Introduction
In 1981, US Secretary of State Alexander Haig threatened to withdraw Norway’s security guarantee. Norway, a founding member of NATO, had proposed to establish a Nordic nuclear weapon free zone (NNWFZ), a regional ban on deployment, development, and transport of nuclear weapons. Haig could not ‘recommend his president to send reinforcements to Norway if such a zone was established’, he told his Norwegian counterpart, Knut Frydenlund.Footnote 1 Frydenlund, who received the message on an official visit to Washington DC, was apparently ‘shocked’ by Haig’s ‘punch to the gut’ response.Footnote 2 One wonders: why would a small state in an asymmetric alliance oppose their nuclear security guarantor? This article analyses the Norwegian decision and policy-making process preceding this seminal meeting. The purpose is twofold: first, to examine the Norwegian case, and second, based on the examination, to develop the autonomy–abandonment dilemma theoretically.
Alliance theory would expect smaller allies to be compliant with the desires of the state that extends nuclear deterrence to them. This expectation rests on an assumption that a state’s action results from a strategic calculus. In asymmetric alliances, no concern looms larger than keeping the nuclear patron’s security guarantee intact. The weaker party must be accommodating to avoid unfavourable responses. In the alliance literature, this follows from the fear of being abandoned.Footnote 3 The same literature accounts for intra-alliance opposition, when it occurs, as triggered by the fear of entrapment, that is, the fear of being dragged into conflicts over interests that are not shared.Footnote 4
Few cases of alliance opposition lead the patron to revoke its security commitments. All such cases deserve exploration because they break with theoretical expectations for umbrella states’ behaviour.Footnote 5 Two reasons stand out as meriting attention in the Norwegian case. First is Norway’s unique geostrategic location. For the Soviet Union to attack the United States with nuclear weapons, the shortest distance would be to fire them from Norway’s border area, the Kola Peninsula.Footnote 6 Norway is therefore exposed to Russian measures to protect their second-strike capability.Footnote 7 This affects Norway’s threat assessment today, as well.Footnote 8 It is therefore surprising when Norway, given their familiarity with Russian nuclear capabilities and threat assessments, opposes its security guarantor’s nuclear policy, thereby risking nuclear abandonment.
Second, Norway has had a track record of navigating a more autonomous position on nuclear policy within NATO and with the United States’ acceptance through so-called self-imposed restrictions.Footnote 9 These restrictions included a ban on permanent foreign bases and storage of nuclear warheads in peacetime. Norway was, therefore, experienced in navigating the balance between asserting autonomy and loyalty as a member of the alliance. Norway, which seems to have recognised the limits of its room for manoeuvring within the alliance, failed to navigate the balance in the early 1980s in connection with the NNWFZ.
To explain why Norway opposed US nuclear policy in 1981, I argue for a theoretical approach, which I call the autonomy–abandonment dilemma, and examine the bargaining process domestically. At its core, autonomy is ‘a state’s ability to decide its own policy’.Footnote 10 To increase the analytical precision of the concept, I distinguish between two subtypes of autonomy: the first views policy as a means to achieve a goal (instrumental autonomy), and the second views it as a means to signal its goal (expressionistic autonomy).
To analyse this case, the Norwegian decision-making process will be examined based on archival material from the Norwegian National Archives, the Norwegian Labour Movement’s Archives, governmental documents, and memoirs.
The article will proceed as follows. First, I outline the alliance literature that explains policy incentives for states in asymmetric alliances and derive expectations for the case study. I also argue for an approach that examines the domestic deliberation process. Then, I present the case analysis of Norway’s proposal to establish a NNWFZ. This analysis shows that two arguments were invoked during the process: (a) opponents feared abandonment and (b) proponents desired autonomy. Examining the bargaining process reveals how the final policy resulted from diplomatic handcraft by specialist policymakers, who balanced strong domestic sentiments in an unfavourable context with allied reactions. In a context where nuclear angst loomed over the Norwegian population and the government’s authority was weak, the desire to assert nuclear autonomy increased.
Based on the data, I then argue for a new framework that is client state–specific – the autonomy – abandonment dilemma – and further develop the concept of autonomy. Importantly, I conclude that the case has implications for how we interpret a state’s intentions based on the outcome of a specific policy. What at first glance appeared as free riding, that is, benefitting from a public good without contribution to it, was on inspection the result of not cynical exploitation of the nuclear patron but rather domestic bargaining. That policy’s strategic implications resulted from the management of the conflict between strategic considerations and domestic political pressures.
Theoretical framework: Understanding why umbrella states oppose their patron’s nuclear policies
Given the dependency the client state has on their patron for security, why do they oppose their policies? In this section, I present the theoretical assumptions underlying this puzzle by first outlining the basic logic of extended deterrence and the alliance dilemma. I then present recent literature on entrapment concerns in client states before deriving three expectations for explaining Norway’s decision to adopt a policy of an NNWFZ in 1981.
In essence, extended deterrence refers to a situation where one state – referred to as the patron – seeks to deter an adversary from attacking a third party – the client.Footnote 11 The patron state has more resources than the client. An adversary of these states must consider the risk that a conflict with one will trigger the inclusion of resources of the other. This patron–client relationship is established to increase the cost of attacking, namely that it deters attacks on the client state as well as the patron state. The unfolding of this logic has been studied in many situations.Footnote 12
States’ management of this extended deterrence relationship is traditionally presented as a dilemma between abandonment and entrapment.Footnote 13 On one hand, states fear being abandoned by their ally. That is, states fear that allied states will defect or fail to act on their security guarantees. On the other hand, states fear entrapment. That is, that the allied state will ‘drag them into conflict over an ally’s interests that one does not share’.Footnote 14
Presumably, fears of abandonment and entrapment are distributed asymmetrically between the patron and the client.Footnote 15 In asymmetric alliances, patron states fear entrapment more than client states because they provide the security guarantee.Footnote 16 Consequently, they must engage their resources on their client’s behalf should they be in conflict. This situation is said to make patron states concerned that client states will pursue more aggressive policies because they know they have the resources of a patron state to back them, usually referred to as a problem of ‘moral hazard’.Footnote 17
On the other hand, client states fear abandonment more than patron states because they receive security. A patron that has nuclear weapons provides the client with the most fundamental guarantee. According to NATO, nuclear weapons are the ‘supreme guarantee of security’ of the alliance.Footnote 18 Presumably, receiving such a security guarantee can be existential for umbrella states.
If a client state faces a more powerful adversary than itself, being alone would increase the risk of both deterrence and defence failure. Because of this, the client states fear that the patron state will fail to honour its guarantee or provide weaker support than needed in a crisis.Footnote 19
Consequently, whether a state primarily fears abandonment or entrapment shapes its incentives to support or oppose its ally’s policies. Fears of abandonment incentivise loyalty to the patron and support of its policy preferences.Footnote 20 This is what makes policy opposition by client states, who are assumed to fear abandonment, appear puzzling. By contrast, opposition can be incentivised behaviour for a state more concerned with being entrapped. Entrapment-concerned states will seek to mitigate risks of being dragged into unwanted conflicts, for example, by opposing aggressive policies by the ally and maintaining flexibility through using ambiguous language in their defence agreement with the ally.Footnote 21
While alliance management theory has emphasised the unequal distribution of abandonment and entrapment concerns in asymmetrical alliances, recent literature by political scientist Lauren Sukin suggests that client states also are concerned with entrapment.Footnote 22 However, she argues that entrapment concerns in client states may be understood differently than in patron states. Instead of worrying whether an ally will involve you in a conflict they have initiated, client states may worry over whether the patron will escalate a conflict unnecessarily. Sukin writes, ‘nonnuclear client states of nuclear allies might worry that they could become involved in a nuclear exchange because of behavior and capabilities of their nuclear ally, [and] the desire to avoid nuclear escalation of a crisis or conflict is what drives the unwanted use theory’.Footnote 23 This ‘unwanted use theory’ is essentially a theory of adverse entrapment.
Entrapment concerns can therefore help explain why client states sometimes oppose their patron’s policies. Although client states are generally more sensitive to fears of abandonment, entrapment concerns are not absent. As Sukin and Alexander Lanoszka argue, ‘we find that allies evaluate reassurance policies in a way that pays greater attention to escalation and entrapment risks than previous work has most often theorized’.Footnote 24 In other words, when client states consider and respond to their patron’s policies, they evaluate the policies and consider more costs than fear of abandonment. Altogether, the literature suggests that client states navigate abandonment and entrapment concerns in response to their patron’s preferred policies.
Based on existing literature on alliance management in extended deterrence relationships, I derive three expectations for the debate prior to Norway’s adoption of an NNWFZ in 1981.
First, if entrapment is the core source of opposition, we should observe Norwegian policymakers presenting arguments that link the proposal of an NNWFZ to escalation management. This means that policymakers will consider the situation without a NNWFZ as risky: that Norway is prioritised as a target because of its connection to US policy, or that Norway lacks control over how the US uses Norwegian territory or capabilities to enhance its interests rather than Norway’s.
Second, if abandonment concerns are present in the debates, we should observe Norwegian policymakers arguing that adopting a policy of an NNWFZ may risk US dissatisfaction or the US leaving the alliance.
Third, because the alliance management literature suggests that alliance members navigate a dilemma, the last expectation for observation is that Norwegian policymakers are torn between fear of abandonment and fear of entrapment. This means that both entrapment and abandonment concerns should be present in the debates.
How the domestic bargaining process affects the alliance dilemma
Although the alliance dilemma illustrates the different interests within an umbrella state, it insufficiently explains why client states would fail to prioritise abandonment concerns. To understand this, it is necessary both to apply the dilemma to the decision-making process domestically and to examine the bargaining process.
Scholars have rarely attended to domestic differences in alliance concerns in the existing literature on the traditional alliance dilemma (entrapment and abandonment). Now, they are calling for research to explore ‘how the public’s nuclear anxiety shapes policy and (…) investigate the differences between public opinion and elite preferences’.Footnote 25 The investigation has started.
Recent literature found differences between public opinion and elite preferences in states allied with the United States, indicating that more attention should be given to how domestic differences shape alliance policies. Sukin and Lanoszka studied public support in US allied states (Estonia, Latvia, Lithuania, Poland, and Romania) for different US (intra-alliance) reassurance policies.Footnote 26 They found that support for economic sanctions or public statements – rather than military deployments, which are often thought to be reassuring to US allies – were seen as reassuring and as a better tool for escalation management.Footnote 27 Allison et al. studied public support for US extended nuclear deterrence policies.Footnote 28 They found that public support for nuclear retaliation is low in the United States, Japan, and South Korea.Footnote 29 Lastly, when asking what the key drivers of nuclear anxiety were, Sukin and Seo found concerns for both sides of the traditional alliance dilemma present in the publics of Australia, Japan, South Korea, Taiwan, and Indonesia.Footnote 30 Contrary to these studies, this article situates these sentiments within the alliance dilemma framework.
Paying attention to the interplay of domestic sentiments and policymakers’ strategic considerations can provide a better understanding of umbrella states’ policies, which is key to understanding how states manage alliances.
Norway’s proposal to establish an NNWFZ in 1981
In this section, I analyse the decision-making process behind Norway’s 1981 proposal to establish an NNWFZ. At the time, Norway was governed by a majority Labour Party government, and I focus on the Labour Party’s internal decision-making process. The discussions were mainly driven by two arguments: (a) opponents feared abandonment and (b) proponents desired autonomy. This means that the data did not concur with the theoretical expectation that proponents of an NNWFZ feared being entrapped or losing control over escalation management. Instead, the proponents of the proposal were motivated by increased autonomy.
In the second part of this section, I show how the bargaining process between the proponents and opponents challenged the expectation that abandonment concerns would dominate the outcome.
Opponents feared pursuing an NNWFZ would risk abandonment
Acting in line with the expectation that umbrella states assert loyalty out of fear of abandonment, key policymakers opposed the NNWFZ proposal. The key positions – all from the Norwegian Labour Party, which was the majority government – include the minister of foreign affairs, the defence minister, and the state secretary to the Ministry of Foreign Affairs. These positions have in common that they interacted with allied states and were exposed to allied reactions to Norwegian policy decisions. They argued that pursuing an NNWFZ could hurt NATO relations and risk the security guarantee.
On multiple occasions, Foreign Minister Knut Frydenlund exhibited scepticism of the proposal, fearing allied reactions and abandonment. In a January 1981 brief to NATO regarding the Norwegian NNWFZ proposal, he emphasised Norwegian loyalty to NATO.Footnote 31 Because this briefing took place three months before the Labour Party officially adopted the proposal, it indicates that allies took a special interest in Norway potentially pursuing a NNWFZ. Frydenlund shared his concerns with allies about what he deemed was a growing anti-nuclear movement in Norway. Fearing it could ‘escalate into an anti-NATO posture’, he stated that the Norwegian government would work to prevent this.Footnote 32 He made it clear to the allied states that Norway would not undermine its position in NATO as a loyal and firm ally.
At the same time, sharing his concern with allies over the anti-nuclear movement’s potential consequences could have been interpreted as Frydenlund trying to prepare allies for the upcoming adoption of the policy, and why he would take what appeared to be a passive role during the debates at the Labour Party’s convention. Although sceptical of the proposal, he did not participate in the debates or express his concerns about abandonment in other public ways at the party convention. However, he communicated his beliefs to NATO and his parliament through briefings he held before and after the convention.
The underlying message of Frydenlund’s opposition was that he was unwilling to risk Norway’s NATO membership to establish an NNWFZ. Later in 1981, after the NNWFZ was adopted, came the infamous meeting in which US Secretary of State Alexander Haig threatened to withdraw the US security guarantee to Norway. After returning to Norway, Frydenlund briefed the Norwegian Parliament’s Enlarged Foreign Affairs Committee on allied reactions to the adoption and the US threat of withdrawing the security guarantee. He told the committee that he had reassured the United States, convincing Haig that Norway would not pursue the NNWFZ without consulting allies.Footnote 33
According to Frydenlund, it was important that the zone be reciprocal. This meant that the Soviet Union would undertake the same measures to remove the weapons in this area, and it would not affect NATO’s arms-control negotiations with the Soviet Union. He was referring to NATO’s 1979 dual-track decision that NATO would deploy intermediate-range nuclear weapons in Europe in response to Soviet deployment, should arms-control negotiations fail. During his meeting with Haig, Frydenlund had also underlined how Norway would not pursue any policies changing their alliance obligations.Footnote 34 His brief to the parliament committee showed that he took on a role in national (parliament) and international (NATO and bilaterally with the United States) arenas to ensure Norway’s strategic interests remained intact by demonstrating alliance commitment.
Similarly, for State Secretary to the Ministry of Foreign Affairs Johan Jørgen Holst, securing alliance loyalty to ensure Norway’s security interests was the chief objective. Providing an even harsher critique of the proposal than Frydenlund, Holst argued that it would remove Norway from NATO’s defence strategy. The strategy (MC-14/3) did not explicitly exclude nuclear first use, but the NNWFZ would.Footnote 35 In a note to the foreign minister, Holst even criticised the proposal’s language as ‘unfortunate’ and ‘poorly conceived’.Footnote 36 He was especially concerned with the formulation stating the proposal was a ‘step in a broader solution for the European area’, which he thought was problematic for European allies who did not agree to pursue this policy.Footnote 37
Holst wrote the note expressing his firm conviction after participating in a meeting with the North Atlantic Council. Evidently, this meeting had convinced him that the proposal ‘conflict[ed] with important foreign policy realities and Norwegian security policy interests’, further arguing that its ambiguities would exacerbate problems in their foreign policy and on a domestic level.Footnote 38 In a different note later in 1981, Holst argued that parts of the NNWFZ proposal were incompatible with alliance membership. Referring to specific parts, he again argued that the proposal would explicitly remove Norway from NATO’s defence strategy and described it as ‘highly unlikely’ that its restrictions would be compatible with alliance membership.Footnote 39 Undoubtedly, Holst’s encounter with allies reinforced his conviction that the proposal was unfit for a NATO member.
Correspondingly, then Defence Minister Thorvald Stoltenberg had not wanted to push the pursuit of an NNWFZ because it would risk the US security guarantee and have ‘dramatic consequences for the defence cooperation between the two countries’.Footnote 40 He wrote in his memoirs that he ‘reacted strongly’ to the proposal when he first heard it. He described the meeting between Frydenlund and Haig and the reactions in Washington as ‘one of the most dramatic meetings between a Norwegian minister of foreign affairs and US Secretary of State after the war’.Footnote 41 At least to him, it was clear that the United States was weary of Norway’s self-imposed restrictions and had met their limit. Contrary to Holst and Frydenlund, Stoltenberg did not refer to meetings in which he had participated. However, as the defence minister, he also participated in NATO and bilateral meetings with the United States.
A caveat is necessary for Stoltenberg’s opposition. Stoltenberg wrote his memoir years later with the benefit of hindsight. In July 1981, he told the Norwegian newspaper Arbeiderbladet that Norway would be willing to give up NATO’s nuclear weapon reinforcements with an NNWFZ.Footnote 42 This statement could be interpreted as Stoltenberg favouring the policy.Footnote 43 At the same time, it could be read as describing what the new policy entailed rather than reflecting his personal views. After all, his interview with Arbeiderbladet occurred after the party had adopted the policy and in the last months before a parliamentary election. With an upcoming election, policymakers might disguise their personal opinions to avoid public disagreement within the party. For that reason, Stoltenberg’s retrospective engagement with the issue is included in this section.
Retrospective engagement with the policy proposal is more challenging to interpret when reading then Prime Minister Oddvar Nordli’s account. In a 2005 interview, he said he ‘was sceptical to declare ourselves a nuclear-free area. I was worried we would upset the Nordic balance that had substantial value for the security policy.’Footnote 44 The balance he referred to was what he considered a stable situation in the High North, a sensitive area. However, in his New Year’s Eve speech in 1981, he said,
Our no to nuclear weapons is in itself a contribution to peace and détente in our part of the world. We must be willing to commit by treaty this nuclear-free status if this can happen in a broader context in our part of the world, where already nuclear weapons are deployed and more are planned to be deployed.Footnote 45
According to Aftenposten’s correspondent to the United States, the United States had reacted strongly to Nordli’s stance in ‘such an important speech without consulting allies’. The Norwegian ambassador had to meet with the US government to provide a further account of Norway’s stance.Footnote 46 In contrast to Stoltenberg, Nordli gave his speech before the party adopted the policy. His statement was also more normative, directly stating that Nordli thought the NNWFZ was what Norway ought to do. For these reasons, Nordli’s 2005 account of his thinking on the NNWFZ is not included in this section as further evidence of policymakers’ opposition.
Frydenlund, Holst, and Stoltenberg were not only elite but also specialist policymakers, in positions specialising in foreign affairs, defence, and security. Specialist policymakers expressing concerns aligned with the fear of abandonment might be interpreted as indicating which positions policymakers hold, whether they engage directly in security and defence issues with allied states, and how they balance the alliance dilemma. The positions they held in this case support the theoretical claim that umbrella states, or specialist policymakers within them, are concerned about being abandoned by their nuclear patron.
Proponents desired autonomy
During the Labour Party’s convention of April 1981, where the NNWFZ policy was adopted, the debates showed massive support for the proposal. Despite the specialist policymakers’ opposition, no participant argued against the proposal during the debates. In this section, I analyse the arguments of those who supported an NNWFZ presented at the Labour Party’s convention and supplement them with a key statement by the party’s prime minister at the time.Footnote 47 Despite entrapment being a central expectation in the alliance literature, I find little evidence that fears of being entrapped in US conflicts or losing control over Norwegian territory motivated the support for an NNWFZ. Instead, proponents consistently emphasised preserving Norway’s room for manoeuvre and self-determination in its security policy.
The NNWFZ proposal was considered a way for Norway to take an active standpoint and make amends for earlier policies.Footnote 48 Some debaters appealed to the passiveness and fear felt during the recent years’ events, referring to NATO’s 1979 dual-track decision.Footnote 49 They worried about it but – contrary to the specialist policymakers, who worried the NNWFZ would undermine NATO’s decision – NNWFZ proponents were concerned about the state of nuclear disarmament. A Labour-led government had supported NATO’s decision publicly and within the party, despite strong domestic opposition.Footnote 50 The dual-track opposers within the party had felt powerless when Norway agreed to deploy nuclear weapons on the European continent if arms-control negotiations failed. To them, the NNWFZ became a way to regain some of the autonomy lost during the dual-track decision.
Because negotiations with the Soviet Union were already difficult following the dual-track decision, an important aspect of the NNWFZ debate was whether to demand the inclusion of the Soviet Union. According to the participant Lauvsnes, this was a hopeless demand for change.Footnote 51 His argument was that to demand inclusion of all parts would be to make no change at all. If it were possible to achieve only a unilateral agreement, it would still be better than no agreement.Footnote 52 One must interpret this statement as a firm belief in the effect of all steps possible to actively engage in abolishing nuclear weapons.
A common attitude among the participants at the convention was that Norway should be active, and not passive, changers of policy.Footnote 53 They viewed the proposal as a specific, concrete step towards nuclear abolition, a way for Norway to change the current nuclear landscape actively and, importantly, to show that expressed beliefs on disarmament and abolition were not just ‘cheap talk’ or rhetoric.Footnote 54
To those participants, NATO membership was a responsibility and an enabler for change. Although the desire to assert autonomy came from forfeiting it to become an alliance member, membership had given Norway access to decision-making processes in a nuclear alliance. The alliance membership was exactly what provided Norway with an opportunity to affect nuclear policy and not be bystanders.Footnote 55
At the same time, the alliance membership clearly affected the experienced autonomy: one participant criticised opponents for ‘adjusting to what the American people can accept or not’.Footnote 56 Because the convention took place before Haig threatened to withdraw their security guarantee, the argument shows that American reactions were part of the calculations, even though the threat was not explicit at the time. It also captured the essence of autonomy loss in asymmetric alliances: adjusting policies to what the patron accepts.
Some elite policymakers favoured the proposal and argued based on a desire to increase Norway’s autonomy, reinforcing the argument that there is a difference between specialist and non-specialist policymakers. Labour Party Secretary Ivar Leveraas highlighted how the proposal could influence and have an impact on current developments.Footnote 57 As the previous section showed, elite policymakers in specialist positions did not share the same strategic considerations in deliberations.
The strongest argument for the NNWFZ as a policy of self-determination even came from Gro Harlem Brundtland, the prime minister during the convention. She made headlines with her statement, ‘Norwegian freedom of action is restricted!’Footnote 58 Brundtland made this statement after losing the election; hence, a caveat is that it came months after the adoption of the policy and the US threat of withdrawal. However, it is included here because it is strong evidence that policymakers in different positions considered Norway’s right to self-determination to be at stake.
Brundtland was commenting on the new government’s white paper on disarmament and their view on the NNWFZ, which contrasted with the Labour Party’s. Brundtland thought the Conservatives’ argument that formalising an NNWFZ would risk US abandonment: ‘a dangerous contribution to creating disagreements on important parts of Norwegian security policy’. She repeated this concern multiple times throughout the interview. She finished by saying that whether one wants to use the room for manoeuvre is irrelevant – but maintaining it is not. She said she was deeply concerned that the new government was limiting Norway’s room for manoeuvre and its foundation for self-determination.Footnote 59
Beyond concerns about room for manoeuvre, some proponents supported an NNWFZ because it expressed virtuous nuclear behaviour. It was therefore not a necessary condition for favouring the proposal to believe that NNWFZ would change the status quo. More specifically, some argued that adopting the NNWFZ would signal the Norwegian will to work for disarmament.Footnote 60
Reiulf Steen, the party leader (1975–81) who led the work with the new party platform ahead of the convention, argued during the debates:
Some believe we do not have the right to take initiatives like these. We believe we have a duty to do so because the situation in the world is dangerous. (…) For us, the proposal is a concrete obligation to take initiative, but it also represents an important symbol.Footnote 61
This quote shows two autonomy arguments: the first part shows a specific belief about what policy produces nuclear security. Steen believed the situation was dangerous, and Norway had a duty to undertake the initiative for change. I interpret the last part, that the NNWFZ presented an important symbol, as an intention to signal a goal.
In sum, most advocates viewed the NNWFZ as a means for Norway to change nuclear policy actively, that is, as a means of self-determination. The arguments they performed were rooted in beliefs on nuclear security that differed from those of NATO and, especially, the United States; they wanted to increase Norway’s nuclear room for manoeuvre within the alliance. The proponents highlighted the opportunity to ‘change the system from within’ and argued that Norway should work to change the nuclear status quo. Some proponents thought that an aspect of foreign policy in general and alliance policies in particular is the ability to signal one’s state’s values and goals to both allies and adversaries.
The domestic bargaining process’s effect on the outcome
An examination of the process revealed that the NNWFZ proposal was made without the specialists’ consent, in a context where the government had weak authority and the public expressed high levels of nuclear angst and powerlessness. This context demanded diplomatic handcraft from the policymakers, balancing strong domestic sentiments with allied reactions. Recognising this domestic bargaining process is necessary to understand why Norway broke with the expectation that they would prioritise their fear of abandonment and adopted the NNWFZ despite realistic abandonment concerns.
The proposal was first made without the knowledge or consent of the government at the time. Jens Evensen (ambassador and advisor in international law at the Ministry of Foreign Affairs) first proposed it at the Norwegian Chemistry Industry Worker Association in 1980 in his lecture, ‘Nuclear Weapons and Insecurity Policy’.Footnote 62 Appealing to the public’s nuclear angst, he argued that failing to adopt a NNWFZ could result in ‘public deception and Ragnarok’.Footnote 63 Minister of Foreign Affairs Frydenlund and Minister of Defence Stoltenberg reacted strongly to Evensen’s lecture. Frydenlund had not been informed ahead of time; on the contrary, he was under the impression that the lecture contained other content.Footnote 64 Stoltenberg thought Evensen had a habit of breaking the political line voiced by the party and the minister.Footnote 65 Evensen’s proposal ‘spread like wildfire’ in a Norwegian public feeling nuclear powerlessness, because the proposal seemed concrete and feasible.Footnote 66
According to Frydenlund, the Norwegian government’s authority was weak at the time,Footnote 67 especially due to NATO’s dual-track decision and Norway’s support. The decision had created tensions both in the public and within the Labour Party, with members of the government dissenting on the question. In the aforementioned brief Frydenlund provided for NATO allies on the upcoming NNWFZ proposal before the April 1981 party convention, he described the nuclear anxiety that the dual-track decision had generated in the public.Footnote 68 He specified in another note for internal ministry use that the decision had led to weak authority for the government, especially for the Ministry of Foreign Affairs and Ministry of Defence.Footnote 69 He was under the impression that even a minority proposal from the programme committee drafting the recommendations for the new party programme would be adopted at the upcoming convention.Footnote 70
The underlying message was that even though Frydenlund and other specialist policymakers in the Labour Party did not favour the proposal, it would be adopted – and they would have to find a way to navigate that situation. Otherwise, he worried that nuclear sentiments could develop into general opposition to NATO.Footnote 71 Frydenlund, it seemed, thought he faced a choice between two evils: a policy he strongly opposed (due to risk of being abandoned) and fear of nuclear opposition leading to more dire consequences (Norway being the one abandoning). His concern with the nuclear opposition was not unfounded.
Public nuclear opposition was strong. In the year after the convention, 125 municipalities declared themselves nuclear free, and 540,268 inhabitants – in a country of just over four million at the time – signed a campaign to declare Norway nuclear free.Footnote 72 Anticipating this engagement, Frydenlund argued ‘it was important that the growing movement had a feeling of being listened to, and that their engagement and anxiety is channelled in a way that can lead to results’,Footnote 73 although he had neither publicly nor privately hidden his beliefs on the matter.Footnote 74
To accommodate nuclear sentiments domestically without provoking allies, specialists worked in the background to change the proposal’s language before the convention. The first draft of the party’s programme said that an NNWFZ was a first step in the work for nuclear-free zones in a broader European context.Footnote 75 According to Holst, this made European allies react because they viewed it as ‘unwanted and unrealistic’. Holst successfully changed the language.Footnote 76 The final wording was, ‘Norway will work for a nuclear weapon free zone in the Nordic region as part of the work to downsize nuclear weapons in a broader European context’.Footnote 77 Unlike the first draft, the final policy did not indicate a belief that other European states would declare themselves nuclear free or challenge NATO’s dual-track policy: nuclear weapons could still be deployed on European territory as a response to Soviet nuclear behaviour.
By avoiding any challenge to NATO’s dual-track decision or alienating the alliance, they managed to make the proposal’s language more palatable to their allies. This is how Frydenlund and the other specialists accommodated the opposition, despite their own convictions, while avoiding the worst reactions. In the end, as we know, this almost failed: Haig threatened Frydenlund with withdrawing the security guarantee. However, it tells us that the US reaction was always part of the calculus – but we do not know what would have happened had they not changed the proposal’s semantics.
In sum, the final policy was the result of diplomatic handcraft by the specialist policymakers, who balanced strong domestic sentiments with allied reactions. Frydenlund, Stoltenberg, and Holst were convinced that the nuclear-anxious domestic audience would adopt even a minority proposal. At the same time, they were wary of allies’ reactions and had their own beliefs, unaligned with the public’s view. Despite the crafters’ abandonment concerns, the public’s desire to assert autonomy after NATO’s dual-track decision gave the specialists reduced leeway after Evensen proposed it publicly without their consent.
The autonomy–abandonment dilemma
Altogether, the data presented showed abandonment concerns, but there were no indications that entrapment concerns were decisive for the adoption of an NNWFZ. Instead, proponents were motivated by asserting autonomy. The case therefore appears as not a traditional dilemma between abandonment and entrapment but rather one between abandonment and autonomy.
In this section, I will review autonomy’s place in alliance literature. I will then connect the finding that autonomy was a decisive motive in the Norwegian case to a similar finding in the case of New Zealand. Based on the review of alliance literature and connection to previous findings, I argue that it is not surprising that client states face an autonomy–abandonment dilemma. To provide more analytical clarity to the concept of autonomy, I conclude this section by contending that there are two subcategories of autonomy: instrumental and expressionistic.
The alliance literature commonly distinguishes between alliance formation and alliance management. The alliance formation literature focuses on why states form alliances. In symmetric alliances, states have an interest in aligning and further managing their alliance because all parties increase their security by aggregating their capabilities.Footnote 78 In asymmetric alliances, such aggregation is thought to incompletely explain states’ interests because the patron state (typically) does not depend on the client for preserving their security interests.
Morrow has argued that in an asymmetric alliance, the patron receives autonomy and the client receives security. Security can be defined materialistically as one state using or threatening to use military power on behalf of another state (extended deterrence). Morrow provided a broader definition, fleshing out the underlying motivation behind defence capabilities: he defined a nation’s security as ‘its ability to maintain the current resolution of the issues that it wishes to preserve’.Footnote 79 Simply, security can be understood as the capability to preserve desired territory and policies.
The state receiving autonomy gains increased room for manoeuvre – the possibility to propose, decide, and execute desired policy more easily. Contrasting with security, autonomy is the ability to change, not preserve, the status quo to one’s own desire.Footnote 80 As Morrow contended, it ‘captures one facet: that of external self-assertion, of the general notion of autonomy as a state’s ability to determine its own policy’.Footnote 81 In sum, patron states increase their policy options by providing security guarantees to client states.
Autonomy is a broader concept than entrapment. Whereas entrapment is the risk of being undesirably involved in conflicts, autonomy captures a state’s ability to determine and engage in all desired foreign, defence, and security policies. The two concepts share the core assumption that states prefer self-determination, and the loss of either constrains a state’s room for manoeuvre. This similarity went largely unnoticed in Morrow’s formulation and has not been adequately addressed in subsequent alliance literature.Footnote 82
Autonomy is not an unfamiliar term in international relations theory, but its operationalisation varies.Footnote 83 It is commonly operationalised as reducing dependencyFootnote 84 or as the option to refrain from integrating undesired policies (what Haugevik and Rieker called ‘formal autonomy’).Footnote 85 Although these definitions, especially the last one, also concern the state’s right or option to self-determine, it is negatively defined: it is freedom from external interference. The operationalisation with which I work, Morrow’s definition, is a positive one: autonomy is the freedom to pursue desired policy changes.Footnote 86
It has also previously been suggested that opposition by small states can be explained by a desire to assert autonomy. Although the literature on small state alliance opposition is scarce, political scientist Amy Catalinac studied a similar case: New Zealand’s nuclear weapon free zone declaration.Footnote 87 Catalinac concluded that asserting autonomy was the primary motivation for small state opposition in that case. Relatedly, Nikolaj Petersen describes US tensions with European allies in the 1980s as a situation in which small countries had to strike a ‘balance between alliance solidarity and political autonomy’.Footnote 88 The Danish ‘footnote policy’ resembled Norway’s in that many of the Danish footnotes concerned nuclear issues, including a proposal for an NNWFZ.Footnote 89 Petersen does not develop ‘autonomy’ as a separate analytical concept, but his work – like the Norwegian case – shows how domestic factors significantly shaped the policy options available to Danish governments.Footnote 90
Based on Morrow’s theoryFootnote 91 on asymmetric alliance formation and Catalinac’s findingsFootnote 92 on small state opposition, it is not surprising to find that Norwegian policymakers in 1981 also wanted to assert autonomy. After all, Morrow’s theory resonates with Catalinac’s findings: states desire autonomy. However, neither Morrow’s theory nor Catalinac’s findings resonate with traditional alliance management theory, which represents the dilemma as one between abandonment and entrapment and does not include autonomy. This indicates that the dilemma between entrapment and abandonment does not fully consider the differing interests of states in asymmetric alliances.
These asymmetric interests shape the incentives and constraints that states face within the alliance, and thus the types of dilemmas they confront. States in this position can face several different dilemmas. Building on the Norwegian case, the literature on asymmetric alliance formation, and previous findings from the New Zealand case, I propose the autonomy–abandonment dilemma as a framework to help fill the gap in the literature on asymmetric alliance management.
If we accept the position that client states give their patron autonomy in exchange for security, then the question for client states is how much autonomy they should, or will, forfeit. At the same time, the threat of being abandoned by their patron looms over policy decisions. Therefore, presenting it as a dilemma between fear of abandonment and desire for autonomy can, in some cases, reflect deliberations and challenges in client states better than the traditional alliance dilemma between entrapment and abandonment.
In other words, the autonomy–abandonment dilemma looks like this: a client state has (domestically decided on) a desired policy change but is concerned that, should they pursue it, their patron will abandon them.
Two subcategories of autonomy: Instrumental and expressionistic
Looking closely at the autonomy motives presented in the data, there are in fact, two distinct categories. To provide more analytical clarity to the concept of autonomy, I therefore distinguish between two subcategories of autonomy-motivated arguments: instrumental and expressionistic.
Instrumental arguments see policies as a means of pursuing a specific goal; they are consequentialist. In the case under scrutiny in this article, advocates whose instrumental arguments favour ratifying an NNWFZ viewed the policy as a means to benefit their preferred goal of nuclear disarmament. This was the dominant subcategory of autonomy arguments. It includes for example the arguments that emphasised how the proposal was a specific, concrete step towards nuclear abolition, and how NATO membership provided Norway an opportunity to change nuclear policy.
Expressionistic arguments see policies as a means to convey specific attitudes; they are symbolic. In this case, they believed pursuing an NNWFZ would signal their goal (nuclear disarmament), whether or not it benefits this goal. This expression is considered important in itself. Although less present than instrumental autonomy, expressionistic autonomy was part of the motivation for an NNWFZ, as evidenced by Reiulf Steen’s argument that the NNWFZ ‘represents an important symbol’.Footnote 93
While instrumental autonomy was the more decisive motive in the Norwegian case, the opposite appears to be true in the New Zealand case. Based on Catalinac’s results, policymakers wanted to assert expressionistic autonomy. Explaining the results of her analysis, she wrote, ‘The act seemed to matter more than the outcome’.Footnote 94 She identified three reasons why New Zealand sought autonomy: (a) to express itself to the world and be known for something, (b) to be unique, and (c) to exercise moral leadership.Footnote 95 These arguments are different than arguments favouring an NNWFZ because of the policy’s ability to benefit their desired outcome; they are indeed primarily aimed at signalling certain moral values.
Both expressionistic and instrumental arguments assert autonomy and self-determination. However, it is analytically fruitful to distinguish between those convinced by the policy’s effect and those convinced by its signal. This distinction illustrates the diversity of the concept of autonomy and can contribute when critically examining the limitations of the different approaches to nuclear policy design.
Conclusion: Balancing and breaking the autonomy–abandonment dilemma
Based on the traditional alliance dilemma, Norway would have been expected not to pursue an NNWFZ because it would risk abandonment. Thus, it is puzzling when an umbrella state like Norway breaks these expectations. In this article, I set out to answer why Norway would break with the expectation that umbrella states are loyal to their nuclear patron.
The findings highlight autonomy as a motive for umbrella states. Proponents of the NNWFZ proposal wanted to assert nuclear autonomy, strengthening the conclusions of previous literature.Footnote 96 To develop the concept of autonomy in the literature further, I argue for distinguishing between two types of autonomy: instrumental and expressionistic. Motivated mainly by instrumental autonomy, proponents were convinced the NNWFZ would bolster disarmament efforts.
The findings also illustrate that small states in asymmetric alliances face a dilemma between autonomy and abandonment. Opponents of the proposal were concerned with being abandoned by the alliance, showing that it was part of the Norwegian calculus. Because NNWFZ proponents, to a lesser degree, showed concern with being entrapped in an unwanted conflict, the traditional alliance management dilemma between abandonment and entrapment unsatisfyingly portrays the dilemma Norway faced. Given the assumption that fears of abandonment and entrapment are distributed asymmetrically between the patron and the client, it is not surprising that another dilemma that is client state–specific would illustrate their different interests. In this case, the alliance dilemma is better represented as an autonomy–abandonment dilemma.
I examined the domestic decision-making process to understand why the expected balance between autonomy and abandonment was broken. My examination revealed that the policy adaptation was the result of a bargaining process between specialists’ strategic considerations and domestic nuclear sentiments. Specialist policymakers in umbrella states must consider not only what the patron state desires but also what their own public desires. These policy goals will not always overlap – a reminder that states are not unified actors, despite the alliance literature often treating them as such. This can lead to compromises where the specialists are dissatisfied with their state’s contribution.
This finding has implications for how we interpret a state’s intentions based on the outcome of a specific policy. If a state decides to restrict their contribution to an alliance, it is common in literature (and in policy) to assume that the state did so out of self-interest. More specifically, the assumed self-interests are either reducing economic burden or avoiding increased military risks. These states are accused of ‘free riding’, given the incentives to benefit from a public good without contributing to it in asymmetric alliances.Footnote 97 In the case of the NNWFZ, this accusation might have been justified. After all, Gro Harlem Brundtland, former prime minister from the Norwegian Labour Party, said,
An arrangement with nuclear weapon free zones does not mean that Norway renounces the right to be defended with nuclear weapons from the territory of other NATO countries. (…) We want nuclear weapons not to be used from Norwegian territory, but that nuclear weapons are sent from the territories of other countries.Footnote 98
This statement could be interpreted as Norway being unwilling to share the risk of contributing to a nuclear policy and only receiving the benefits (having nuclear deterrence) – ‘free riding’. However, the examination of the process showed that nuclear free riding was not a result of strategic considerations designed to cynically exploit the nuclear states but rather an unintended side effect of solving domestic disregard for alliance policy. Because policymakers motivated by a desire to change the nuclear status quo prevailed over specialist policymakers with strategic concerns, the outcome may appear strategically suboptimal when judged against alliance theory alone. However, the policy outcome is the product of domestic bargaining between policymakers with different policy goals rather than deliberate efforts to exploit the nuclear patron.
In the case of Norway’s decision to pursue an NNWFZ, opposition never led to the dire consequences Haig threatened. National and international events overtook the policy. In 1981, the new government was opposed to the proposal. When Labour formed a government again in 1986, they joined the Nordic initiative and cooperation for pursuing this policy. However, after the Cold War ended, so did the Labour Party’s nuclear policy.Footnote 99 This was not the case for New Zealand, which the United States abandoned because of the very same nuclear opposition.
Importantly, the findings also increase our understanding of how and why small states occasionally miscalculate the alliance dilemma. While autonomy and abandonment can be considered as competing aspirations, policymakers try to diminish the competition. Instead, they seek to navigate a balance between both aspirations where they maximise autonomy while minimising the risk of abandonment. This can be done intentionally, or it can be done out of necessity, as in the case under scrutiny in this article. Nevertheless, it indicates that the competition between autonomy and autonomy is not a dichotomy, where one chooses between the two aspirations to begin with, but rather a balance where both aspirations are present but with the potential to collide.
Future research could examine whether this autonomy–abandonment framework explains behaviour in other client states beyond Norway and New Zealand, to assess its empirical scope. Indications from existing work suggest that it may also be relevant for the Danish ‘footnote’ policy.Footnote 100 Several other NATO client states have imposed restrictions on nuclear policy.Footnote 101 Applying the autonomy–abandonment dilemma to their decisions could further test whether a client state–specific dilemma helps explain these outcomes. Finally, more research is needed to explore whether this framework can account for client state behaviour in policy areas beyond the nuclear realm.
Beyond testing the theory, future research could also investigate the United States’ motivations to abandon New Zealand while maintaining its security guarantee for Norway. Exploring this could increase our understanding of why patrons abandon their umbrella states. The different outcomes for Norway and New Zealand might indicate that different alliance commitments can create different ways of influencing patron states’ policies. New Zealand’s alliance, ANZUS, did not have an integrated defence structure, nor was the alliance postured to have member states assist each other with armed forces, as NATO does. Another potential road to explore is the United States’ use of coercion against more ‘peace’-oriented allies. In the alliance literature, it is better known how the United States coerces nuclear-ambitious allies, such as how they ‘thwarted’ Germany’s nuclear ambitions.Footnote 102 The US desire to avoid nuclear proliferation among allies has been considered a key interest for extending nuclear deterrence.Footnote 103 Less known in the academic literature is how the United States has managed allies working for more ambitious nuclear disarmament policies than themselves.
Acknowledgements
The author would like to thank James Cameron, Johannes Gullestad Rø, Raino Malnes, Henrik Stålhane Hiim, and the two anonymous reviewers for valuable feedback that improved the quality of the article. The author is also thankful for feedback received from participants at ‘Nasjonal fagkonferanse i statsvitenskap 2025’ and colleagues at SPOL.
Funding statement
This work was supported by the Norwegian Ministry of Defence.
Competing interests
There are no competing interests to declare.
Adelina Trolle Andersen is a doctoral research fellow in political science at the Norwegian Institute for Defence Studies (IFS) and the University of Oslo. Andersen is part of the research programme Norwegian Security Policy in Strategic Perspective (SPOL). Her thesis focuses on umbrella states’ nuclear policies.
Open access
