Columbia University Press, 1975). For a recent divergence from this pattern, see Nuno P. Monteiro and Alexandre Debs, “The Strategic Logic of Nuclear Proliferation,” International Security, Vol. 39, No. 2 (Fall 2014), pp. 7–51.

10. 

In particular, scholars have ignored accounts of German nuclear ambitions in Hans-Peter Schwarz, Konrad Adenauer: A German Politician and Statesman in a Period of War, Revolution, and Reconstruction, Vol. 2: The Statesman, 1952–1967 (Providence: Berghahn, 1997); Marc Trachtenberg, A Constructed Peace: The Making of the European Settlement, 1945–1963 (Princeton, N.J.: Princeton University Press, 1999); and Francis J. Gavin, Nuclear Statecraft: History and Strategy in America's Atomic Age (Ithaca, N.Y.: Cornell University Press, 2012).

11. 

See, in particular, Ariel E. Levite, “Never Say Never Again: Nuclear Reversal Revisited,” International Security, Vol. 27, No. 3 (Winter 2002/03), pp. 59–88; Nicholas L. Miller, “Nuclear Dominoes: A Self-Defeating Prophecy?” Security Studies, Vol. 23, No. 1 (January/March 2014), pp. 33–73; and Nicholas L. Miller, “The Secret Success of Nonproliferation Sanctions,” International Organization, Vol. 68, No. 4 (Fall 2014), pp. 914–944.

12. 

Kenneth N. Waltz, “More May Be Better,” in Scott D. Sagan and Waltz, eds., The Spread of Nuclear Weapons: A Debate Renewed (New York: W.W. Norton, 2003), p. 38.

13. 

Hymans, “Veto Players, Nuclear Energy, and Nonproliferation,” p. 154.

14. 

See Glenn H. Snyder, Alliance Politics (Ithaca, N.Y.: Cornell University Press, 1997), p. 170.

15. 

For example, then-Secretary of State Hillary Clinton suggested this as a method of containing proliferation in response to potential Iranian nuclear acquisition. See Mark Landler and David E. Sanger, “Clinton Speaks of Shielding Mideast from Iran,” New York Times, July 22, 2009, http://www.nytimes.com/2009/07/23/world/asia/23diplo.html.

16. 

Paul, Power versus Prudence, p. 22.

17. 

Benjamin Frankel, “The Brooding Shadow: Systemic Incentives and Nuclear Weapons Proliferation,” Security Studies, Vol. 2, Nos. 3–4 (Spring/Summer 1993), p. 46.

18. 

Bleek and Lorber, “Security Guarantees and Allied Nuclear Proliferation,” p. 432.

19. 

On hedging, see Levite, “Never Say Never Again,” pp. 69–73.

20. 

Paul, Power versus Prudence, p. 23. See also Knopf, “Security Assurances.”

21. 

Monteiro and Debs, “The Strategic Logic of Nuclear Proliferation,” p. 10. This prediction is not conditional on the client's level of threat. Consistent with the logic of nuclear dependence, Monteiro and Debs argue that the patron's credibility is highest when it provides “formal pledges of support and deployments of troops or nuclear weapons.” See ibid., p. 16.

22. 

Kenneth N. Waltz, Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979), p. 118

23. 

Charles Glaser argues that self-help logic does not necessarily lead to security competition via balancing, but rather can prompt states to cooperate with adversaries via bilateral arms control. This article only considers countries that have already selected a competitive strategy by participating in an alliance. Instead of predicting whether balancing will occur, it explains the type of balancing that is optimal. See Charles L. Glaser, “Realists as Optimists: Cooperation as Self-Help,” International Security, Vol. 19, No. 3 (Winter 1994/95), pp. 50–90.

24. 

On the idea that dependence on a superpower for military protection is suboptimal to acquisition of an indigenous deterrent, see Avery Goldstein, Deterrence and Security in the 21st Century: China, Britain, France, and the Enduring Legacy of the Nuclear Revolution (Redwood City, Calif.: Stanford University Press, 2000), pp. 21–25.

25. 

John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W.W. Norton, 2001), p. 33. Put differently, “states do not willingly place themselves in situations of increased dependence.” See Waltz, Theory of International Politics, p. 107.

26. 

Robert Jervis, The Meaning of the Nuclear Revolution: Statecraft and the Prospect of Armageddon (Ithaca, N.Y.: Cornell University Press, 1989), pp. 29–35; Bernard Brodie, Strategy in the Missile Age (Princeton, N.J.: Princeton University Press, 1959); and Thomas C. Schelling, Arms and Influence (New Haven, Conn.: Yale University Press, 2008).

27. 

On this view, see Kenneth N. Waltz, “Nuclear Myths and Political Realities,” American Political Science Review, Vol. 84, No. 3 (September 1990), pp. 731–745; John J. Mearsheimer, “Back to the Future: Instability in Europe after the Cold War,” International Security, Vol. 15, No. 1 (Summer 1990), pp. 5–56; and John J. Mearsheimer, “The Case for a Ukrainian Nuclear Deterrent,” Foreign Affairs, Vol. 72, No. 3 (Summer 1993), pp. 50–66.

28. 

I thank an anonymous reviewer for bringing this critique to my attention.

29. 

See Goldstein, Deterrence and Security in the 21st Century, chap. 5. According to Vipin Narang, other states that have achieved survivable second-strike forces are China, India, and Israel. See Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conflict (Princeton, N.J.: Princeton University Press, 2014), p. 22.

30. 

Vipin Narang, “What Does It Take to Deter? Regional Power Nuclear Postures and International Conflict,” Journal of Conflict Resolution, Vol. 57, No. 3 (June 2013), pp. 478–508.

31. 

On this “catalytic” strategy, see Vipin Narang, “Posturing for Peace? Pakistan's Nuclear Postures and South Asian Stability,” International Security, Vol. 34, No. 3 (Winter 2009/10), pp. 41–43.

32. 

This argument follows from the contention that states balance against threats rather than against capabilities. On this claim, see Stephen M. Walt, The Origins of Alliances (Ithaca, N.Y.: Cornell University Press, 1987).

33. 

The behavior of Sweden, which faced a modest probability of conflict with the Soviet Union and benefited from an informal U.S. extended deterrence guarantee, corresponds with this logic. Sweden also depended on the United States for nuclear technology, a benefit it risked losing if it pursued an independent deterrent. See Thomas Jonter, “The United States and Swedish Plans to Build the Bomb, 1945–68,” in Knopf, Security Assurances and Nuclear Nonproliferation, pp. 219–245. On the idea that dependent states restrain their ambitions to avoid sanctions, see Miller, “The Secret Success of Nonproliferation Sanctions.”

34. 

To determine which states had formal defense pacts with the United States from 1946 to 1989, I use codings from the Alliance Treaty Obligations and Provisions (ATOP) dataset. I exclude Israel because ATOP codes it as having received U.S. security guarantees in 1981, after Israel had already acquired nuclear weapons. I also exclude Pakistan, because the United States explicitly refused to provide security guarantees against India, its main rival. On ATOP, see Brett Ashley Leeds et al., “Alliance Treaty Obligations and Provisions, 1815–1944,” International Interactions, Vol. 28, No. 3 (2002), pp. 237–260. On Pakistan, see Husain Haqqani, Magnificent Delusions: Pakistan, the United States, and an Epic History of Misunderstanding (New York: PublicAffairs, 2013). To estimate nuclear technical capability, I included any U.S. client that reached five or more on Jo and Gartzke's seven-point scale of nuclear latency at any point during the alliance. See Jo and Gartzke, “Determinants of Nuclear Weapons Proliferation,” pp. 172–173. For a critique of this scale as a measure of latency, see Scott D. Sagan, “Nuclear Latency and Nuclear Proliferation,” in Potter and Mukhatzhanova, Forecasting Nuclear Proliferation, Vol. 1, pp. 80–101. The countries matching these scope conditions that did not go down the nuclear path were Belgium, Canada, Chile, Colombia, Denmark, Greece, Mexico, the Netherlands, Peru, the Philippines, Portugal, Spain, Thailand, Turkey, and Venezuela.

35. 

For evidence of nuclear ambitions and U.S. pressure, see Mitchell Reiss, Bridled Ambition: Why Countries Constrain Their Nuclear Capabilities (Washington, D.C.: Woodrow Wilson Center Press, 1995), chap. 3; James J. Walsh, “Surprise Down Under: The Secret History of Australia's Nuclear Ambitions,” Nonproliferation Review, Vol. 5, No. 1 (Fall 1997), pp. 1–20; Goldstein, Deterrence and Security in the 21st Century, chaps. 5–6; James J. Walsh, “Security Assurances and Iran: Assessment and Re-conceptualization,” in Knopf, Security Assurances and Nuclear Nonproliferation, pp. 111–136; Leopoldo Nuti, “Me Too, Please: Italy and the Politics of Nuclear Weapons, 1945–1975,” Diplomacy & Statecraft, Vol. 4, No. 1 (March 1993), pp. 114–148; Llewelyn Hughes, “Why Japan Will Not Go Nuclear (Yet): International and Domestic Constraints on the Nuclearization of Japan,” International Security, Vol. 31, No. 4 (Spring 2007), pp. 67–96; Astrid Forland, “Norway's Nuclear Odyssey: From Optimistic Proponent to Nonproliferator,” Nonproliferation Review, Vol. 4, No. 2 (Winter 1997), pp. 1–16; Scott Snyder, “South Korean Nuclear Decision Making,” in William C. Potter and Gaukhar Mukhatzhanova, eds., Forecasting Nuclear Proliferation in the 21st Century, Vol. 2: A Comparative Perspective (Redwood City, Calif.: Stanford University Press, 2010), pp. 158–181; and Derek J. Mitchell, “Taiwan's Hsin Chu Program: Deterrence, Abandonment, and Honor,” in Campbell, Einhorn, and Reiss, The Nuclear Tipping Point, pp. 293–316.

36. 

Bleek and Lorber argue that their findings support the logic of nuclear dependence, but their statistical analysis omits any indicators of coercive pressure, and their examination of the South Korean case actually demonstrates that the United States had to employ threats of abandonment to keep Seoul from acquiring nuclear weapons. See Bleek and Lorber, “Security Guarantees and Allied Nuclear Proliferation,” p. 442.

37. 

On the idea that coercive influence varies with dependence, see Richard M. Emerson, “Power-Dependence Relations,” American Sociological Review, Vol. 27, No. 1 (February 1962), pp. 31–41; Robert A. Dahl, “The Concept of Power,” Behavioral Science, Vol. 2, No. 3 (July 1957), pp. 201–215; and David A. Baldwin, “Interdependence and Power: A Conceptual Analysis,” International Organization, Vol. 34, No. 4 (Autumn 1980), pp. 471–506.

38. 

Glenn H. Snyder, “The Security Dilemma in Alliance Politics,” World Politics, Vol. 36, No. 4 (July 1984), p. 472.

39. 

Although any nuclear-armed state may provide security guarantees, during the Cold War, the United States and the Soviet Union provided the majority of such commitments.

40. 

Earlier nuclear scholarship was based on the idea of “existential deterrence,” which assumes that acquisition of a handful of nuclear weapons is sufficient to deter threats. Vipin Narang, however, demonstrates that states cannot independently deter military aggression without implementing either a posture of assured retaliation or one of asymmetric escalation, both of which require procedures and capabilities beyond the mere possession of nuclear explosive devices. This finding suggests that states remain at least somewhat vulnerable to alliance coercion even after they cross the nuclear threshold. See Narang, Nuclear Strategy in the Modern Era. On existential deterrence, see Marc Trachtenberg, “The Influence of Nuclear Weapons in the Cuban Missile Crisis,” International Security, Vol. 10, No. 1 (Summer 1985), pp. 139–140.

41. 

Even after a nuclear aspirant crosses the nuclear threshold, the nonproliferator may still derive leverage from its provision of military or economic resources. The value of these resources, however, is smaller than that of a security guarantee, reducing the odds that military embargoes or economic sanctions will induce militarily threatened aspirants to reverse their ambitions absent promises of protection. For a fuller discussion, see Gene Gerzhoy, “Coercive Nonproliferation: Security, Leverage, and Nuclear Reversals,” Ph.D. dissertation, University of Chicago, 2014.

42. 

Snyder, “The Security Dilemma in Alliance Politics,” p. 484. On the broader notion that one state's strategic interest in another's security undermines coercive threats of abandonment, see Snyder, Alliance Politics, pp. 168–170.

43. 

Thomas C. Schelling, The Strategy of Conflict (Cambridge, Mass.: Harvard University Press, 1980), p. 187.

44. 

Waltz, Theory of International Politics, p. 169.

45. 

The idea that domestic politics has a greater effect when systemic constraints are weak comes from the neoclassical realist paradigm. See the contributions in Jeffrey W. Taliaferro, Steven E. Lobell, and Norrin M. Ripsman, eds., Neoclassical Realism, the State, and Foreign Policy (Cambridge: Cambridge University Press, 2009).

46. 

Denis Healey, The Time of My Life (London: Michael Joseph, 1989), p. 243.

47. 

Schelling, Arms and Influence, pp. 74–75.

48. 

If these commitments do not exist prior to the bargaining process, they must be supplied during it.

49. 

These numbers exclude trained reservists, as well as personnel in each country's air and naval forces, which would add approximately 885,000 soldiers to West Germany's estimated strength and 1.5 million soldiers to the Soviet Union's. On Soviet military capabilities, see Congressional Budget Office, Assessing the NATO / Warsaw Pact Military Balance (Washington, D.C.: U.S. Government Printing Office [GPO], 1977), pp. 39–42. On West German military strength, see International Institute for Strategic Studies, The Military Balance, Vol. 67 (London: International Institute for Strategic Studies, 1967). Although Western military analysts concluded in the 1960s that NATO could plausibly maintain a collective conventional defense against a Soviet invasion, “U.S. forces [were] needed to maintain the balance.” Thus, even with more optimistic estimates of West German military capabilities, the Federal Republic would have been outgunned without U.S. military assistance. See Alain C. Enthoven and K. Wayne Smith, How Much Is Enough? Shaping the Defense Program, 1961–1969 (New York: Harper and Row, 1971), p. 162.

50. 

Schwarz, Konrad Adenauer, p. 577. From 1965 to 1970, the Soviet military budget increased 40 percent, enabling the Soviet Union to achieve nuclear parity with the United States by 1972. See Svetlana Savranskaya and William Taubman, “Soviet Foreign Policy, 1962–1975,” in Melvyn P. Leffler and Odd Arne Westad, eds., The Cambridge History of the Cold War, Vol. 2: Crises and Détente (Cambridge: Cambridge University Press, 2010), pp. 134–157.

51. 

Erhard-Johnson Meeting, September 26, 1966, Papers of Lyndon B. Johnson: Presidential Papers, National Security File (NSF), National Security Council (NSC) Histories, Trilateral Negotiations, box 50, Lyndon Baines Johnson Presidential Library, Austin, Texas (hereafter LBJL). The presence of U.S. troops also pacified intra-alliance tensions, ensuring NATO's survival as a political entity. See Timothy P. Ireland, Creating the Entangling Alliance: The Origins of the North Atlantic Treaty Organization (Westport, Conn.: Greenwood, 1981).

52. 

Schwarz, Konrad Adenauer, p. 481. On Adenauer's support for nuclear deployments, see ibid., pp. 219, 263. Defense Minister Franz-Josef Strauss believed that because “land forces no longer had their previous, classical importance,” nuclear weapons would allow Germany to deter the Soviet Union's superior conventional power. See Pertti Ahonen, “Franz-Josef Strauss and the German Nuclear Question, 1956–1962,” Journal of Strategic Studies, Vol. 18, No. 2 (June 1995), p. 28.

53. 

On Khrushchev's military ultimatums, see Vladislav Martinovich Zubok and Konstantin Pleshakov, Inside the Kremlin's Cold War: From Stalin to Khrushchev (Cambridge, Mass.: Harvard University Press, 1996), pp. 194–195. On Adenauer's response, see Carl G. Anthon, “Adenauer's Ostpolitik, 1955–1963,” World Affairs, Vol. 139, No. 2 (Fall 1976), p. 120; and Schwarz, Konrad Adenauer, pp. 142–143, 308–309.

54. 

On the “little détente,” see Vladislav Zubok, “The Case of Divided Germany, 1953–1964,” in William Taubman, Sergei Khrushchev, and Gleason Abbott, eds., Nikita Khrushchev (New Haven, Conn.: Yale University Press, 2000), pp. 275–300; William Burr and David Alan Rosenberg, “Nuclear Competition in an Era of Stalemate, 1963–1975,” in Leffler and Westad, The Cambridge History of the Cold War, pp. 88–111; and Marc Trachtenberg, “The Structure of Great Power Politics, 1963– 1975,” in Leffler and Westad, The Cambridge History of the Cold War, pp. 482–502.

55. 

On peace efforts under Adenauer and Erhard, see Anthon, “Adenauer's Ostpolitik, 1955–1963”; Benedikt Schoenborn, “Bargaining with the Bear: Chancellor Erhard's Bid to Buy German Reunification, 1963–64,” Cold War History, Vol. 8, No. 1 (February 2008), pp. 23–53. On Brandt's “New Ostpolitik,” which led to agreement on a formal nonaggression pact, see Julie von Dannenberg, The Foundations of Ostpolitik: The Making of the Moscow Treaty between West Germany and the USSR (Oxford: Oxford University Press, 2008).

56. 

George McGhee to State Department, January 15, 1965, Foreign Relations of the United States (FRUS), 1964–1968, Vol. 13: Western Europe Region (Washington, D.C.: GPO, 1995), p. 175.

57. 

Georg Kiesinger to Lyndon Johnson, April 4, 1967, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL. U.S. analyses concurred that the “enormous strategic nuclear capabilities” of the Soviet Union meant there was a continued risk of Soviet military probes and coercive diplomacy. See John McCloy, “Conclusions and Recommendations of the Trilateral Commission,” November 21, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL. See also Helmut Sonnenfeldt and Raymond L. Garthoff, “Soviet Intentions: Possible Soviet Uses of Military Force in Europe,” October 24, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 51, LBJL.

58. 

Ronald J. Granieri, The Ambivalent Alliance: Konrad Adenauer, the CDU/CSU, and the West, 1949– 1966 (New York: Berghahn, 2004), p. 96.

59. 

Ibid., p. 137.

60. 

Ibid., pp. 81–83.

61. 

NSC meeting, October 7, 1953, FRUS, 1952–1954, Vol. 2, Pt. 1: National Security Affairs (Washington, D.C.: GPO, 1979), p. 527; and NSC Meeting, December 11, 1958, FRUS, 1958–1960, Vol. 7, Pt. 1: Western European Integration and Security, Canada (Washington, D.C.: GPO, 1993), p. 367. Eisenhower's goal was to eventually remove all six U.S. divisions from Europe, but he understood the danger of doing so too rapidly. See NSC Meeting, August 18, 1959, FRUS, 1958–1960, Vol. 7, Pt. 1, p. 249. See also Trachtenberg, A Constructed Peace, pp. 144–150.

62. 

As Eisenhower said in October 1956, “We cannot take divisions out of Europe at this time. The effect on Adenauer would be unacceptably damaging.” See NSC meeting, August 17, 1956, FRUS, 1955–1957, Vol. 4: Western European Security and Integration (Washington, D.C.: GPO, 1988), p. 99.

63. 

Schwarz, Konrad Adenauer, p. 235.

64. 

Ibid., p. 236.

65. 

Ibid., pp. 239–240.

66. 

Ibid., p. 264. Adenauer believed that there was nothing unusual about his nuclear ambitions. When, in April 1957, the Soviet ambassador to Bonn raised the possibility that Germany could one day gain a nuclear capability, Adenauer did not deny it. The German foreign minister asked, “If England and other powers have atomic weapons, why should the FRG not have them?” Quoted in Zubok and Pleshakov, Inside the Kremlin's Cold War, pp. 195–196.

67. 

As French Prime Minister Félix Gaillard said in December 1957, the “Soviets have caught up technically and passed [the] U.S. in [the] atomic and missile fields. All [are] now in direct danger [of a] paralyzing attack.” See Telegram from the United States Delegation at the NATO Heads of Government meeting to the Department of State, December 17, 1957, FRUS, 1955–1957, Vol. 4, pp. 232–242.

68. 

Schwarz, Konrad Adenauer, p. 321

69. 

Ibid., p. 319.

70. 

Adenauer personally informed Dulles. See ibid., p. 322. On the Eisenhower administration's support for nuclear sharing, see Trachtenberg, A Constructed Peace, especially pp. 176, 207–210.

71. 

Ibid., pp. 193–194, 209. See also John D. Steinbruner, The Cybernetic Theory of Decision: New Dimensions of Political Analysis (Princeton, N.J.: Princeton University Press, 1974), pp. 182–183; and Robert S. Norris, William M. Arkin, and William Burr, “Where They Were,” Bulletin of the Atomic Scientists, Vol. 55, No. 6 (November/December 1999), p. 30.

72. 

Trachtenberg, A Constructed Peace, p. 246.

73. 

Ahonen, “Franz-Josef Strauss and the German Nuclear Question, 1956–1962,” p. 33.

74. 

Eisenhower hoped that NATO would become a “third great power bloc,” a hope that Adenauer wholeheartedly shared. See Trachtenberg, A Constructed Peace, pp. 147–156; and Schwarz, Konrad Adenauer, pp. 471–474.

75. 

Schwarz, Konrad Adenauer, pp. 473, 568, 631; Mackby and Slocombe, “Germany,” p. 186; and Kelleher, Germany and the Politics of Nuclear Weapons, p. 140.

76. 

Both Kennedy and Johnson hinted at joint control in discussions with European allies, but the Bowie report—which first articulated the idea of the MLF—intimated that it would never operate except with a U.S. veto. See Frank Costigliola, “Lyndon B. Johnson, Germany, and ‘the End of the Cold War,’” in Warren I. Cohen and Nancy Bernkopf Tucker, eds., Lyndon Johnson Confronts the World: American Foreign Policy, 1963–1968 (New York: Cambridge University Press, 1994), pp. 173– 210; and Robert R. Bowie, The North Atlantic Nations: Tasks for the 1960s, Vol. 7 (College Park: University of Maryland, 1991).

77. 

Adenauer was especially skeptical of sea-based forces, which he felt could be withdrawn more easily than land-based ballistic missiles. See Schwarz, Konrad Adenauer, pp. 663–665.

78. 

Rusk-Adenauer conversation, June 22, 1962, FRUS, 1961–1963, Vol. 13: Western Europe and Canada (Washington, D.C.: GPO, 1994), pp. 419–422, at p. 420; and Rusk-Couve conversation, June 20, 1962, FRUS, 1961–1963, Vol. 13, pp. 413–419.

79. 

Zubok and Pleshakov, Inside the Kremlin's Cold War, pp. 245–246.

80. 

Quote from Rusk in Trachtenberg, A Constructed Peace, pp. 304–305. On the U.S. understanding of Soviet fears, see NSC Excomm (Executive Committee) meeting, February 12, 1963, FRUS, 1961– 1963, Vol. 13, pp. 494–502.

81. 

Kennedy-Khrushchev meetings, June 4, 1961, FRUS, 1961–1963, Vol. 14: Berlin Crisis, 1961– 1962 (Washington, D.C.: GPO, 1993), pp. 87–98. Kennedy was even more explicit with Rusk, and Rusk conveyed Kennedy's proposals to Gromyko later that month. See Memorandum (Memo) of Action, September 5, 1961, FRUS, 1961–1963, Vol. 14, p. 393; and Rusk-Gromyko meetings, September 28 and 30, 1961, FRUS, 1961–1963, Vol. 14, pp. 439–441, 459.

82. 

On JFK's promise, see Schwarz, Konrad Adenauer, p. 518. On Adenauer's reaction to the Rusk-Gromyko talks, see ibid., p. 559.

83. 

Kennedy-Adenauer meetings, November 21–22, 1961, FRUS, 1961–1963, Vol. 14, pp. 616–618. See also Trachtenberg, A Constructed Peace, p. 340.

84. 

Rusk-Adenauer conversation, June 22, 1962, FRUS, 1961–1963, Vol. 13, pp. 419–422, at p. 422.

85. 

Trachtenberg, A Constructed Peace, p. 341.

86. 

Rusk-Gromyko meeting, March 22, 1962, FRUS, 1961–1963, Vol. 15: Berlin Crisis, 1962–1963 (Washington, D.C., GPO, 1994), pp. 67–71.

87. 

Schwarz, Konrad Adenauer, p. 612. Adenauer also used two press conferences to snub Kennedy, openly criticizing the whole of his Berlin policy and the negotiating positions he had taken with the Soviets.

88. 

Trachtenberg, A Constructed Peace, pp. 338, 368–371.

89. 

For the authoritative account of these disputes, see Francis J. Gavin, Gold, Dollars, and Power: The Politics of International Monetary Relations, 1958–1971 (Chapel Hill: University of North Carolina Press, 2004).

90. 

Schwarz, Konrad Adenauer, p. 462.

91. 

Trachtenberg, A Constructed Peace, p. 373.

92. 

Ibid., n. 84.

93. 

Schwarz, Konrad Adenauer, p. 670.

94. 

Ibid., pp. 596–598, 718.

95. 

Adenauer informed George Ball of his agreement to the MLF on January 14, 1963. See ibid., pp. 665–666. On July 4, 1963, Adenauer encouraged de Gaulle to proceed rapidly with nuclear weapons development. See ibid., p. 693.

96. 

Adenauer doubted the reliability of French extended deterrence for the same reasons he mistrusted U.S. security guarantees. Thus, he sought to purchase nuclear weapons from France outright. De Gaulle misled Adenauer, however, hinting heavily at nuclear cooperation, while telling the United States and the United Kingdom that France would never allow Germany an independent arsenal. See Granieri, The Ambivalent Alliance, p. 199.

97. 

Gavin, Nuclear Statecraft, p. 42. For more evidence that the United States was aware of Germany's nuclear aims, see Marc Trachtenberg, “Appendices to A Constructed Peace,” Appendix Six: “The U.S. Assessment of German Nuclear Aspirations,” 1999, http://www.sscnet.ucla.edu/polisci/faculty/trachtenberg/appendices/annexlist.html.

98. 

Kennedy-Malreaux meeting, May 11, 1962, FRUS, 1961–1963, Vol. 13, p. 696.

99. 

Schwarz, Konrad Adenauer, p. 629.

100. 

NSC Excomm meeting, January 25, 1963, FRUS, 1961–1963, Vol. 13, pp. 487–491. During this meeting, Kennedy declared that if the Europeans were “getting ready to throw [the] U.S. out of Europe,” he preferred “to march out.”

101. 

NSC Excomm meeting, February 5, 1963, FRUS, 1961–1963, Vol. 13, pp. 173–179.

102. 

Quote from U.S. ambassador Walter C. Dowling. See ibid, p. 175.

103. 

Kennedy started by sending Adenauer a “very agitated” letter on January 22, and his “barely veiled threat” was delivered to Adenauer by Ambassador Dowling on February 4. See Schwarz, Konrad Adenauer, pp. 673–674.

104. 

Rusk told the West German foreign minister, Heinrich von Brentano, that the French-German treaty suggested to the American people that the United States was no longer welcome in Europe, and could make it “impossible” to keep U.S. forces there. See Rusk-Brentano meeting, March 22, 1963, FRUS, 1961–1963, Vol. 13, p. 191. For Rusk's stark description of German dependence, see Trachtenberg, A Constructed Peace, p. 375.

105. 

NSC meeting, February 5, 1963, FRUS, 1961–1963, Vol. 13, pp. 173–179.

106. 

Adenauer understood that a unilateral preamble affirming Germany's partnership with the United States would be seen as a “direct provocation to de Gaulle.” See Schwarz, Konrad Adenauer, p. 675

107. 

Granieri, The Ambivalent Alliance, p. 200.

108. 

See Schwarz, Konrad Adenauer, p. 682.

109. 

Ibid., pp. 663–664.

110. 

See Trachtenberg, A Constructed Peace, pp. 382–383, especially n. 125 on p. 383.

111. 

Kennedy-Khrushchev conversation, June 4, 1961, FRUS 1961–1963, Vol. 7: Arms Control and Disarmament (Washington, D.C.: GPO, 1995), p. 88.

112. 

See Trachtenberg, A Constructed Peace, pp. 384–391.

113. 

Schwarz, Konrad Adenauer, pp. 692–693.

114. 

Granieri, The Ambivalent Alliance, pp. 179–180.

115. 

National Security Action Memo (NSAM) 270, October 29, 1963, FRUS, 1961–1963, Vol. 13, pp. 624–626. Rusk made the promise to maintain deployments in an October 27 speech in Frankfurt.

116. 

On U.S.-Soviet disagreements on the MLF, see Khrushchev to Kennedy, n.d., mid-1963, FRUS, 1961–1963, Vol. 6: Kennedy-Khrushchev Exchanges (Washington, D.C.: GPO, 1996), pp. 271–279. On British opposition, see Hal Brands, “Non-Proliferation and the Dynamics of the Middle Cold War: The Superpowers, the MLF, and the NPT,” Cold War History, Vol. 7, No. 3 (August 2007), pp. 395, 405–406.

117. 

NSC Excomm meeting, February 12, 1963, FRUS, 1961–1963, Vol. 13, p. 593. Kennedy feared that without the MLF, there would be a “very real possibility of a Franco/German cooperation in nuclear weapons systems.” See Bundy to de Zulueta, May 10, 1963, FRUS, 1961–1963, Vol. 13, p. 575.

118. 

MLF Conversation, August 30, 1963, FRUS, 1961–1963, Vol. 13, pp. 606–607; and Lyndon B. Johnson meeting, April 10, 1964, LBJ Presidential Papers, NSF, Subject File, Multilateral Force, box 23, LBJL.

119. 

On British resistance, see Richard Neustadt's memo, July 6, 1964, LBJ Presidential Papers, NSF, Subject File, Multilateral Force, box 23, LBJL. On Erhard's desire for a bilateral version of the force, see Erhard to Johnson, September 30, 1964, LBJ Presidential Papers, NSF, Subject File, Multilateral Force, box 23, LBJL. On Bundy's advice to scrap the plan, see Bundy to Rusk, Ball, and McNamara, November 25, 1964, LBJ Presidential Papers, NSF, Files of McGeorge Bundy, box 5, LBJL. For Johnson's decision, see NSAM 322, December 17, 1964, LBJ Presidential Papers, NSF, NSAMs, LBJL.

120. 

Publicly, Johnson was committed to the “hardware solution,” but in private he was maneuvering the Germans and British into negotiations that he hoped would kill the MLF. See Costigliola, “Lyndon B. Johnson, Germany, and ‘the End of the Cold War,’” pp. 189–192.

121. 

McGhee to Johnson, January 16, 1965, LBJ Presidential Papers, NSF, Country File, box 194, LBJL. McGhee later recalled that Johnson had “pulled the rug out from under” Erhard. Quoted in Brands, “Non-Proliferation and the Dynamics of the Middle Cold War,” p. 404.

122. 

Kissinger-Erhard conversation, January 28, 1966, Personal Papers of Francis M. Bator, Subject File, box 28, LBJL; and Bator to Johnson, “A Nuclear Role for Germany: What Do the Germans Want?” April 4, 1966, Personal Papers of Francis M. Bator, Subject File, box 28, LBJL.

123. 

Although the first draft of the treaty was only revealed in 1966, discussions about a prospective nondissemination pact began in the early 1960s.

124. 

Gavin, Nuclear Statecraft, p. 96.

125. 

Johnson-Erhard meeting, December 20, 1965, FRUS, 1964–1968, Vol. 13, p. 291.

126. 

NSC meeting, January 22, 1963, FRUS 1961–1963, Vol. 13, p. 486; Kennedy-Adenauer meeting, June 24, 1963, FRUS, 1961–1963, Vol. 13, p. 170; and Gavin, Gold, Dollars, and Power, pp. 113– 114, 138–141.

127. 

Erhard to Johnson, July 5, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL. In this letter, Erhard also demanded that Johnson put the MLF back on track. On Erhard's domestic difficulties, see Bator's memo to Johnson, August 23, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

128. 

Bator to Johnson, August 11, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL; and McGhee to State, September 20, 1966, FRUS, 1964–1968, Vol. 15: Germany and Berlin (Washington, D.C.: GPO, 1999), pp. 417–419.

129. 

Gavin, Nuclear Statecraft, pp. 47–48.

130. 

Germany had promised to offset British spending, and pressure was growing in London for force withdrawals if Bonn reneged. See Leddy to Rusk, August 23, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

131. 

Ball to Johnson, September 21, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

132. 

Erhard-Johnson meeting, September 26, 1966.

133. 

McCloy-Erhard conversation, October 20, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

134. 

Brands, “Non-Proliferation and the Dynamics of the Middle Cold War,” p. 408.

135. 

Gavin, Gold, Dollars, and Power, pp. 150–152.

136. 

See “The Mansfield Resolution,” LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 51, LBJL. On the collapse of the Erhard government, see Gavin, Gold, Dollars, and Power, p. 157; and Alfred C. Mierzejewski, Ludwig Erhard: A Biography (Chapel Hill: University of North Carolina Press, 2004), pp. 203–204.

137. 

McCloy to Johnson, November 21, 1966, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

138. 

Costigliola, “Lyndon B. Johnson, Germany, and ‘the End of the Cold War,’” p. 201.

139. 

Gavin, Nuclear Statecraft, p. 97.

140. 

McGhee to Rusk, August 25, 1966, FRUS, 1964–1968, Vol. 15, p. 395.

141. 

McCloy-Kiesinger meeting, March 5, 1967, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

142. 

Kiesinger to Johnson, April 4, 1967, LBJ Presidential Papers, NSF, NSC Histories, Trilateral Negotiations, box 50, LBJL.

143. 

Department of State Policy Planning Council, “Implications of a More Independent German Foreign Policy,” November 16, 1967, Bator Papers, Subject File, box 22, LBJL.

144. 

From January 1967 to May 1968, the United States and Germany conducted extensive negotiations over Article 3 of the treaty, which dealt with safeguards. See paper prepared in the Department of State, May 5, 1968, FRUS, 1964–1968, Vol. 15, pp. 662–669.

145. 

See Bonn to State, December 20, 1967, FRUS, 1964–1968, Vol. 15, p. 617. On the fear of the NPT's unlimited duration, see State Department paper, n.d., mid-June 1968, FRUS, 1964–1968, Vol. 13, pp. 712–716; and Fisher to Rusk, March 4, 1967, FRUS, 1964–1968, Vol. 11: Arms Control and Disarmament (Washington, D.C.: GPO, 1997), pp. 454–458. The right-wing Christian Social Union, in particular, argued that the treaty could forbid a future united Europe from inheriting the nuclear status of its nuclear-armed components. See Memorandum of Conversation (Memcon), July 23, 1968, FRUS, 1964–1968, Vol. 11, pp. 649–651.

146. 

The United States did not give this guarantee. See Duckwitz to Rostow, May 2, 1968, LBJ Presidential Papers, NSF, Files of Walt W. Rostow, box 2, LBJL; and Rostow to Johnson, June 11, 1968, LBJ Presidential Papers, NSF, Files of Walt W. Rostow, box 2, LBJL.

147. 

Rostow to Johnson, November 7, 1967, FRUS, 1964–1968, Vol. 11, pp. 523–524.

148. 

Rostow-Barzel conversation, February 23, 1968, FRUS, 1964–1968, Vol. 15, pp. 630–637.

149. 

Munich to State, July 15, 1967, LBJ Presidential Papers, NSF, Files of Spurgeon M. Keeny, box 6, LBJL. As Ambassador McGhee argued, “The Germans are reluctantly drawn along by the course of events toward the signing of the NPT.” See McGhee to Johnson, May 31, 1968, FRUS, 1964–1968, Vol. 15, pp. 675–677. U.S. officials were convinced in 1968 that despite right-wing opposition and ongoing safeguards disputes, Germany would “probably sign.” See Rusk-Kuznetsov conversation, May 17, 1968, FRUS, 1964–1968, Vol. 11, pp. 598–603.

150. 

West German leaders feared that the Soviet Union would continue to claim the right—under UN Articles 53 and 107—to intervene militarily in the Federal Republic. In March 1969, the Soviets hinted that they would be willing to renounce this right if West Germany signed the NPT. See Kiesinger-Nixon conversation, February 26, 1969, FRUS, 1969–1976, Vol. 40: Germany and Berlin, 1969–1972 (Washington, D.C.: GPO, 2008), pp. 34–38. On the Soviet assurances, see editorial note, FRUS, 1969–1976, Vol. 40, pp. 48–51. On Brandt's accelerated timetable for NPT signature, see Kissinger to Nixon, October 20, 1969, FRUS, 1969–1976, Vol. 40, pp. 103–105.

151. 

Matthias Küntzel, Bonn and the Bomb: German Politics and the Nuclear Option (London: Pluto, 1995).

152. 

On this logic, see Peter Gourevitch, “The Second Image Reversed: The International Sources of Domestic Politics,” International Organization, Vol. 32, No. 4 (Autumn 1978), pp. 881–912.

153. 

In response to public protests against nuclear weapons that year, Adenauer disingenuously promised to seek an agreement that would “avoid arming the armies in the East and the West with nuclear weapons,” while secretly attempting to do precisely the opposite. See Schwarz, Konrad Adenauer, p. 269.

154. 

See Memcon, February 8, 1967, FRUS, 1964–1968, Vol. 11, p. 436.

155. 

Mitchell Reiss, Without the Bomb: The Politics of Nuclear Nonproliferation (New York: Columbia University Press, 1988), pp. 98–99; and David Albright and Corey Gay, “Taiwan: Nuclear Nightmare Averted,” Bulletin of the Atomic Scientists, Vol. 54, No. 1 (January 1998), pp. 54–60.

156. 

See Stephen M. Walt, “Alliances in a Unipolar World,” World Politics, Vol. 61, No. 1 (January 2009), p. 95.

157. 

Barbara Demick, “More South Koreans Support Developing Nuclear Weapons,” Los Angeles Times, May 18, 2013, http://articles.latimes.com/2013/may/18/world/las-fg-south-korea-nuclear-20130519; and Robert Windrem, “Japan Has Nuclear ‘Bomb in the Basement,’ and China Isn't Happy,” NBC News, March 11, 2014, http://www.nbcnews.com/storyline/fukushimaanniversary/japan-has-nuclear-bomb-basement-china-isnt-happy-n48976.

158. 

Mark Urban, “Saudi Nuclear Weapons ‘On Order’ from Pakistan,” BBC, November 6, 2013, http://www.bbc.com/news/world-middle-east-24823846.

159. 

For example, in Pakistan the threat from India overturned the military's earlier hesitancy to acquire nuclear weapons, and the country eventually overcame supply-side and resource constraints. See Feroz H. Khan, Eating Grass: The Making of the Pakistani Bomb (Redwood City, Calif.: Stanford University Press, 2012). On the limitations of supply-side constraints, see Scott R. Kemp, “The Nonproliferation Emperor Has No Clothes: The Gas Centrifuge, Supply-Side Controls, and the Future of Nuclear Proliferation,” International Security, Vol. 38, No. 4 (Spring 2014), pp. 39–78.

160. 

On the idea that military retrenchment will accelerate proliferation, see Stephen G. Brooks, G. John Ikenberry, and William C. Wohlforth, “Don't Come Home, America: The Case against Retrenchment,” International Security, Vol. 37, No. 3 (Winter 2012/13), p. 35.