Faculty of Social and Behavioural Sciences MSc Political Science: International Organisation
Final Thesis
Supervisor: Dr. N.J.G. van Willigen Second reader: Prof. dr. D.C. Thomas
NATO
N
UCLEAR
S
HARING
,
B
ETWEEN
D
ETERRENCE
AND
D
ISARMAMENT
MIHAILO JOVETIC 1720686
11 June 2018
Word Count: 9801
Abstract: US tactical nuclear weapons (TNWs) have been deployed to European NATO allies since the early years of the Cold War. This arrangement, called nuclear sharing or forward deployment, persists to this day, as five European NATO states still host US TNWs. This paper seeks to explain why in the period 1991-2014 certain European nuclear hosts have ceased hosting US TNWs (Greece), others pursued proactive policies in arms-control, disarmament, and non-proliferation (ADN) forums (the Netherlands), while others have maintained a more passive stance (Italy). Through a neoclassical realist framework, it is hypothesised that public domestic pressures influence state behaviour in regard to nuclear hosting and activity in ADN forums, which ultimately the findings disprove. Intrinsic interests by a small state (the Netherlands) to maximise its influence in matters of nuclear strategy – a domain otherwise dominated by superpowers – lead it to attempt to diplomatically bridge nuclear and non-nuclear states. The strategic culture and aspirations of a medium-large state (Italy) to belong to the group of big players steering
T
ABLE OFC
ONTENTSIntroduction ... 3
Literature Review ... 7
Deterrence ... 7
Disarmament and Non-Proliferation ... 8
Why Neoclassical Realism ... 11
Theoretical Framework... 14
The Adapted Model ... 16
Methodology & Research Design... 21
Findings & Analysis ... 24
The Netherlands ... 24
Italy ... 27
Conclusion ... 31
Theoretical Implications and Suggestions for Further Research ... 32
References ... 34
I
NTRODUCTIONThe North Atlantic Treaty Organisation (NATO) clearly identifies collective defence as one of its three core tasks (NATO, 2010).1 Collective defence includes the notion of
deterrence, which is achieved through both conventional and nuclear capabilities, the latter being supplied by the United Kingdom (UK) and, most importantly, the United States (US).2 NATO explicitly states it will remain a nuclear alliance as long as nuclear
weapons exists. US nuclear warheads have been positioned in Europe since the 1950s in several countries as part of NATO’s extended nuclear deterrence (also known as nuclear forward deployment or nuclear sharing) to counter the threat of Soviet invasion.
Despite the dissolution of the Soviet Union (USSR) in 1991 and the rapprochement between NATO and Russia throughout the 1990s and 2000s (Rachwald, 2011, p. 119-124), US tactical nuclear weapons (TNWs) remained in Europe and are currently stationed in five countries: Belgium, Germany, Italy, the Netherlands and Turkey (Norris & Kristensen, 2011, p. 65). TNWs3 are smaller than strategic nuclear
weapons (SNWs) in terms of yield and range, and those remaining in Europe are all gravity bombs to be delivered by aircraft. TNWs are also called “theatre” weapons, since they are meant to be used in war theatre against an enemy army, not a strategic target such as for instance a city. Countries hosting US TNWs are called NATO nuclear host states: this means the TNWs are under US custody, but in case of nuclear war are to be delivered by the host state’s dual capable aircraft, making it a de facto nuclear
1 The other two being crisis management and cooperative security.
2 France is also a nuclear weapon state and member of NATO; however, its nuclear arsenal is not
assigned to NATO (Andreasen et al., p. 41).
3 These are sometimes also referred to as nonstrategic nuclear weapons, however in this research the term
weapon state.4 However, there is variation across nuclear host states’ policies on the
subject; most are believed to be hosting 10-20 US TNWs to be delivered by their own aircraft, while for instance Italy is believed to also be hosting between 25 and 50 US TNWs to be delivered by US Air Force aircraft based in Italy (Norris & Kristensen, 2011, p. 69; Kristensen, 2015). Furthermore, certain nuclear hosts have experienced vocal grass-roots movements advocating nuclear disarmament and have been an active proponent of arms control, disarmament, and non-proliferation (ADN) agreements on an international level – for instance through the Non-proliferation and Disarmament Initiative (NPDI) (e.g. the Netherlands), while others have not (e.g. Italy). There are instances of states ceasing to be nuclear hosts, namely the UK, Canada and Greece (Kristensen, 2005, p. 13). The UK acquired its own nuclear arsenal, while Canada stopped hosting in 1984 – the strategic importance for the latter being a nuclear host also being debatable. Greece is most relevant for this paper; it can be considered a strategically valuable nuclear host, is in Europe, but stopped hosting in 2001 (Sauer & Van der Zwaan, 2012, p. 81).
Extensive literature has sought to explain NATO nuclear sharing, in the vast majority of cases either from a classical deterrence theory viewpoint or from a disarmament and non-proliferation angle (Sagan, 1997; Hymans, 2006; Sagan, 2011; Foradori, 2012; Sechser, 2016). Little, however, of what has been written on this topic addresses (1) why in the 23 years from the dissolution of the Soviet Union in 1991 until the Ukraine crisis in 2014 there has not been a complete withdrawal of US TNWs from Europe, and moreover (2) why there has been variation between nuclear hosts’ policies on US TNWs on their territories and on ADN initiatives. This two-folded question is
4 This is also called the ‘dual-key arrangement’: the host state can veto delivery of nuclear weapons
empirically and theoretically relevant because NATO’s nuclear sharing policy in this period, as well as the variation between the different nuclear hosts’ policies, are an anomaly for neorealism (defensive and offensive), neoliberalism (understood as neoinstitutionalism), and social constructivism. Of these three grand theories, neorealism seems to hold most truth – why would NATO remain a nuclear alliance otherwise, and why would US TNWs still be in Europe? Yet, neorealism falls short in explaining why the various European nuclear hosts (which have experienced the same systemic change: the end of the Cold War) have adopted different policies; some disarmed (Greece), others took international initiatives in promoting nuclear disarmament (the Netherlands), while others remained passive (Italy). This calls for a more nuanced theory, maintaining neorealist assumptions, but allowing for intervening (domestic) variables explaining the variation in policies: neoclassical realism. The following research question will therefore be answered through a neoclassical realist theoretical framework:
What explains variation in nuclear sharing policies between NATO nuclear hosts in the period 1991-2014?
This paper starts by reviewing the literature relevant to nuclear sharing; the dichotomy laid out in the literature leads to identifying neoclassical realism as the most adequate theory to research the puzzle at hand. Next is an explanation of the theoretical framework – which takes inspiration from Schweller’s underbalancing theory (2004) and concepts from other neoclassical realist scholars (Ripsman et al., 2016), most importantly that of topic salience, leading to the causal mechanism and subsequent three hypotheses. In the methodology and research design section, the comparative nature of
the study, the case selection (the Netherlands and Italy)5, and the research methods
(document analysis and elite interviews) are explained and justified. This is followed by a case-by-case analysis of the findings, showing why the Netherlands is active in ADN forums (out of self-interest to maximise its influence as a small state by diplomatically bridging nuclear and non-nuclear states) and Italy is passive (due to its strategic culture of privileging alliance policy over unilateral action and satisfaction of the status-quo by foreign policy-making elites). This is followed by the overall conclusions of the paper, including the theoretical implications of the findings and suggestions for further research.
5 Greece is incorporated in the theoretical framework and not the findings section because it is a case of
successful nuclear disarmament, while this study is concerned with current nuclear hosts, and due to insufficient information to assess the impact of domestic factors.
L
ITERATURER
EVIEWAs was alluded to in the introduction, there are two somewhat opposed narratives authors have adhered to when addressing the subject of nuclear sharing: either from a classical deterrence theory perspective or from an ADN angle.
Deterrence
The “classic deterrence theory” camp (Schelling, 1966; Fearon, 1997; Schultz, 1998; Freedman, 2004; Schultz & Goodby, 2015), is grounded in exactly its name – the notion of deterrence: ‘discouraging the enemy from taking military action by posing for him a prospect of cost and risk outweighing his prospective gain’ (Snyder, 1961, p. 3).6 Due to
its adversarial nature and intrinsic element of balancing, it has its roots in the school of realism, which posits that international politics is an endless struggle for power among states who will therefore seek conflict to maximise their relative gains (Morgenthau, 1948; Waltz, 1979; Walt, 1990; Mearsheimer, 1994). Neorealism – also called structural realism; a modernised refurbishment of classical realism which points at the anarchic structure of the world as the driver of state behaviour – neatly explains why states pursue nuclear weapon programs: to balance against and deter a state they regard as a security threat (May, 1994; Thayer, 1995; Frankel, 1993). Within neorealism, there is in turn another dichotomy: defensive realism, the main exponent being Waltz; and offensive realism, the main exponent being Mearsheimer (Snyder, 2002, p. 150). If a state is unable to pursue a nuclear armament program, it will join an alliance with another state that is prepared to offer extended deterrence (Sagan, 1997, p. 57). In this
6 Deterrence exists in two forms: by punishment; i.e. by the threat of a major reprisal (usually with
conventional forces), and by denial; i.e. by minimising the incentives for a potential enemy to start a military incursion (with the threat of nuclear retaliation) (Snyder, 1961, p. 14-16). Deterrence by denial can therefore only be exercised by states possessing nuclear capabilities.
two-level dynamic, the deterrer (or patron) commits to protecting the protégé and can signal it in a variety of ways, for example by placing nuclear weapons on the latter’s soil (Schelling, 1966; Fearon, 1997; Schultz, 1998). The stronger the patron signals its commitment to the protégé, the more real the deterrence effect vis-à-vis a potential adversary. This is precisely the policy followed by the US during the Cold War, when it placed up to 7300 nuclear weapons (in 1971) in Europe (Norris & Kristensen, 2011, p. 65). When the threat of an enemy ceases to exist, deterrence theory in the eyes of defensive realists forecasts the dismantlement of nuclear arsenals; the 1991 South African decision is a case in point (Sagan, 1997, p. 60), as was the decision of states that were “born” nuclear – Ukraine, Kazakhstan and Belarus – to denuclearise due to the lack of an external threat (Garnett, 1995; Sagan, 1997). More broadly, defensive realism would predict that the systemic change (i.e. the dissolution of the Soviet Union) would lead to mutual nuclear disarmament, while offensive realism would predict that the systemic change would certainly not lead to nuclear disarmament – quite the opposite: it would predict the stronger side will move to maximise its power and further consolidate its hegemonic status (Mearsheimer, 2001, p. 21), by for example increasing or modernising its nuclear arsenal. Deterrence theory, according to both defensive and offensive realism, does therefore not neatly explain why the dissolution of the Soviet Union did not lead to the complete removal of US nuclear weapons from European soil, or an increase/modernisation of the existing arsenals in the period 1991-2014.
Disarmament and Non-Proliferation
The “disarmament and non-proliferation” camp (Hymans, 2006; Sagan, 2011), on the other hand, finds its theoretical roots partly in liberalism (particularly in neoinstitutionalism) and partly in social constructivism (Hymans, 2006, p. 458-460).
Liberalism is a rationalist theory – as is realism – and also sees states as the primary actors in world politics, but it is significantly different from realism because it emphasises cooperation between states and predicts this will lead states to create international institutions in the search for maximised absolute gains (Keohane & Martin, 1995; Grieco, 1993). Social constructivism goes a step further and posits that international institutions, norms and values are diffused between elites and therefore become the main driving factors in world politics, rather than states as rational actors (Karns et al., 2015).
Since nuclear weapons were invented (and used) in the 1940s, several multilateral and bilateral treaties have been signed restricting the testing, developing or producing of nuclear weapons, as well as reducing existing arsenals. Most of these treaties are time bound, that is to say: they have an expiry date after (or before) which a treaty renewal – or a new treaty altogether – is negotiated (ideally). Pivotal in this series of treaties and arms control agreements is the 1968 multilateral Treaty on the Non-Proliferation of Nuclear Weapons (NPT), which prohibits the transfer of nuclear weapons from nuclear to non-nuclear states, furthers the goal of complete nuclear disarmament (Alberque, 2017, p. 11), while committing nuclear states to aid non-nuclear states in developing peaceful non-nuclear technologies (Karns et al., 2015, p. 356). The NPT is a prime example of multilateralism in arms control and non-proliferation, although in itself it did not prevent existing nuclear states from increasing their arsenals or lead them to nuclear disarmament. It did, however, pave the way for bilateral and multilateral agreements that did just that.7 A first such example are the Strategic Arms
7 As this paper is concerned with the non-disarmament of nuclear weapons on European soil, the
examples given here include mainly instances that led to actual nuclear disarmament, not treaties on nuclear-free zones or instruments aimed at preventing the spread of nuclear weapons to non-nuclear states.
Limitation Talks (SALT), resulting in SALT I (1972) and SALT II (1979 – which was however never ratified), which limited the number of intercontinental ballistic missiles (ICBMs) of both US and Soviet Union. SALT was followed by the Strategic Arms Reduction Treaty (START), negotiated throughout the 1980s and signed in 1991, which committed the US and USSR to reducing their nuclear arsenals to 7200 warheads each. The two superpowers also signed the Intermediate-Range Nuclear Forces Treaty (INF) in 1987, banning all nuclear and conventional missiles with ranges of 500-1000 kilometres. START II was followed by the Strategic Offensive Reduction Treaty (SORT), signed in 2002 by the US and Russia, which reduces nuclear warheads even further, to a maximum of 2200 operationally deployed warheads each. Most recently, in 2011 the US and Russia signed and ratified the New START, replacing SORT, pledging to half the number of strategic nuclear missile launchers and revising the inspection and verification mechanism, which was introduced in SORT. New START expires in 2021.
These agreements were successful in reducing both US and Soviet/Russian nuclear arsenals, and numbers show the end of the Cold War led to a drastic reduction of US nuclear weapons on European territory (Norris & Kristensen, 2011, p. 65). The overwhelming majority did not address TNWs, the only exception being the Presidential Nuclear Initiatives (PNIs) in late 1980s and early 1990s. The PNIs were a series of reciprocal unilateral actions by US and USSR leaders of removing forward deployed TNWs mounted on short-range missiles and artillery shells.8 Yet, they did not
concern gravity TNWs, which are those that are today placed in Europe. Neoinstitutionalism would predict that the previously discussed agreements would lead to further cooperation, resulting in including all nuclear weapons (thus also all TNWs)
and all sides would denuclearise, because since they are already cooperating through various international regimes which include inspection and control mechanism, why would any side have the need to hold on to their nuclear weapons? Empirical evidence shows this not to be the case, despite concrete attempts: the 2017 Treaty on the Prohibition of Nuclear Weapons (TPNW), championed by a group of NGOs under the banner of the International Campaign to Abolish Nuclear Weapons (ICAN) and which prohibits the possession of nuclear arms, was passed in the United Nations General Assembly and signed by 122 states. All nuclear states abstained, including NATO nuclear hosts, with the Netherlands even voting against. Likewise, social constructivism would predict that the spread of international norms of nuclear disarmament and non-proliferation will inevitably lead to full nuclear disarmament, which the empirical evidence given above has thus far disproven.
Why Neoclassical Realism
Neorealism (defensive and offensive), neoinstitutionalism, and social constructivism hence seem to fall short in explaining continued nuclear sharing, which is an anomaly for all three theoretical views. Of the three grand-theories, neorealism – and its core assumption of deterrence – seems to hold the most truth: otherwise why would NATO so explicitly state it will remain a nuclear alliance as long as nuclear weapons exist? And furthermore, why would any US nuclear weapons still be in Europe? This shows that increased cooperation does not necessarily lead to more cooperation (as per neoinstitutionalism), and that norms and ideas of disarmament and non-proliferation do evidently not spread ubiquitously (as per social constructivism). At this level of abstraction, however, neorealism cannot explain why some nuclear hosts have disarmed (e.g. Greece) and others have not (e.g. the Netherlands and Italy), or why certain nuclear
hosts have taken international initiatives in promoting nuclear disarmament (e.g. the Netherlands) while others have not (e.g. Italy). All these countries have, to a certain degree, experienced the same systemic changes, but their foreign policy responses have differed.
A variation of neorealism – neoclassical realism – advances causal mechanisms that can successfully lead to an understanding of variation between policies of different NATO nuclear hosts. Neoclassical realism (whose original proponents are Brown, 1995; Christensen 1996; Schweller, 1998; Wohlforth, 1993; and Zakaria, 1998; and more recently Taliaferro, 2006; Lobell, 2009; Kitchen, 2010; and Ripsman et al., 2016)9
posits that ‘systemic pressures are filtered through intervening domestic variables to produce foreign policy behavior’ and that ‘states assess and adapt to changes in their external environment partly as a result of their peculiar domestic structures and political situation’, whereby their specific political processes act as transmission belts for policy responses to systemic changes (Schweller, 2004, p. 164). Neoclassical realism therefore fits with deterrence theory, as it was proposed earlier, but crucially adds explanatory power by including unit-level (domestic) variables which contribute to answering the central focus of this thesis: variation between the policies of NATO nuclear hosts. Within the school of neoclassical realism, several mid-range theories have been developed; the theoretical framework that follows in the next section takes inspiration from the causal mechanisms proposed in Schweller’s theory of underbalancing (2004) and from concepts other neoclassical realist scholars (Ripsman et al., 2016) have developed when explaining foreign policy behaviour. Schweller’s theory of underbalancing is suitable and relevant for answering the research question because it
brings ulterior explanatory power by elaborating on causal mechanisms brought forward by earlier realist scholars (namely Waltz’s balance of power), and by introducing new concepts that help in understanding the question at hand. Pure neorealist or neoliberal theories have given up such explanatory power for the sake of parsimony; for instance, defensive realism predicts that the disappearing of a clear enemy would logically lead to a complete removal of nuclear arsenals in order to maintain the balance of power; while offensive realism predicts that it would lead to increased and/or modernised nuclear arsenals. Likewise, neoliberalism would predict states to increasingly cooperate and hence fully dismantle their nuclear arsenals as they seek to maximise absolute gains. The theoretical framework given in the following chapter can therefore help to systematically shed light on what unit-level intervening variables explain variation across different NATO nuclear hosts in reacting to systemic changes; an empirical phenomenon that for the other theories is an anomaly.
T
HEORETICALF
RAMEWORKSchweller’s theory of underbalancing (2004) is a neoclassical realist variation and extension of Waltz’s balance of power theory. The original balance of power theory posits that states balance against external threats, which they do systematically and timely: ‘balancing means the creation or aggregation of military power through internal mobilization or the forging of alliances to prevent or deter the territorial occupation or political and military domination of the state by a foreign power or coalition’ (Schweller, 2004, p. 166). The opposite of balancing is underbalancing, which occurs when ‘the state does not balance or does so inefficiently in response to a dangerous and unappeasable aggressor’ (p. 168). Schweller puts forward four unit-level variables which contribute to understanding why a democratic state would underbalance:
(1) elite consensus – ‘states do not make policies, governments through their leaders do’ (p. 170); this concept concerns the degree of shared perception among policy-makers about (a) the degree of external threats, (b) what policies are required (or not) to deal with such threats, and (c) the domestic political risks and costs associated with such policy options.
(2) government or regime vulnerability – ‘what is the likelihood that the current leadership will be removed from political office’ (p. 173), this concept concerns the relationship between rulers and ruled: is the government legitimate in the eyes of the public? Are people’s expectations met? Does the government enjoy broad support from the masses? Most importantly, can the government minimise domestic interference when making policy decisions?
(3) social cohesion – ‘do all members of society accept the rules of the game; that is, they support the society's institutions as legitimate and appropriate
mechanisms to settle disputes among them no matter how profound their disagreements or grievances (i.e. loyal opposition)’ (p. 175), it concerns a longer-term trend in terms of the general public being “on the same page”, and not occasional political disagreements on epiphenomenal issues.
(4) elite cohesion – ‘concerns the degree to which a central government’s political leadership is fragmented by persistent internal divisions’ (p. 180).
Government/regime vulnerability and social cohesion determine the ability of the state to extract resources to balance, while elite consensus and elite cohesion determine whether this ability will be set in motion. These four variables combined determine the degree of state coherence (p. 169). Schweller therefore proposes the following chain of events, including the aforementioned intervening variables, which better explains adjustments in state foreign policy than Waltz’s balance of power theory does:
change in relative power
elite consensus about the nature of the threat and the degree of elite cohesion
mobilization hurdles as a function of regime vulnerability and social cohesion
continuity or change in foreign policy (i.e. balancing, bandwagoning, appeasement, half measures, etc.)
Figure 1 – Schweller's causal mechanism explaining change in foreign policy
Schweller’s central thesis is that when “the change in relative power” is the rise of an external threat, the response of states with low coherence (as a function of the four variables) will be to underbalance.
Now, there is no rise of an external threat NATO nuclear hosts may or may not (under)balance against in the period that is analysed in this study, which is why the model does not perfectly align with the case of this research. However, Schweller’s concepts and causal mechanism do offer inspiration for the development of a more adequate model that can help in understanding variation in the policies of nuclear hosts on the question of ADN efforts. A few adjustments in the causal mechanism and the introduction of two additional neoclassical realist concepts are in order.
The Adapted Model
First and foremost, topic salience in the public debate is added as a conditional variable in the adapted model. This concept (often referred to as “issue salience” in Foreign Policy Analysis (FPA) literature10) is defined as ‘the relative importance and
significance that an actor ascribes to a given issue on the political agenda’ (Soroka, 2003, p. 28-29; Wlezien, 2005, p. 556-561; in Oppermann, 2014, p. 26). In this case, the “actor” in question is the public, since it concerns salience in the public debate. This concept is added because ‘foreign policy decision-making can be expected to follow an entirely different path in high salience environments than in low-salience environments’ (Oppermann, 2010, p. 7), and because preliminary findings suggest this factor might be an explanation for the varying policies of nuclear hosts.11
The concept of social cohesion remains as explained by Schweller and summarised earlier, while elite consensus, elite cohesion, and government vulnerability are incorporated in the concept of foreign policy executive (FPE). This is done to simplify the model, and because the FPE is a category of decision makers who ‘sit at the helm of the state, it includes president, prime minister, dictator, key cabinet members,
10 Soroka, 2003; Carlsnaers, 2008; Hudson, 2013; Oppermann & Spencer, 2013.
advisors charged with the conduct of foreign and defence policies’ (Ripsman et al., p. 61).12 The strength or weakness of the FPE therefore equates to the combination of what
Schweller calls elite consensus and cohesion, and it also determines to what extent government is vulnerable.13 In short, a strong FPE enjoys support amongst elites and
can be expected to act somewhat freely; conversely, a weak FPE has to make careful considerations prior to every decision and is therefore more prone to inaction.
What follows are three schematic overviews showing what combinations of explanatory variable values lead to the three types of empirically witnessed behaviours by nuclear hosts, namely: former (Greece, ceased hosting TNWs), reluctant (the Netherlands, still hosts TNWs but is an international advocate of nuclear disarmament) and passive (Italy, still hosts TNWs and is passive on the subject).
decline of an external threat
high topic salience in the public debate
high social cohesion + weak FPE (strong/weak)
nuclear disarmament (Greece) Figure 2a – Explanatory variables value combination resulting in nuclear disarmament
12 The FPE can be considered the equivalent of what Putnam calls ‘the negotiator’ in his two-level games
theory (1988), whereby the main goal of the negotiator is to maximise their own win-sets by balancing domestic and international interests.
13 A strong FPE does not necessarily mean they are a dictator or authoritarian leader, it only means they
decline of an external threat
high topic salience in the public debate
high social cohesion + strong FPE
nuclear hosting but advocating nuclear disarmament internationally (the Netherlands) Figure 2b – Explanatory variables value combination resulting in reluctant nuclear hosting
decline of an external threat
low topic salience in the public debate
high social cohesion + weak FPE
only maintaining nuclear host status (Italy) Figure 2c – Explanatory variables value combination resulting in passive nuclear hosting
Since the Greek case is not as thoroughly researched as the second two, it is addressed here, with a particular focus on its theoretical implications.
All US TNWs were secretly removed from Greek bases in 2001 (Sauer & Van der Zwaan, 2012, p. 81), with various media outlets reporting different stories (Kristensen, 2005, p. 56). Concrete reasons for the withdrawal remain unclear. It is reported that in 1998 NATO asked Greece to replace its ageing aircraft fleet with F-16s,
which would include nuclear strike capability, but the ‘Greek government declined because its scarce resources were more urgently needed for air defence and conventional missions’ (p. 56). This seems plausible, as other news sources reported that the TNWs were removed because new aircraft Greece had ordered could not carry nuclear weapons (Harrell, 2008).
The limited information on the Greek case does not support the hypothesised cruciality of topic salience, however – and most importantly – it highlights the importance of the FPE and stresses the need to introduce another neoclassical realist concept: strategic culture. This concept is closely linked to the FPE, as it includes ‘a set of inter-related beliefs, norms, and assumptions’ (Ripsman et al., 2016, p. 67) which can fundamentally shape how the FPE undertakes strategic adjustments (p. 69).14 Strategic
culture is also reflective of the size, geographic location and history of a particular state, for instance: the Netherlands is surrounded by NATO allies and has a reputation of a diplomatic bridgebuilder (interview: Van der Meer, 2018); while Italy is adjacent to Northern Africa, making it much more prone to migratory streams and maritime threats. Italy even briefly pursued its own nuclear program. In the Greek case, the FPE seems to have strategically prioritised conventional deterrence, which was assented by NATO, suggesting Greece was also not vital in NATO’s overall nuclear deterrent.
Considering the three empirical policy outcomes identified in the causal mechanisms, the following hypotheses are derived:
H1: When nuclear sharing is salient in the public debate, social cohesion high, and the FPE weak, the state will cease being a nuclear host.
14 It is not added as a separate variable in order to maintain model simplicity and most importantly
H2: When nuclear sharing is salient in the public debate, social cohesion high, and the FPE strong, the state will be a reluctant nuclear host.
H3: When nuclear sharing is not salient in the public debate, social cohesion high, and the FPE weak, the state will be a passive nuclear host.
These hypotheses will be tested by using the adapted model previously explained. The next section gives an explanation of how the variables in the hypotheses are operationalised (measured through which indicators), what the exact research method is, what criteria are used to select the cases to be studied, and what data sources are used.
M
ETHODOLOGY&
R
ESEARCHD
ESIGNBefore delving in the indicators which are used to operationalise the concepts of the hypotheses, it is first necessary to state and justify which cases will be studied, in order to more concretely address the question of indicators.
This is a comparative case study using the method of difference. As such, two cases with similar values on the independent variable but different values on the dependent variable will be studied in order to answer the research question. There are only five NATO nuclear hosts, ideally all would be studied and compared in order to make a broader claim on nuclear sharing policies from the viewpoint of hosts. Due to time and resource constraints, however, only two host states will be studied: Italy and the Netherlands. These two nuclear hosts were selected because one experienced substantial anti-nuclear grass-roots movements throughout the years (the Netherlands), while one did not (Italy). One is a militarily large NATO country, while the other is comparatively small NATO. Both host US TNWs to be delivered by their own aircraft, but while the Netherlands has been an active proponent of nuclear disarmament internationally, Italy has been more passive. Furthermore, comparing the Netherlands and Italy is theoretically relevant because it enables the testing of a theory (neoclassical realism) which claims domestic level variables determine foreign policy change in response to systemic change. In the case of these two states, the systemic change is the same15, but the foreign policy response is not.
The indicators used to measure the variables of the hypotheses include both primary and secondary sources and will be researched using process tracing16.
15 The exposure to different strategic and security threats (see p. 19) will be kept in consideration.
16 Defined as the tracing of ‘the decision process by which various initial conditions are translated into
Topic salience in the public debate is estimated by looking at the number of instances “nuclear weapons” and “NATO” are discussed,17 which is achieved through
document analysis of parliamentary debates, media outlets, and the amount of academic publications on the subject. This data is triangulated by conducting semi-structured interviews with subject matter experts in the respective countries in order to minimise reverse-causation or a reciprocal relationship between the variables.
Social cohesion is determined by analysing the content of the sources found when determining topic salience and comparing the degree of agreement on what course of action the government should follow. Less agreement points at less social cohesion, more agreement points at more social cohesion, most likely resulting in topic salience quickly dropping – unless the government decides to go against public opinion. This is triangulated with interviews.
FPE strength is gauged by observing the constitution of the government (single-party, coalition, grand coalition, etc.) – a single-party government points at a stronger FPE, while a grand coalition or minority government points at a weaker FPE – and the recent historical record of governments (do they often fall or do they regularly complete their mandates)18. Semi-structured interviews are conducted with experts who have been
involved in the shaping and/or analysing of policy in the respective countries.
Finally, this research aims at answering the research question given on page five through a neoclassical realist theoretical framework. Of course, it contains limitations, the main being that due to time and resource constraints only two current nuclear hosts
17 These two keywords were selected because their combination is most approximate in referring to
questions of nuclear sharing, rather than other nuclear weapons matters such as e.g. Pakistani, North Korean or Iranian nuclear programmes.
18 Since the cases to be studied in this study concern widely considered democratic states, these indicators
are researched, while studying all five would be desirable to maximise external validity. This limitation is mitigated by studying two nuclear hosts that represent extremes in terms of the parameters to be studied; a pro-ADN and small NATO nuclear host on the one hand, and a passive and large NATO nuclear host on the other. It must also be noted that this study explicitly focuses on the variables explained previously, but that it could occur that other factors not considered by the theoretical model may be influencing state behaviour. This possible discrepancy will be addressed in the final considerations of the paper.
F
INDINGS&
A
NALYSISThe two cases discussed in this section are supported by document analysis and through a total of five interviews with experts from the two countries.
The Netherlands
US TNWs were first deployed to the Netherlands in the 1950s and were often discussed in Parliament during the Cold War, with a Minister of Defence in the 1960s justifying nuclear hosting by stating that ‘you are either an ally, or you are not’ (interview: Voorhoeve, 2018). In the 1980s, nuclear hosting reached peak salience levels in the public debate as extensive protests took place across the country opposing NATO’s Dual Track Decision19 to place nuclear cruise missiles (the so-called Euromissiles) in
the Netherlands and other European states. The missiles were in the end never placed in the Netherlands due to the 1987 INF treaty, and the number of US TNWs drastically decreased from 6.000 warheads across Europe in 1987 to 500 in 1995 (Norris & Kristensen, 2011, p. 65).
The 1990s and 2000s, however, saw an equally drastic decrease in public interest in nuclear sharing policy compared to the 1980s (Van der Meer, 2018). The question of nuclear deterrence vanished together with the USSR as NATO transformed from an explicitly defensive alliance to an adaptable crisis management force (interviews: Dutch Policy Officer, 2018; Voorhoeve, 2018). This is further shown by the fact that PAX – a Dutch NGO – needed more than a year to collect enough signatures (40.000) for a citizens’ initiative requesting the Parliament to debate the possibility of banning nuclear
weapons (Van der Meer, 2018).20 On top of that, it is the only citizens’ initiative on the
matter, and parliamentary debates on nuclear weapons have also been scarce.21 Indeed,
in the 1990s NATO summits of Defence Ministers were mostly concerning the situation in Yugoslavia, and Russia was regarded as a constructive partner, with Russian delegations participating in the meetings (Voorhoeve, 2018). However, according to a Dutch Policy Officer at the Ministry of Foreign Affairs (MFA), PAX’s efforts in raising public awareness have contributed to the Netherlands developing a more elaborate policy regarding nuclear sharing in recent years, as representatives of PAX often meet with officers of the Dutch MFA. Views on the degree of topic salience and the weight of public opinion in policy-making are hence differing. What all interviewees agree on and remains undeniable, is that salience in the public debate has never come close to what it was in the 1980s.22
Most news and academic publications on the matter advocate the removal of the TNWs, on the grounds of them being outdated, expensive, the NPT, and for the sake of security (arguing the warheads could spontaneously explode). However, there were (and still are) deep divisions in the public and in political parties on what policy the government ought to adopt regarding nuclear sharing (Voorhoeve, 2018; Van der Meer, 2018). For instance, Voorhoeve and Van der Meer agree that the military importance of US TNWs in Europe has become negligible, confirming views elaborated by others (Nichols et al., 2012; Van der Zeijden, 2014), but disagree on the degree and nuance of
20 The citizens’ initiative ‘Teken tegen kernwapens’ (‘Sign against nuclear weapons’) was started in
September 2014, and the 45.000 signatures were delivered to the Parliament in September 2015 (See PAX, 2016).
21 In this period there have been 65 parliamentary questions (in the lower house) involving the keywords
“nuclear weapons” and “NATO”, which is an average of just below three per year. For an overview of the standpoints of political parties against nuclear sharing in the 2012 election, see Van der Zeijden (2014, p. 47).
22 In that case, it was evident the Dutch government, led by Lubbers, was strongly influenced by public
their political-strategic value. On the one hand, Van der Meer is sceptical of the argument that being a nuclear host elevates a country’s position within the alliance, and believes the main reason US TNWs have been kept in the Netherlands is because policymakers (wrongly) think the TNWs can be part of a deal with Russia to mutually disarm.23 On the other hand, Voorhoeve believes nuclear sharing is pure political
symbolism: the US showing its commitment to European defence and security, and European states showing their willingness for burden sharing to the US. This view is shared by Cotta-Ramusino (interview, 2018) and is in line with the following statement made by the Netherlands in the 2005 NPT Review Conference: ‘the purpose of the remaining non-strategic nuclear weapons held by NATO is fundamentally political’ (UN Office of Disarmament Affairs, 2005, p. 364).
As far as political parties and government stability are concerned, there have been ten cabinets and four different Prime Ministers from 1991 to 2014. All cabinets consisted of two or more party coalitions, and three have fallen. Political parties on the left of the political spectrum have traditionally been against (some more explicitly than others), while those on the right have generally been in favour, although internal divisions in for example the Christian Democrats (CDA) were and are present (Voorhoeve, 2018). A thorough debate on the topic would have been challenging for certain parties (e.g. the social liberals – D66, and the CDA) (Voorhoeve, 2018), which may be part of the explanation for why the question has been barely touched upon.24
The fact that all governments were coalitions with a right-leaning party as largest or second-largest coalition partner may also be part of the reason for the lack of discussion and action on nuclear sharing. However, this aspect is not covered by the theoretical
23 Russian interlocutors Van der Meer has spoken with have manifested their concerns lie with US SNWs
(such as the Minuteman missiles), not at all with the comparatively small TNWs.
model used here, therefore this conclusion cannot be drawn with certainty. Moreover, whether a purely left-leaning cabinet would have been more proactive on the question is also difficult to assess; in fact, Van der Meer stresses that no matter the political colour of the government, the Netherlands has a deep-rooted tradition and interest in valuing and protecting multilateralism, which reflects in it having a reputation of bridgebuilder and consensus-finder. This contributes to understanding why, on the one hand, the Netherlands is proactive in ADN arenas, while on the other still hosting US TNWs, therefore factually linking nuclear and non-nuclear weapon states. This claim is seconded by Voorhoeve; who agrees with Van der Meer and Cotta-Ramusino that the 1990s and 2000s were a missed chance to fully withdraw US TNWs. The amount of US TNWs did indeed drastically decrease, but their intrinsic politically symbolic value resulted in a reduced number remaining in Europe.
Italy
In many aspects, Italy finds itself in a very similar situation to the Netherlands.25 The
main differences are that Italy hosts more TNWs, to be delivered by both US and Italian aircraft, and is far less active in ADN forums. The first two distinctions are factual (Norris & Kristensen, 2011, p. 65), the third is more prone to interpretation. Voorhoeve, Van der Meer, and Cotta-Ramusino all strongly agree Italy is much less proactive in ADN efforts than the Netherlands, which is also evidenced by it not being a member of the NPDI, not participating in TPNW negotiations, and voting against it (Ronzitti, 2017, p. 3). Despite this, the Dutch Policy Officer pointed out that their Italian counterparts are at times very proactive in ADN efforts, though no specific details were given.
25 Many NATO and bilateral US-host state arrangements that were explained for the Netherlands apply
also to Italy. To avoid repetition, this section mainly focuses on highlighting the differences between the two and analysing them in order to explain their causes.
Salience in the public debate is extremely low, if not zero (interview: Simoncelli, 2018; Cotta-Ramusino, 2018). According to Cotta-Ramusino and Simoncelli, this is due to the general Italian population being unaware that US TNWs are present on Italian soil. Mogherini (as cited in Foradori, 2011, p. 30 – who at the time was a Parliamentarian), has said that this lack of public interest is a remnant of the Cold War, during which questions of military and defence were taboo and any kind of public scrutiny or transparency of decision-making in their regard was undesirable. Furthermore, during the Cold War, US TNWs played an important role in reaffirming Italy’s belonging to the West (Simoncelli, 2018) – which is particularly important because Italy’s Communist Party was the largest in Europe (Foradori, 2014, p. 33). However, the end of the East-West rivalry depoliticised nuclear hosting, making it worthless in political debates, resulting in a general disinterest by the public, mass-media, and political elites (Pistelli26 and Silvestri27, as cited in Foradori, 2011, p. 30). This is reflected in the low
number of parliamentary questions submitted in 1991-2014 including the keywords “nuclear weapons” and “NATO”, which stands at 33; an average of just below one and a half per year.28
News items on TNWs are extremely scarce (Simoncelli, 2018), and the occasional ones published address the matter as if it were a new phenomenon (Cotta-Ramusino, 2018) and often include misinformation29 (Simoncelli, 2018), although in
the large majority of cases they call for their removal. Academic articles are also scarce,
26 Lapo Pistelli, at the time, was Parliamentarian and head of the Foreign Affairs Council. In 2013-2015
he was Deputy Minister of Foreign Affairs under Federica Mogherini.
27 Stefano Silvestri, at the time, was President of the Istituto Affari Internazionali (IAI - International
Affairs Institute) in Rome.
28 This is half the amount submitted in the Dutch Parliament in the same period. Also, the lower house of
the Italian Parliament consists of 630 members, while the Dutch lower house of 150.
and themselves point out how nuclear sharing has been forgotten (Cotta-Ramusino, 1996; Foradori, 2011), though an overwhelming majority of them advocate the withdrawal of TNWs, primarily on the grounds that nuclear sharing sets a precedent of how the NPT can be outflanked (Cotta-Ramusino, 2018; Ronzitti, 2007). There have also been citizens’ initiatives against nuclear weapons (Simoncelli, 2018), but from 1991 onwards none of them gained substantial support. This further shows that most of the public is unaware of TNWs being in Italy, in spite of the fact that when told about their presence, most people say they should be removed (Simoncelli, 2018). However, this high degree of social cohesion is virtually irrelevant due to poor interest by the larger public.
Likewise, the degree of government stability is irrelevant, despite the fact that in the studied period there have been 15 different governments (two of which care-takers), nine different Prime Ministers, and 14 of which have fallen. As Cotta-Ramusino points out, every new government needs to set priorities, and for the previously explained reasons those never included nuclear sharing. They have rather concerned questions of economic reform, sovereign debt, unemployment, and migration. Moreover, foreign policy decisions are shaped by elites in pertinent ministries, with the political class often not fundamentally involved in the decision-making (Simoncelli, 2018). This is illustrated by the fact that throughout the years following the end of the Cold War, Italy has consolidated its foreign policy as pro-NATO both under left-leaning governments (e.g. by participating in the 1999 intervention in Kosovo), as well as under right-leaning governments (e.g. by participating in the mission to Afghanistan). Therefore, regardless of who was in power, Rome’s foreign policy always followed Washington, D.C. (Nuti,
2005). This further shows that the TNWs function of signalling Italy’s pro-NATO stance is no longer valid.
Italy’s reluctance for proactive action in ADN forums is therefore caused by the essential importance it ascribes to NATO solidarity in combination with its deeply rooted strategic culture which privileges multilateralism over any form of unilateralism (Foradori, 2011, p. 34-35).30 The Italian approach to foreign and defence policies has
always been of moderation rather than change (Foradori, 2014, p. 39). It is hence not too surprising that Italy has somewhat distanced itself from the more “progressive camp” of NATO nuclear hosts (p. 32), as newer NATO members – particularly those who feel more vulnerable to Russian influence – have expressed a will for US TNWs remaining in Europe (Cotta-Ramusino, 2018). Nobody in Italy denies the military negligibility of US TNWs in Europe (Simoncelli, 2018; Cotta-Ramusino, 2018), but the combination of their political (Atlantic) value, views of newer NATO members, and Italy’s strategic culture, have resulted in Italy maintaining a path of conservative policy on nuclear sharing.
C
ONCLUSIONThe research has shown that the theoretical model and hypotheses proposed are only partly accurate. First of all, categorising topic salience in the public debate as a conditional variable does not hold true, even though salience was indeed higher in the Netherlands than in Italy. Still, in neither countries topic salience was recorded as high, yet they adopted different policies regarding nuclear sharing. Social cohesion loses its prominence consequentially. The core factor determining the nuclear sharing policy course of each country lies in FPE strength and their respective strategic cultures.
US TNWs have not been fully withdrawn from Europe due to them still carrying political value. Between their former role of deterrence and the pressures of disarmament there is reciprocal Atlantic solidarity; US commitment towards Europe’s security, and European states’ commitment to burden sharing in NATO. As Freedman wrote in 1994: ‘US nuclear guarantee to Europe is crucial to the authenticity of the Atlantic Community’ (p. 44). This has held true, as Thränert confirmed in 2010: ‘a full-scale debate about US nuclear withdrawal from Europe could trigger a controversy that would undermine NATO cohesion’ (in Sauer & Van der Zwaan, 2012, p. 92).
Matters of nuclear strategy are generally superpower themes, where the US, Russia, or China (and to a lesser extent the UK and France) steer developments. In this context, the Netherlands is a small state, and in order to maximise its power has an intrinsic interest in adhering to and implementing international norms and values, thereby maintaining and furthering multilateralism. This explains the Dutch policy of proactive engagement in ADN forums, which aims at attaining diplomatic bridging between nuclear and non-nuclear states, while preserving its close ties with NATO
allies (particularly the US) by de facto following alliance policy (i.e. no unilateral nuclear disarmament).
In the case of Italy, its size makes it aspire to power levels of France or the UK, but it does not have its own nuclear deterrent. Nuclear hosting is therefore a valid alternative, especially politically. Its policy of passivity in ADN forums is therefore explained by the combination of (1) Italy’s strategic culture (privileging multilateralism over unilateralism, i.e. following NATO policy), (2) the limited space this gives for controversial policies without upsetting the alliance, and (3) the fact that Italy’s power in the status quo is satisfactory to its foreign policy-making elites.
Theoretical Implications and Suggestions for Further Research
The theoretical model proved partly accurate for this research question, however it may also be applicable to other questions and/or cases. The main adaptation that can be made based on these findings is that when the policy in question is of continuation, public salience is not a conditional variable, but a regular variable. As the Euromissiles crisis in the Netherlands showed, public salience can sway government decision-making another way than systemic pressures may prescribe. The theoretical implication is that the policy in question needs to be of change, rather than of continuation, and needs to be visible (physically or in the media) to the public, in order for public salience to be a conditional variable. On the other hand, to augment support for this argument in the case of nuclear sharing, the remaining three host states need to be studied.
Further understanding of how NATO states react to alliance policies can be attained by using the theoretical model used in this paper. Suggestions include: why and how do NATO members modernise their defence apparatuses – see Greece refusing to buy F-16s in the 1990s, or certain NATO allies purchasing F-35s while others do not,
and what vested interests may be at play in those decisions? How has NATO enlargement affected nuclear policy dynamics in the alliance? And finally, what implications does increasing defence cooperation in Europe have on NATO? These are only a few examples of possible topics worth exploring to better understand NATO and further develop the neoclassical realist theoretical model used here.
R
EFERENCESAlberque, W. (2017, February 7). The NPT and the Origins of NATO’s Nuclear Sharing Arrangements. IFRI, Proliferation Papers, 57.
Andreasen, S., Williams, I., Rose, B., Kristensen, H. M., & Lunn, S. (2018, January). Building a Safe, Secure, and Credible NATO Nuclear Posture. The Nuclear Threat Initiative.
Brown, M. E., Lynn-Jones, S. M., & Miller, S. E. (Eds.). (1995). The perils of anarchy: contemporary realism and international security. MIT Press.
Carlsnaes, W. (2008). Actors, structures, and foreign policy analysis. Foreign policy: theories, actors, cases, 85-100.
Christensen, T. J. (1996). Useful adversaries: Grand strategy, domestic mobilization, and Sino-American conflict, 1947-1958. Princeton University Press.
Cotta-Ramusino, P. (1996). La presenza di armi nucleari in Italia: Un problema dimenticato.
Fearon, J. D. (1997). Signaling foreign policy interests: Tying hands versus sinking costs. Journal of Conflict Resolution, 41(1), 68-90.
Foradori, P. (2011). Il dibattito sulle armi nucleari tattiche in Italia: tra impegni di disarmo e solidarietà atlantiche. Documenti IAI, 11(4).
Foradori, P. (2012). Tactical nuclear weapons in Italy: striking a balance between disarmament aspirations and alliance obligations. The Nonproliferation Review, 19(1), 13-29.
Foradori, P. (2014). Reluctant disarmer: Italy's ambiguous attitude toward NATO's nuclear weapons policy. European Security, 23(1), 31-44.
Foradori, P. & Rosa, P. (2010). Italy: Hard Tests and Soft Responses. In: Kirchner, E. J., & Sperling, J. (Eds.). (2010). National security cultures: patterns of global governance. Routledge.
Frankel, B. (1993). The brooding shadow: Systemic incentives and nuclear weapons proliferation. Security Studies, 2(3-4), 37-78.
Freedman, L. (1994). Great powers, vital interests and nuclear weapons. Survival, 36(4), 35-52.
Freedman, L. (2004). Deterrence. Cambridge, UK.
Garnett, S. W. (1995). Ukraine's Decision to Join the NPT. Arms Control Today, 25(1), 7.
George, A. L., & McKeown, T. J. (1985). Case studies and theories of organizational decision making. Advances in information processing in organizations, 2(1), 21-58.
Grieco, J. M. (1993). Anarchy and the Limits of Cooperation: A Realist Critique of the Newest Liberal Institutionalism. In: D. A. Baldwin (Ed.), Neorealism and Neoliberalism: The Contemporary Debate. (pp. 116-140). New York: Columbia University Press.
Harrell, E. (2008, June 19). Are US Nukes in Europe Safe? TIME. Retrieved 2018, April 17 from:
http://content.time.com/time/world/article/0,8599,1816035,00.html Hudson, V. M. (2013). Foreign policy analysis: classic and contemporary theory.
Hymans, J. E. (2006). Theories of nuclear proliferation: The state of the field. Nonproliferation Review, 13(3), 455-465.
Karns, M. P., Mingst, K. A., Stiles, K. W. (2015). International Organizations: The Politics and Processes of Global Governance. London: Lyenne Rienner Publishers.
Keohane, R. O., & Martin, L. L. (1995). The promise of institutionalist theory. International security, 20(1), 39-51.
Kimball, D. (2017, July). The Presidential Nuclear Initiatives (PNIs) on Tactical
Nuclear Weapons at a Glance. Fact Sheets & Briefs – Arms Control Association. Retrieved 2018, March 13 from:
https://www.armscontrol.org/factsheets/pniglance
Kitchen, N. (2010). Systemic pressures and domestic ideas: a neoclassical realist model of grand strategy formation. Review of international studies, 36(1), 117-143. Kristensen, H. M. (2005). US nuclear weapons in Europe. Natural Resources Defense
Council.
Kristensen, H. M. (2015, September 10). Upgrades at US Nuclear Bases In Europe Acknowledge Security Risk. Strategic Security – Federation of American Scientists. Retrieved 2018, March 15 from:
https://fas.org/blogs/security/2015/09/nuclear-insecurity/
Lobell, S. E. (2009). Threat assessment, the state, and foreign policy: a neoclassical realist model. Neoclassical realism, the state, and foreign policy, 42-74. May, M. M. (1994). Nuclear weapons supply and demand. American Scientist, 82(6),
526-537.
Mearsheimer, J. J. (1994). The false promise of international institutions. International security, 19(3), 5-49.
Mearsheimer, J. J. (2001). The tragedy of great power politics. WW Norton & Company.
Morgenthau, H. J. (1948). Politics Among Nations: The Struggle for Power and Peace. New York: Alfred A. Knopf.
NATO (2010, November 19). Active Engagement, Modern Defence – Strategic Concept for the Defence and Security of the Members of the North Atlantic Treaty
Organization.
Nichols, T., Stuart, D., & McCausland, J. D. (2012). Tactical Nuclear Weapons and NATO. Carlisle, PA: Strategic Studies Institute. US Army War College http://www.strategicstudiesinstitute.army.mil/pdffiles/PUB1103.pdf. Norris, R.S. & Kristensen, H. M. (2011). US tactical nuclear weapons in Europe.
Bulletin of the Atomic Scientists, 67(1), 64-73.
Nuti, L. (2005). The richest and farthest master is always best: US-Italian relations in historical perspective. The Atlantic alliance under stress: US-European relations after Iraq, 177-198.
Oppermann, K. (2010, September). The concept of issue salience in foreign policy analysis: Delineating the scope conditions of theoretical approaches in the field. In SGIR 7th Pan-European Conference on IR (pp. 9-11).
Oppermann, K., & Spencer, A. (2013). Thinking alike? Salience and metaphor analysis as cognitive approaches to foreign policy analysis. Foreign Policy
Oppermann, K. (2014). Delineating the scope conditions of the poliheuristic theory of foreign policy decision making: The noncompensatory principle and the domestic salience of foreign policy. Foreign Policy Analysis, 10(1), 23-41. PAX (2016). Teken tegen kernwapens. Retrieved 2018, April 23 from:
https://www.paxvoorvrede.nl/wat-wij-doen/campagnes/teken-tegen-kernwapens#
Putnam, R. D. (1988). Diplomacy and domestic politics: the logic of two-level games. International organization, 42(3), 427-460.
Rachwald, A. R. (2011). A ‘reset’ of NATO–Russia relations: real or imaginary? European Security, 20(1), 117-126.
Ripsman, N. M., Taliaferro, J. W., & Lobell, S. E. (2016). Neoclassical realist theory of international politics. Oxford University Press.
Ronzitti, N. (2007). Le basi americane in Italia – problemi aperti. Senato della Repubblica – servizio studi, servizio affari internazionali, 70.
Ronzitti, N. (2017, November). Lo stato del disarmo nucleare. Osservatorio di Politica Internazionale – IAI, 77.
Rose, G. (1998). Neoclassical realism and theories of foreign policy. World politics, 51(1), 144-172.
Sagan, S. D. (1997). Why do states build nuclear weapons? Three models in search of a bomb. International security, 21(3), 54-86.
Sagan, S. D. (2011). The causes of nuclear weapons proliferation. Annual Review of Political Science, 14, 225-244.
Sauer, T., & Van der Zwaan, B. (2012). US tactical nuclear weapons in Europe after NATO’s Lisbon summit: why their withdrawal is desirable and
feasible. International relations, 26(1), 78-100.
Schelling, T. C. (1966). Arms and Influence. New Haven: Yale.
Schultz, K. A. (1998). Domestic opposition and signaling in international crises. American Political Science Review, 92(4), 829-844.
Shultz, G. P., & Goodby, J. E. (Eds.). (2015). The War that Must Never be Fought: Dilemmas of Nuclear Deterrence. Hoover Press.
Schweller, R. L. (1998). Deadly imbalances: Tripolarity and Hitler's strategy of world conquest. Columbia University Press.
Schweller, R. L. (2004). Unanswered threats: A neoclassical realist theory of underbalancing. International security, 29(2), 159-201.
Shea, J. (2009, March 4). 1979: The Soviet Union deploys its SS20 missiles and NATO responds. Lecture. Retrieved 2018, April 22 from:
https://www.nato.int/cps/en/natohq/opinions_139274.htm?selectedLocale=en Snyder, G. H. (1961). Deterrence and Defense: Toward a Theory of National Security.
Princeton, NJ: Princeton University Press.
Snyder, G. H. (2002). Mearsheimer's World—Offensive Realism and the Struggle for Security: A Review Essay. International Security, 27(1), 149-173.
Soroka, S. N. (2003). Media, public opinion, and foreign policy. Harvard International Journal of Press/Politics, 8(1), 27-48.
Taliaferro, J. W. (2006). State building for future wars: Neoclassical realism and the resource-extractive state. Security Studies, 15(3), 464-495.
Thayer, B. A. (1995). The causes of nuclear proliferation and the utility of the nuclear non-proliferation regime. Security Studies, 4(3), 463-519.
UN Office of Disarmament Affairs (2005). 2005 Review Conference of the Parties to the Treaty on the Nonproliferation of Nuclear Weapons (NPT/CONF.2005/57 – Part II).
Van der Zeijden, W. (2014). A Dutch revolt? The salience of the nonstrategic nuclear weapons issue in Dutch politics. European Security, 23(1), 45-57.
Walt, S. M. (1990). The Origins of Alliance. Cornell University Press.
Waltz, K. N. (1979). Theory of International Politics. Reading, MA.: Addison-Wesley Publishers.
Wlezien, C. (2005). On the salience of political issues: The problem with ‘most important problem’. Electoral Studies, 24(4), 555-579.
Wohlforth, W. C. (1993). The elusive balance: power and perceptions during the Cold War. Cornell University Press.
Zakaria, F. (1998). From wealth to power: The unusual origins of America's world role. Princeton University Press.
Interviews
Cotta-Ramusino, P. (2018, April 30). Secretary General of Pugwash Conferences on Science and World Affairs. Professor of Mathematical Physics at the University of Milan. Associate of the John F. Kennedy School of Government at Harvard University. Member of the International Institute for Strategic Studies (IISS). Founder of the Italian Union of Scientists for Disarmament (USPID).
Policy Officer at the Netherlands Ministry of Foreign Affairs (2018, April 23). Simoncelli, M. (2018, May 10). Vice-president and cofounder of Istituto di Ricerche
Internazionali Archivio Disarmo (IRIAD – Institute for International Research Disarmament Archive).
Van der Meer, S. (2018, April 16). Research Fellow at The Netherlands Institute of International Relations Clingendael.
Voorhoeve, J. J. C. (2018, April 16). Chair of the Peace and Security Committee of the Advisory Council on International Affairs (AIV) of the Netherlands. Member of the AIV. Professor of International Organisations at Leiden University. Former Minister of Defence of the Netherlands (1994-1998). Former political and parliamentary leader of the People's Party for Freedom and Democracy – VVD (1986-1990).
