The influence of the EU Neutrals on European Security and Defence Policy

Table of Contents

Introduction
1. A possible explanatory factor: The preferences of neutral states
1.1. A Change in the International System…
1.2. …yields a change in state preferences…
1.2.1. A changed concept of “neutrality”
1.2.2. Changed Neutral State Preferences for ESDP functions
1.3. … and a change in the cooperation problem.

2. Operationalisation

3. Case Study: Ireland
3.1. 1996/1997 Intergovernmental Conference and the Treaty of Amsterdam (2001)
3.2. Cologne European Council (1999)
3.3. The Helsinki (1999) and Feira (2000) European Councils
3.4. Nice Intergovernmental Conference (2000) and the Treaty of Nice (2001)
3.5. The 2003/04 ICG and the European Constitutional Convention (2002-2004)

4. Conclusion
4.1. Summary and Discussion of Findings
4.2. Representativeness
4.3. Assessing the Applicability of Rational Choice Institutionalism

Works Cited

Introduction

In the space of a few years, the EU has made more progress on developing its European Security and Defence Policy (ESDP) than in the previous forty years of European integration. This has occurred despite the fact that four EU member states that are historically “neutral”, i.e. they are not members of NATO and are merely observers (not members) of WEU. Namely, these states are Austria, Finland, Ireland and Sweden (henceforth: EU neutrals). Many of the ESDP measures (unanimously!) adopted by the EU seem incompatible with such neutrality policies. How, then, has it been possible for ESDP to evolve so significantly since 1997 in light of the fact that the Union must accommodate the concerns of the EU neutrals?

In this paper, I argue that ESDP has been able to develop so rapidly because it has been crafted so as to allow for the participation of Austria, Finland, Ireland and Sweden without jeopardizing their neutrality policies. Specifically, this means that ESDP has adopted solely functions that reflect the preferences of EU neutral states^[1] – a preference that ESDP encompass matters concerning “security and defence cooperation”, whilst excluding a “common defence”. Thus my independent variable is the preferences of neutral states for ESDP. My dependent variable are the institutional functions that ESDP adopted between 1997 and 2004, as reflected in actual policy agreed upon by the European Council. I limit my investigation to a case study of the Irish government’s preferences for ESDP functions in relation to several specific instances of ESDP evolution (European Council meetings and Intergovernmental Conferences).

My paper proceeds as follows: The first section specifies my explanatory factor – the preferences of neutral states for ESDP – and draws on rationalist institutionalist theory to lay out my hypothesis that ESDP will only encompass those functions the neutral states favour. The second section discusses the operationalisation of the independent and dependent variables and presents the data sources to be used. In the third section, I engage in the case study analysis of Irish influence on ESDP. The fourth section summarises the findings, discusses the strengths and shortcomings of my study and rationalist institutionalism’s application to it.

1. A possible explanatory factor: The preferences of neutral states

The notion that governmental preferences are crucial in determining institutional functions derives from rationalist institutionalist theory. Rational choice (rationalistic) institutionalism holds that actors have fixed sets of exogenously determined preferences and behave in an instrumental, strategic manner in pursuing them.^[2] Maximizing the attainment of preferences often requires actors to direct their attention to institutions. This is because “institutions matter”: States have an interest in determining the function of institutions, because institutions have an impact on states’ abilities to pursue their interests. Thus “institutions not only set constraints to strategic action, they are themselves the object and outcome of strategic action”^[3]. Accordingly, we can expect states to invest a great deal of energy in ensuring that institutions are shaped in a way that reflects their preferences.

I do not claim that rationalist institutionalism has a monopoly on explaining institutional evolution. Indeed, a secondary goal of this paper is to critically assess rational choice institutionalism’s explanatory power with respect to the evolution of security institutions. I do so by testing a rationalist institutionalist claim (that government preferences are crucial determinants of institutional function) by means of a specific instance (Irish government preferences and ESDP evolution).

In the following, I use rational institutionalist theory to explain the evolution of ESDP and the impact that (neutral) state preferences have on this process.

1.1. A Change in the International System…

“When fundamental shocks occur, or over long periods of time, actors and their preferences do change. Under these conditions, the analyst has to explore how the fundamental interests of societies, and therefore the preferences of their agents (states or governments), may change.”^[4]

For rationalist institutionalists, the source of changes in state preferences – and thus the source of institutional evolution – begins at the level of the international system. For ESDP, this ‘fundamental shock’ is the end of the Cold War, which has left European states, neutral or not, faced with a fundamentally changed international security environment. The international system is no longer bipolar, reducing the chances of a major East-West confrontation, from which neutral states might be able to ‘stand aside’. Increasing interdependence between states has altered the implications for European states of insecurity elsewhere in world. Risks and threats are of a transnational, and, increasingly, global nature. Europe can thus be viewed as a “vulnerable island of stability, surrounded by instability and injustice”^[5]. Ensuring security requires preventing the development of crises, by means of preventive measures, containment, intervention to end conflicts and stabilisation efforts. As Naumann puts it, the task after the end of the Cold War became to create security not only in, but for Europe.^[6] The goal was to create security for Europe as well as for its immediate periphery.^[7]

This is a fact that is recognised by the neutral states. The 2000 Irish White Paper on Defence notes: “The new security environment in greater Europe, […] is marked by a lower degree of risk of large scale military conflict, but also by new challenges and uncertainties… leading to humanitarian crises and refugee flows which have affected every country in the EU… these countries have resulted in substantial zones of instability on or close to the border of the EU”^[8]. This awareness of the changed European security environment is reflected in the preferences of neutral states for ESDP, to which I now turn.

1.2. …yields a change in state preferences…

How, then, have these changes in the international system affected the EU neutral states’ historic neutrality policies? I explore this question in general first, and then go on to examine how I expect this to be reflected in the EU neutrals’ preferences for ESDP functions.

1.2.1. A changed concept of “neutrality”

“Neutrality”, during the Cold War, referred to an international legal status adopted voluntarily by a state committing itself to stand aside from conflagrations. Connolly defines neutrality as "a permanent policy orientation in peace as well as war of keeping out of pacts or alliances in peacetime with view to remaining neutral in any future war”.^[9] Moreover, the neutral state must constrain itself from aiding any of the contending parties and doesn’t participate in economic sanctions against one or the other party. During the Cold War, Austria legally bound itself to this status by enshrining it in its Constitution, while Sweden, Finland and Ireland adopted their neutrality policies in an informal – but equally convincing – fashion.^[10] During the Cold War, a policy of neutrality had several advantages for states, particularly for smaller ones: Firstly, a neutrality policy reduced the likelihood of being drawn into a conflict between major powers.^[11] Secondly, neutrality enhanced the diplomatic weight of small states by offering, if it so desired, the chance to adopt a mediatory function between greater powers.^[12] Finally, neutrality increases the sovereignty of a state, in that they are able to make autonomous decisions on matters such as defence spending.^[13]

“Neutrality” in the traditional, Cold War era sense is no longer of much value to states in the post-Cold War security environment. The case of the EU neutrals is a peculiar one, because their security policy of neutrality is challenged more by peacetime than it is by war. As Katsumi Ishizuka points out:

The post Cold War era has witnesses a more equal distribution of opportunities for states to contribute to peacekeeping operations, but it has also reduced the status of middle and neutral powers like Sweden, Finland and Ireland. Middle and neutral powers have also been faced with a situation that has forced them to reconsider the maintenance of their military status of neutrality in the current unipolar system of world politics.^[14]

Does this mean that the EU neutrals are abandoning neutrality as a policy? I believe that this is not the case. Indeed, the governments of all of the states in question have committed themselves to upholding ‘neutrality’, as is evident from Government White Papers and in statements by government officials^[15]. Moreover, the proposition that the EU neutrals are abandoning their neutrality fails to explain their behaviour with respect to ESDP – why the EU neutrals have not only supported some ESDP policies but also been adamant about rejecting certain other proposals.

The safeguarding of neutrality has two primary reasons, both related to the domestic politics level of analysis. Firstly, it reflects the desire of (most of the) political elite in EU neutral states to retain their traditional national sovereignty in defence matters.^[16] In Ireland, the Fianna Fail/Progressive Democrat government has continued – as have all past Governments – to reassure the electorate that nothing fundamentally has changed in Irish security and defence policy and that Ireland maintains its traditional policy of ‘military neutrality’. Strikingly, no political party – until quite recently^[17] – has formally called for any fundamental reappraisal of that policy. Secondly, and linked to the above, public opinion in the neutral member states is overwhelmingly in favour of “neutrality” as a policy. In Ireland, a 1996 opinion poll shows 69% of the population in favour of retaining Ireland’s policy of neutrality.^[18] The popularity of the policy is illustrated by the fact that attempts by opposition political parties to mobilise on the grounds of an “impending loss of neutrality” have proven to be highly successful. This sort of mobilisation campaign is frequently cited as the reason for the rejection of the Nice Treaty by the Irish people during the first referendum in 2001.^[19]

Nevertheless, it would be wrong to conclude that the neutral EU states have simply retained neutrality policies in the Cold War sense – indeed, “neutrality as we know it from the Cold War days is a ‘dead concept’”^[20]. Rather, the concept of “neutrality” has been redefined to suit their interests in an interdependent, post-Cold War security environment. This redefined concept of “neutrality”, to which the neutral EU states are adhering, requires further specification. There is a sense of vagueness surrounding the concept, which, Portela-Sais argues, is reflected by the fact that neutrality is rarely called by its name^[21]: Thus the Treaty of the European Union speaks of the “specific character of the security and defence policies of certain Member States”. Academic discussion has dubbed the group of states the “post-neutrals”. Indigenous policymakers, including Finish foreign minister Erkki Tuomioja claim that their countries are following a policy of “non-alignment”, though they haven’t joined the Non-Aligned Movement.^[22]

However, I argue that this vagueness is primarily a matter of political rhetoric as well as a public confusion that comes with a shift in conceptual meaning. Following rationalist institutionalist theory on the fixed preferences of states, I argue that the post-Cold War policy of “neutrality” has become quite unambiguous in recent years, and that this is reflected in state behaviour. Accordingly, I hypothesize, neutrality is now considered by the EU neutrals to imply not being bound into an “alliance” by a mutual defence commitment.^[23] What makes such a neutrality policy rationally credible is the fact the collective defence is of lesser importance as an institutional function than it was during the Cold War. As CFSP chief Javier Solana notes: “The challenges we face call for a variety of responses with an increasingly diverse use of resources. In general, they are less dependent on traditional military intervention. The question of collective defence is less relevant to the majority of crisis situations which confront us these days”^[24].

1.2.2. Changed Neutral State Preferences for ESDP functions

The new neutrality policies of the EU neutrals are reflected, I argue, in their preferences for the functions ESDP adopts. The functions, or “functional scope”^[25] of an institution, refer to the specific policy areas an institution covers. Thus the functions ascribed to ESDP indicate the scope of European integration in the security and defence field. Neutral state preferences for ESDP functions are thus preferences regarding policy areas as well as substantive policy issues.^[26]

Haftendorn et al. identify four different ‘ideal types’ of functions for security institutions: security management, out-of-area intervention, collective security and collective defence. These are differentiated on the basis of whether they respond to threats or risks ^[27] and whether they institution is inclusive or exclusive with respect to the potential aggressor^[28]. Thus security management involves members of the security institution forming an inclusive coalition that responds to risks. Out-of-area intervention involves cooperation amongst states seeking to address risks emanating from non-members. Collective security involves member states committing themselves to come to the aid of one member if it is attacked or threatened by another member. Finally, the purpose of collective defence is to protect the members of an alliance against threats from non-members.^[29]

The concept of “collective defence” requires further explanation, as it in not entirely unambiguous. In collective defence (also referred to as common or mutual defence) institutions, a military attack on one member state is viewed as an attack on all member states. In turn, member states are required to assist the attacked state by all possible means – military or other. However, collective defence institutions do differ in whether this mutual assistance obligation is automatic or not: In the case of the Western European Union, it is, yet in NATO it is not.^[30] For the purposes of this paper, I will consider both automatic and non-automatic mutual assistance provisions, as detailed above, to be a “common defence”.

Graphed, Carsten Tams’ schema looks as follows:

illustration not visible in this excerpt

Given my hypothesis that neutrality is now defined by the neutral states to mean not being bound into an “alliance” by a mutual defence commitment, I argue that the neutral states preferences for ESDP functions include security management, out-of-area intervention, and collective security (functions 1-3). However, the neutrals are opposed to ESDP adopting a collective defence function (function 4). For purposes of simplicity, I will refer to Tams’ functions 1, 2 and 3 as “security and defence cooperation” and to function 4 as “common defence”.^[31] Accordingly, I expect the neutral states to agree to ESDP functions and form relating to “security and defence cooperation”, but to reject proposals concerning a “common defence”.

1.3. … and a change in the cooperation problem.

What sort of cooperation problem are we confronted with as far as the evolution of ESDP is concerned? What is certain is that the neutral states’ accession to the EU, and their activities in the realm of ESDP since then, reflect a willingness to cooperate in creating a European Security and Defence Policy. Accordingly, I believe, the interaction between the EU Neutrals and the remaining EU member states on ESDP is best explained as a game-theoretic coordination game^[32]: “Although cooperation is not thwarted by dominant strategies of defection, it may never be launched, due to disagreements about coordination points”^[33]. All member states desire further development of ESDP, yet problems arise due to the existence of multiple equilibria, i.e. differing preferences for ESDP function. I argue that one major difference in preference for ESDP is reflected in the divide between EU states with and without neutrality policies.

In a coordination game, the key problem is one of choosing between different equilibria: “The central dilemma in this situation is deciding which […equilibrium] will prevail”^[34]. Which focal point will be chosen from these equilibria? In other words, what institutional functions will be agreed upon from the conflicting preferences of the EU member states? I hypothesize that the actual functions adopted by ESDP (as is evident) will reflect the neutral states’ preferences for ESDP.^[35] The neutral states wield such bargaining power because ESDP decisions require unanimity^[36] in the European Council (Article 23(2) TEU).^[37] As a result, we may also expect the neutral states to resist any move towards qualified majority voting on ESDP matters. However, as voting procedures are a matter of form rather than function, I will only consider this preference when relevant to function, and not in any further detail within the framework of this study.

Accordingly, rationalist institutionalist theory leads us to believe that the neutral states have significant impact in the shaping of European Security and Defence Policy. As Wallace and Wallace note: “Questions of sovereignty and of distinctive national interests have remained important; cooperation has edged forward on the basis of consensus”^[38]. Thus I question the validity of the (neo-)realist prediction, that the small, neutral states would be forced to surrender their longstanding neutrality policies to the martial preferences of ‘EU hegemons’ (Germany, France?), who would be primarily responsible for shaping ESDP. It is the purpose of my paper to substantiate the claim I make to the contrary.

My reasoning so far can be summarised as follows:

To the neutral EU states in the post-Cold War environment, “neutrality” means not being bound into an alliance by a common defence commitment.

Accordingly, neutral states preferences for ESDP are for functions relating to “security and defence cooperation”, but against a “common defence”.

The neutral states’ preferences will be reflected in ESDP, because ESDP decisions require unanimity in the European Council.

[...]

^[1] It is necessary to note that this is, of course, a simplification. The functions of ESDP can, more accurately, be said to be a product of two major sets of preferences: Firstly, there are the neutrality policies of the EU neutrals. Secondly, there are the Alliance commitments of those EU states who are also members of NATO. However, my focus here is to investigate the impact of the former - neutrality. On occasion, the latter set of preferences becomes relevant to the former, at which point I will also turn to the influence of NATO commitments on shaping ESDP.

^[2] Hall/Taylor (1997): 944-945.

^[3] König-Archibugi (2004): 142.

^[4] Haftendorn et al. (1997): 6.

^[5] Naumann (2002): 32 (my translation).

^[6] Ibid.: 26.

^[7] Ibid.: 27.

^[8] White Paper on Defence (2000): 13.

^[9] Connolly (1992): 5.

^[10] Portela-Sais: 1.

^[11] Väyrynen (1987)

^[12] Binter (1991): 113-126.

^[13] This is, of course, simultaneously a disadvantage of neutrality: Neutral states must provide for their own security, and are thus not able to ‘free-ride’ on the security and defence provisions of alliance members.

^[14] Ishizuka (1999).

^[15] For evidence on Irish commitment to neutrality see, for example, the Irish Government White Paper on Foreign Policy (1996) and ‘Ireland and the European Union’ (2003) .

^[16] Daniel Keohane (2001): 14-16.

^[17] The opposition Fine Gael party questioned the usefulness of neutrality in a 2003 document: Fine Gael, Beyond Neutrality: Security, Social Justice and Responsibility.

^[18] Sloan.

^[19] This was, for example, the position espoused by the Irish Socialist Party as well as by the Peace and Neutrality Alliance (PANA).

^[20] Portela-Sais: 1

^[21] Ibid.: 1-2.

^[22] quoted in Kirk (2002): 1.

^[23] My purpose here is not to make a normative judgement as to whether this is a too-devalued definition. I refer those with an interest in this question to the advocacy group Action from Ireland (AFRI), which has argued that non-participation in a mutual defence pact is too narrow a definition of neutrality. See: A. Storey. 2001. The Treaty of Nice, NATO and a European Army: Implications for Ireland. Afri Position Paper No.3. www.afri.buz.org

^[24] Solana (2003): 109.

^[25] Lindberg/Scheingold (1970): 65.

^[26] The question may arise as to why I chose to consider how the preferences of states shape institutional function, over studying a form-follows-function relationship. My reasons for rejecting a function-form study have to do with the difficulty in observing how function causes form. Institutional form (Lindberg/Scheingold’s equivalent term is ‘institutional capacity’) are the procedures and processes by which decisions are taken, and which are necessary for states to cooperate despite the presence of partially conflicting interests. Though I study aspects of ESDP form (such as unanimous decision making), determining how and whether function impacts ESDP form, and what role the neutrals play in this, is, I believe, methodologically very difficult. Moreover, I wish to avoid coming up with the indeterminate, as frequently circular, research results that are a frequent problem of function-form analysis. Cf Hafendorn et al. (1999): 329. On the function-form issue when studying ESDP, see also König-Archibugi (2004): fn. 9.

^[27] “Threats pertain when there are [identifiable] actors that have the capabilities to harm the security of others and that are perceived by their potential targets as having intentions to do so. When no such threat exists, either because states do no have the intention or the capability to harm the security of others, states may nevertheless face a security risk.” Haftendorn et al. (1999): 25. See also Tams (1999): 82-85.

^[28] Inclusive security institutions are designed to deal with threats among members; exclusive security institutions deter and defend against external threats. cf. Hafendorn/Keohane/Keohane (1999): 26.

^[29] Tams (1999): 81-84.

^[30] Woyke (2000): 278-282, 317.

^[31] These terms draw on EU terminology, which distinguishes between ‘common defence policy’ and a ‘common defence’. ‘Common defence policy’ is the equivalent of my term ‘security and defence policy’; the EU term ‘common defence’ denotes a mutual defence commitment (a la NATO or WEU). Cf. Institute of European Affairs (2001): 2.

^[32] The neutral’s concern is not defection on the part of other states (which rules out a collaboration problem), nor is it to exploit partners’ with a dominant strategy (suasion problem) or incomplete information about another state’s references (assurance problem).

^[33] Haftendorn et al. (1999): 7.

^[34] Martin (19992): 775.

^[35] In bargaining terminology, this can be stated as follows: ESDP measures relating to the security and defence cooperation fall within the neutral states’ “zone of agreement”. However, collective defence does not, thus its exclusion from ESDP policy forms the EU neutrals’ “reservation price”.

^[36] Unanimity does not include Denmark, which has a “blanket opt-out” on ESDP and thus doesn’t participate in or vote on ESDP matters in the European Council. For a discussion of the importance of unanimity/intergovernmentalism in CFSP/ESDP in the European Council see Müller-Brandeck-Bocquet (2002): 261ff.

^[37] In order to ensure that their constellation of preferences is not compromised, one might also expect the neutral states to resist encroachments on their national sovereignty on this matter: Accordingly, a neutral state preference for institutional form is unanimous Council decision making on ESDP – and a rejection qualified majority voting (QMV). However, in this paper I focus on the institutional functions (“functional scope”) of the EU, rather than its form (“institutional capacity”). For a study of how government preferences for ESDP form vary, see König-Archibugi (2005).

^[38] Wallace/Wallace (2000): 489.