‘should i stay or should i go ?’—a critical analysis of the right to withdraw from the eu

eulj_524 589..603

‘Should I Stay or Should I Go ?’—A Critical Analysis of the Right to

Withdraw from the EU

Hannes Hofmeister*

Abstract: This article seeks to answer one of the key questions facing the EU in thefuture: what effect will the new right to withdraw have on the EU? Will it lead to a gradualfragmentation of what was supposed to be ‘an ever closer union of unlimited duration’? Orwill it even mark the beginning of the end of the Union? In order to answer these complexquestions, this article first briefly analyses the pre-Lisbon situation regarding withdrawal.It then critically examines the newly inserted Article 50, which codifies the right towithdraw. Having done so, it will then examine whether non-legal considerations, such aspolitical and economic reasons, will render withdrawal a theoretical rather than realisticoption.

I Introduction

Three years after the adoption of the Constitution for Europe, and two years after itsrejection by Dutch and French voters,1 the continent’s leaders managed the impos-sible: they saved ‘large parts of the old text and stitched them together into a newtreaty’,2 the so-called Lisbon Treaty. In essence, the Lisbon Treaty will bring aboutmajor institutional changes, such as the creation of the post of President of the Euro-pean Council or High Representative for Foreign and Security Affairs. Yet it alsocontains another key amendment which has hardly received any media attentiondespite its potential to shake the very foundations of the EU: the unilateral right towithdraw from the EU.

For the first time in its history, EU law now provides for an express right to withdrawfrom the Union. EU membership is thus no longer a ‘marriage for life’, but rather aso-called ‘Lebensabschnittspartnerschaft’.3

* Research Fellow, Lauterpacht Centre for International Law, University of Cambridge. The author wouldlike to thank Professor Marise Cremona, European University Institute, Florence and the anonymousreferees for their helpful comments.

1 In popular referenda held in May/June 2005; for details see G. Amato and J. Ziller, The EuropeanConstitution (Edward Elgar, 2007), at 56–60.

2 ‘Tailoring a Compromise’, The Economist, 24 June 2007.3 ie ‘A partnership for a certain phase in one’s life’.

European Law Journal, Vol. 16, No. 5, September 2010, pp. 589–603.© 2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford, OX4 2DQ, UKand 350 Main Street, Malden, MA 02148, USA

What effect will this new right have on the EU? Will it lead to a gradual fragmen-tation of what was supposed to be an ever-closer union of unlimited duration? Or willit even mark the beginning of the end of the Union?

In order to answer these complex questions, this article proceeds as follows. It firstbriefly analyses the pre-Lisbon Treaty situation regarding withdrawal. It will thencritically examine the newly inserted Article 50, which codifies the right to withdraw.Having done so, it will then examine whether non-legal considerations, such as politicaland economic reasons, will render withdrawal unlikely.

II The Right to Withdraw—The Pre-Lisbon Situation

Until recently, the existence of a unilateral right to withdraw has been highly contro-versial. Various attempts have been made to derive such a right both from the EC/EUTreaties themselves and from the general principles of international law. Both sourcesof law will therefore be briefly analysed.

A The EC/EU Treaty

Neither the EC Treaty (EC) nor the EU Treaty (TEU) contained an express provi-sion allowing for unilateral withdrawal from the Union. Moreover, attempts todeduce an implicit right to withdraw were difficult to reconcile with the wording ofArticle 312 EC,4 which provides that ‘[t]his treaty is concluded for an unlimitedperiod’. Article 1(2) of the EU Treaty,5 which referred expressly to ‘an ever closerunion’ made it even more complicated to argue for such an implicit right.6 Systematicconsiderations also militated against the existence of such a right: Since Article 49TEU made accession to the Union contingent upon the consent of all MemberStates, it would have been odd if withdrawal—as the actus contrarius—could havebeen declared unilaterally.7

In short, neither the EC Treaty nor the EU Treaty provided for an express or implicitright to unilateral withdrawal.

B International Law

It might, nevertheless, be possible to derive such a right from the general principles ofinternational law. There is no doubt that the Founding Treaties were originally bornout of international law, even if they had some unique features from their beginning. Itis, however, questionable whether these general principles of international law stillapply to the Community legal order.

4 As well as the identical Article 51 TEU.5 As well as the Preamble to the EU Treaty.6 H. Herrnfeld, in: J. Schwarze (ed.), EU Kommentar (Nomos, 2000), Art 51 EUV, 1.7 See K. Hailbronner et al, Handkommentar zum EUV/EGV (Heymanns, 1998), Art Q EUV, 2. Though this

will be the case under the Lisbon Treaty.

European Law Journal Volume 16

590© 2010 Blackwell Publishing Ltd.

According to some scholars,8 the EC Treaty has created a new and autonomouslegal order9 and therefore the applicability of general rules of international law isexcluded.10

Others, however, argue that the EC/EU Treaties are still part and parcel of interna-tional law. In their view, the notion of an autonomous community legal order is anillusion. Yet, even proponents of this view restrict the applicability of general rules ofinternational law to cases in which there is no specific Community provision dealingwith the subject matter. The existence of such a specific Community provision wouldrender general rules of international law inapplicable because of the principle of ‘lexspecialis derogat legi generali’. However, since Community law did not contain anyprovision dealing with treaty termination11 and withdrawal, recourse to the generalrules of international law was possible. Proponents of this theory therefore had littledifficulty in applying the principles of international law—in particular the ViennaConvention on the Law of Treaties (VCLT)12—to the EU.13 Subject to the conditionslaid down in Articles 56, 60 and 62 of the VCLT, a state may thus withdraw from theUnion. These Articles, however, allow for withdrawal only in very exceptional circum-stances, thus making withdrawal a theoretical rather than a real option.14

C Conclusion

In short, in the pre-Lisbon era, the situation regarding unilateral withdrawal wasunclear. Although the EC Treaty and the EU Treaty did not provide for an express or

8 See, eg, M. Hilf, in H. von der Groeben, J. Thiesing and C. Ehlermann (eds), Kommentar zum EU/EG-Vertrag (Nomos, 1997), Art 240, 9; H. P. Ipsen, Europäisches Gemeinschaftsrecht (Mohr Siebeck, 1972), at101.

9 See also the ECJ’s ruling in Costa v ENEL: ‘By contrast with ordinary international treaties, the EEC hascreated its own legal system which on the entry into force of the Treaty became an integral part of the legalsystems of the Member States and which their courts are bound to apply. By creating a community ofunlimited duration, having its own institutions, its own personality, its own legal capacity and capacity ofrepresentation on the international plane and, more particularly, real powers stemming from a limitationof sovereignty or a transfer of powers from the states to the Community, the Member States have limitedtheir sovereign rights, albeit within limited fields, and have thus created a body of law which binds boththeir nationals and themselves’; Case 6/64, Costa v ENEL, Slg 1964, 1251, at 1259.

10 See Hilf, op cit n 8 supra and Ipsen, op cit n 8 supra.11 Until the entry into force of the Lisbon Treaty on 1 December 2009.12 The application of the VCLT to the EC is problematic for the following reasons. First, not all Member

States of the EC have ratified the VCLT. Moreover, due to the non-retroactivity of the VCLT (see Art 4),the EC Treaty does not fall under the direct application of the VCLT. However, in view of the legal natureof these codified rules as part of customary international law, all members of the EC are bound indirectlyby provisions of the VCLT; M. Geistlinger, Some Ideas on the Possibility of Unilateral Withdrawal from theEuratom Treaty, Conference Paper (Copenhagen, 23 September 2005), at 2.

13 Against this, some argue that even in the absence of a specific Community provision, rules of generalinternational law are inapplicable if their application would have so-called ‘disintegrating effects’. Sincesuch effects cannot be excluded when deriving a right to withdraw from general rules of international law,recourse to these rules is not possible; see, eg, J. Zeh, ‘Recht auf Austritt’, (2004) 7 Zeitschrift fürEuropa—rechtliche Studien 182.

14 Yet even if one accepts the applicability of international law at all—in the given case the applicability ofArts 54–62 VCLT—it would be difficult to construe a right to withdraw unilaterally based on theseArticles. For instance, Art 62 of the VCLT which regulates withdrawal in case a fundamental change ofcircumstances occurs (so-called ‘clausula rebus sic stantibus’) is of very limited relevance in the context ofthe EU, because Art 48 TEU—as lex specialis—already allows for Treaty modifications.

September 2010 Right to Withdraw from the EU


implicit right to unilateral withdrawal, some scholars nevertheless derived such a right15

from the general rules of international law, in particular from Articles 56–62 of theVCLT.

The Lisbon Treaty has finally put an end to this debate by inserting Article 50 intothe revised EU Treaty. This article provides for an unfettered right to unilateral with-drawal, to which I will now turn.

III The New Right to Withdraw under Article 50

A Article 50—An Outline

The right to withdraw from the EU is codified in Article 50 of the revised EU Treaty.The specific wording of the provision is as follows:

(1). Any Member State may decide to withdraw from the European Union in accordance with its ownconstitutional provisions.

(2). A Member State which decides to withdraw shall notify the European Council of its intentions; theEuropean Council shall examine that notification. In light of the guidelines provided by the EuropeanCouncil, the Union shall negotiate and conclude an agreement with that State setting out the arrange-ments of withdrawal taking account of the framework of its future relationship with the Union. Thatagreement shall be negotiated in accordance with Article 218(3) of the Treaty on the Functioning of theEuropean Union. It shall be concluded on behalf of the Union by the Council of Ministers, acting by aqualified majority, after obtaining the consent of the European Parliament.

(3). The Treaties shall cease to apply to the State in question from the date of entry into force of thewithdrawal agreement, or failing that, two years after the notification referred to in paragraph 2, unlessthe European Council, in agreement with the Member State concerned, decides to extend this period.

(4). For the purposes of paragraphs 2 and 3, the member of the European Council or of the Councilrepresenting the withdrawing Member State shall not participate in the European Council or theCouncil discussions or decisions concerning it. A qualified majority shall be defined in accordance withArticle 238(3)(b) of the Treaty on the Functioning of the European Union.

(5). If a State which has withdrawn from the Union asks to re-join, its request shall be subject to theprocedure referred to in Article 49.

This complex provision will be analysed in more detail in the following section. Inparticular, I will examine (a) the ‘conditions’ for withdrawal; (b) the procedural aspects;and last, but not least, (c) the legal effects of withdrawal.

a) Conditions for WithdrawalAccording to Article 50(1), any Member State may decide to withdraw from the EU inaccordance with its own constitutional provisions. EU law itself does not impose anyspecific substantive conditions on a Member State’s right to withdraw.16 The underly-ing rationale of this rule is to reassure potential Member States that they will remain the‘masters of the treaties’, able to withdraw from the EU at any time subject only to theirown constitutional provisions.17

15 Albeit a heavily restricted right. In contrast, Art 50 of the revised EU Treaty provides for an unfettered,unilateral right to withdraw.

16 See CIG 4/2003, Regierungskonferenz 2003—Redaktionelle und juristische Anmerkungen zu dem Entwurfeines Vertrages über eine Verfassung für Europa- Basisdokument, at 126.

17 See, eg, W. Heintschel von Heinegg, in C. Vedder and W. Heinschtel von Heinegg (eds), EuropäischerVerfassungsvertrag, Handkommentar (Nomos, 2007), at 256.



b) ProcedureA Member State that decides to withdraw must, first, formally notify the EuropeanCouncil of its intentions.18 On the basis of the guidelines provided by the EuropeanCouncil, the Union will then negotiate an agreement, specifying the details of with-drawal and setting out the contours of the future relationship between the Union andthe withdrawing state. The agreement thus only comprises the details of the with-drawal, whereas the future relationship between the Union and the withdrawing stateis only cursorily addressed in the agreement. Practically speaking, this will necessitatethe conclusion of a further agreement at a later point in time which also regulates indetail the future relationship between the two parties.

The agreement will be concluded on behalf of the Union by the Council of Ministers,acting by a qualified majority, after obtaining the consent of the European Parliament.Interestingly, the withdrawal agreement is not designed as an ‘actus contrarius’19 to theaccession agreement. As such, it would have to be concluded by the Member States—see Article 49 TEU. Parties to the withdrawal agreement are, however, the withdrawingstate and the Union itself—see Article 50(2). This makes sense insofar as the with-drawal agreement only concerns the relationship between the Union and the withdraw-ing state.20

c) Legal EffectsAccording to Article 50(3), EU Law shall cease to apply to the state in question fromthe date of entry into force of the withdrawal agreement or, failing that, two years afterthe notification of the EU Council. A close analysis of Article 50(3) thus reveals that thewithdrawal agreement referred to above is not an essential condition for withdrawalfrom the EU. Even if the EU and the withdrawing Member State fail to reach anagreement, withdrawal nevertheless becomes effective two years after notification.

In short, Article 50 represents a compromise solution: although Member States havea ‘unilateral right to withdraw, they do not have an immediate right to do so’.21 Beforewithdrawal eventually becomes effective, either the two-year period must have expiredor a withdrawal agreement must have been successfully concluded.

While the introduction of a right to withdraw from the Union has been welcomed bymany politicians, the next section will show that Article 50 is not a masterpiece of legaldraftsmanship and might lead to significant problems.

B Critical Analysis

a) Lack of Detailed Procedural RegulationsThe fairly brief procedural rules on withdrawal have already been outlined above.Unfortunately, they fail to address many key issues that would arise on withdrawal.For instance: although the representatives of the withdrawing Member State on theCouncil are expressly prohibited from taking part in the vote or in discussions regard-ing the putative withdrawal, Article 50 remains silent with respect to the role of

18 See Art 50(2) TEU.19 ie as a ‘contrary act’ to the accession agreement.20 See, eg, Heintschel von Heinegg, op cit n 17 supra, at 256–257.21 R. Friel, ‘Providing a Constitutional Framework for Withdrawal from the EU’, (2004) 53 International

and Comparative Law Quarterly 425.



parliamentarians from that Member State in securing the consent of Parliament. Canthese MEPs vote? Or be heard in the discussions?22 Nor do the procedural rules addressthe important question of whether the withdrawal notification itself can be withdrawnduring the two-year period.

The fact that Article 50 does not define the new status of the withdrawing MemberState23 during the negotiation period also raises a number of interesting questions. Forinstance, can the withdrawing Member State still take part in the general24 decision-making process, thereby influencing decisions of an institution which it is about toleave? If that were actually the case, then the withdrawing state would be in a powerfulposition: it could delay or even block decisions that would benefit the remainingmembers of the EU but would be detrimental to its own future status.

Furthermore, what happens to ongoing projects, involving the withdrawing state?And what about the rights of EU citizens residing in the withdrawing state: forinstance, do they retain the right to vote in local elections?25 Article 50 does notaddress any of these essential questions. Of course, it may be argued that most ofthese points can be dealt with in the withdrawal agreement. Yet, such an agreementis not obligatory.26 Even if the parties fail to reach an agreement, withdrawal willnevertheless become effective two years after notification of the Council.27 Hence,there is the risk that the parties may eventually find themselves in a situation ofeffective withdrawal but without an agreement regulating the key aspects outlinedabove. Unfortunately, this risk is rather high since any agreement, dealing with anissue as complex as withdrawal, would require a sense of cooperation among theparties. Yet, such cooperation may be difficult to achieve without an existing legalframework, setting out the procedure for withdrawal in detail. This is particularly truefor tense and conflict-laden situations such as withdrawal. In such circumstances, itmay be very difficult for the conflicting parties to agree even on basic (procedural)issues.28 Hence, a detailed procedural legal framework is essential to avoid the risk ofdeadlock.29

The procedural rules on withdrawal are deficient in many ways. Not only do theylack adequate mechanisms to regulate the relationship between the EU and the with-drawing Member State as explained above; they also fail to address adequately proce-dural issues arising on an internal Union level. For instance, the withdrawal ofany Member State would change the composition as well as the operation of EUinstitutions considerably. ‘Swift and rapid treaty amendment would be required toaccommodate these changes. Any Article on withdrawal should therefore containsuch expedited provisions in order to enable the EU to accommodate quickly to the

22 ibid, at 426.23 And of its representatives.24 ie the ones that do not relate to the withdrawal strictly speaking.25 Another interesting question left unanswered by Art 50 is: what happens to cases pending before the

Community courts involving the withdrawing state?26 See, eg, Heintschel von Heinegg, op cit n 17 supra, at 256–257.27 See Art 50(3) of the revised EU Treaty.28 See, eg, the long and complex negotiations that preceded the withdrawal of Greenland. For further details

on the withdrawal of Greenland, see W. Ungerer, ‘Der Austritt Grönlands aus der Europäischen Gemein-schaft’ (1984) 39 Europaarchiv 345.

29 Which may ultimately distract the EU from its main functions.



profound changes’30 caused by a state’s withdrawal. Article 50, however, is also silenton this aspect.31

In sum, the procedural provisions lack clarity and—perhaps even worse—are incom-plete. The drafters of Article 50 could have prevented much confusion by taking intoaccount the above-mentioned aspects and by formulating a clear and comprehensiveprovision.32 Regrettably, the contrary is the case. Article 50 is remarkably deficient inthis respect.

b) Withdrawal and the ‘Nature’ of the EUIt can also be argued that the introduction of a unilateral right to withdraw, asenvisaged by Article 50, is incompatible with the nature of the Union. In its famousjurisdiction clarifying the supranational characteristics of the EC legal order, theEuropean Court of Justice (ECJ) has consistently emphasised two aspects: (a) theunlimited duration of the founding treaties and (b) the irreversible nature of the inte-gration process. For instance, in Costa v ENEL,33 the court held:

By contrast with ordinary international treaties, the EEC has created its own legal system which . . . be-came an integral part of the legal systems of the Member States and which their courts are bound toapply. By creating a community of unlimited duration, having its own institutions, its own personality,its own legal capacity and capacity of representation on the international plane and, more particularly,real powers stemming from a limitation of sovereignty or a transfer of powers from the states to theCommunity, the Member States have limited their sovereign rights, albeit within limited fields, and havethus created a body of law which binds both their nationals and themselves. It follows that . . . the lawstemming from the Treaty, an independent source of law, could not, because of its special and originalnature, be over-ridden by domestic legal provisions, however framed, without being deprived of itscharacter as Community law and without the legal basis of the Community itself being called intoquestion. The transfer by the States from their domestic legal system to the Community legal system ofthe rights and obligations arising under the Treaty carries with it a permanent limitation of their sovereignrights, against which a subsequent unilateral act incompatible with the concept of the Community cannotprevail.34

In Commission v France,35 the ECJ upheld this approach, highlighting once more the‘unlimited duration’ of the Community: ‘The Member States agreed to establish aCommunity of unlimited duration, having permanent institutions vested with realpowers, stemming from a limitation of authority or a transfer of powers from the Statesto that Community. Power thus conferred could not, therefore, be withdrawn from theCommunity . . . ’.36 This jurisprudence is in line with Articles 53 TEU and 356 TFEU,37

which both provide that ‘this treaty is concluded for an unlimited period’. The intro-duction of a unilateral right to withdraw, however, seems to be irreconcilable with theidea of an ever-closer Union of unlimited duration. In particular, it appears to castdoubt on the irreversible nature of the integration process.

30 Friel, op cit n 21 supra, at 426.31 It therefore presumably relies upon the ordinary treaty amendment provision, Art 48 TEU, which seems

ill-suited to the specific needs which arise on withdrawal: ibid, at 426.32 ibid, at 426.33 Costa v ENEL, n 9 supra.34 ibid, at 1269–1271.35 Case 7/71, Commission v France, Slg 1971, 1003.36 ibid, at 1018.37 Formerly Arts 51 TEU and 312 EC.



This line of argument is not convincing, though, for two reasons. First, both thejurisprudence of the ECJ and the wording of Article 53 TEU38 cannot be interpreted insuch a way as to require the permanence of EU membership and thereby exclude thewithdrawal of a Member State. Second, it must not be forgotten that the MemberStates remain the sovereign ‘masters of the treaties’.39 As such, they have the power toalter any existing provision of the treaties or to introduce new ones. I will deal with botharguments in turn.

First, both the jurisprudence of the ECJ and the wording of Article 53 TEU40

cannot be interpreted in such way as to require the permanence of EU membershipand thereby exclude the withdrawal of individual Member States. True, Article 53TEU41 provides that ‘this treaty is concluded for an unlimited period’. Given theperpetual tenor of this phrasing one might be tempted to argue that if the treaty isconcluded for an unlimited period, then it must be permanent; and if it is permanent,then there must be no right to withdraw. But is this really the correct way to interpretthis enigmatic Article?

Interesting insights can be gained, for example, by comparing Article 53 TEU withArticle 1 of the abortive European Political Community (EPC),42 on whose ashes theFounding Treaties were eventually drafted.43 Article 1 EPC employed a strongerterm: the EPC ‘was to be indissoluble. Arguing a-contrario it could be said that allArticle 53 TEU44 intends to convey is that the duration of the EU is not to be limitedin advance (unlike the ECSC which was so limited) and not that it necessarily bepermanent’.45

Moreover, even if one (hypothetically) assumes for the moment that the EU Treatyis permanent, this does not necessarily mean that individual membership has to bepermanent, too. One has to distinguish clearly between the permanence of the treatyitself and the permanence of individual membership. An organisation itself can bepermanent despite a change in the composition of its Member States. Hence, with-drawal of an individual Member State would still be possible even if the EU Treatywere to be permanent.46

In short, both the jurisprudence of the ECJ and the wording of Article 5347 cannot beinterpreted in such a way as to exclude the withdrawal of individual Member States.

Second, it must not be forgotten that the Member States remain the sovereign‘masters of the treaties’. As such, they have the power to alter any existing provision

38 And the identical Art 356 TFEU, as well as their ‘predecessors’: Arts 51 TEU and 312 EC.39 This phrase was coined by the German Constitutional Court; see, eg, 89 BVerfGE 155, at 190.40 And the identical Art 356 TFEU and their ‘predecessors’: Arts 51 TEU and 312 EC.41 And its predecessors; see n 40 supra.42 Article 1 of the draft EPC provided: ‘The present treaty sets up a European Community of a supranational

character. The community is founded upon a union of peoples and states, upon respect for their person-ality and upon equal rights and duties for all. It shall be indissoluble’. The text of the EPC is available athttp://aei.pitt.edu/991/01/political_union_draft_treaty_1.pdf

43 J. Weiler, ‘Alternatives to Withdrawal from an International Organization: The Case of the EuropeanEconomic Community’, (1985) 20 Israel Law Review 284.

44 And its predecessors; see n 40 supra.45 Weiler, op cit n 43 supra, at 285.46 An organisation itself can be permanent despite a change in the composition of its Member States. The

only problematic—though highly unlikely—scenario would be if all Member States withdrew at the sametime.

47 And the identical Art 356 TFEU and their ‘predecessors’ Arts 51 TEU and 312 EC.



of the treaties or to introduce new ones.48 It is sometimes claimed, though, that theECJ49 has developed a number of fundamental principles that can no longer bealtered.50 Whether such an ‘unalterable core’ of principles actually exists is thoughvery controversial.51 But even assuming that such a core of fundamental principlesactually exists, ‘permanence of individual membership’52 would not be part ofit.53

In conclusion, it can be argued that the introduction of a unilateral right to withdrawis not irreconcilable with the nature of the EU.

c) The Right to Withdraw and the Principle of UniversalityAnother argument, closely related to the foregoing, is derived from the principle ofuniversality. Some international lawyers54 tend to deny Member States the right towithdraw unilaterally from international organisations based on the ‘principle of uni-versality’. Against the backdrop of increasing globalisation and the equality of sover-eign states, the principle of universality requires all states to cooperate in internationalorganisations. For instance, such a duty to cooperate continuously may be derivedfrom the ‘Friendly Relations Declaration’.55 If states are required to cooperate con-tinuously, it is argued, then they may not withdraw. Yet applying these rather vagueconcepts to the specific case of withdrawal from the EU would take matters too far. Forinstance, the Friendly Relations Declarations does not specify in detail the degree ofcooperation required. In particular, it does not refer to a ‘continuous’ cooperation.Moreover, (limited) cooperation may still be possible even if a state withdraws.56

The application of the concept of universality in the given context is therefore notconvincing.

48 Subject to the conditions laid down in Art 48 TEU; but see also T. Hartley, ‘The Constitutional Foun-dations of the European Union’, (2001) 117 The Law Quarterly Review 236, who even argues: ‘Article 48cannot deprive the Member States, acting unanimously, of the power to amend the Treaties withoutcomplying with its requirements’.

49 In particular in the delphic opinion ECJ 1/91, Slg, 1991, I-6079 (71).50 See, eg, J. L. Da Cruz Vilaca and N. Picarra, ‘Y-a-t-il des limites matérielles à la révision des traités

institutant les communautés europeénnes?’, (1993) 29 Cahiers de droit européen 3.51 In this author’s opinion, the better view seems to be that no such ‘unalterable core’ of Community law

exists. ‘There is no evidence that the Member States ever accepted any such substantive limits to theirtreaty amending power’: B. de Witte, ‘Treaty Revision in the European Union: Constitutional Changethrough International Law’, (2004) 35 Netherlands Yearbook of International Law 57.

52 Or to put it differently: the ‘non-existence’ of a right to withdraw.53 See M. Herdegen, Europarecht (Beck, 8th edn, 2006), at 86. The principles of democracy, respect for the

rule of law and human rights as well as the jurisdictional role of the ECJ are considered by some to belongto this unalterable core of fundamental principles; see, eg, M. Heintzen, ‘Hierarchisierungsprozesseinnerhalb des Primärrechts der Europäischen Gemeinschaft’, (1994) 29 Europarecht 35; Da Cruz Vilacaand Picarra, op cit n 50 supra.

54 See, eg, G. Czerwinski, Das Universalitätsprinzip und die Mitgliedschaft in internationalen universalenVerträgen und Organisationen (Duncker & Humblot, 1974), at 40.

55 The relevant passage of the Friendly Relations Declaration reads as follows: ‘States have the duty toco-operate with one another, irrespective of the differences in their political, economic and socialsystems . . . ’: Declaration on Principles of International Law concerning Friendly Relations and Coop-eration among States in accordance with the Charter of the United Nations, GA Res 2625 (1970), availableat http://daccessdds.un.org/doc/RESOLUTION/GEN/NR0/348/90/IMG/NR034890.pdf?OpenElement.

56 eg, such a (limited) form of cooperation would be possible, if the withdrawing state decides to remain amember of the European Economic Area.



d) Article 50 Privileges Large StatesArticle 50 also tends to favour big states in a number of ways. Prima facie, this mayseem perplexing, given that some of the most fervent supporters of this provision weresmall states. However, they seem to have neglected the fact that Article 50 does not laydown the specific details57 to be regulated in the withdrawal agreement,58 in particularthe compensation mechanism59 to be employed once a state withdraws. The Union andthe withdrawing state are thus basically free to include any clause they think fit.However, there is no doubt that large and economically powerful states have morenegotiating power than small states.60 Just imagine France or the UK contemplatingwithdrawal. No one would seriously doubt that their bargaining power is greater thanthat of a small state, say Malta. They can thus significantly influence the content of theagreement to their advantage. This unsatisfactory situation could have beenavoided—at least to a certain extent—if Article 50 had made (a) the withdrawalagreement mandatory and (b) provided for the specific details61 to be regulated in thisagreement.

e) The Wording of Article 50According to Article 50(2), ‘a Member State which decides to withdraw shall notify theEuropean Council of its intentions; the European Council shall examine that notifica-tion. In light of the guidelines provided by the European Council, the Union shallnegotiate and conclude an agreement with that State . . . ’.62 This wording is confusing.It implies that the conclusion of an agreement is a necessary pre-condition for with-drawal. This is, however, not the case. When reading paragraph (2) in conjunction withparagraph (3), it becomes evident that the agreement referred to in paragraph (2) is notan essential condition for withdrawal. The terminology used in paragraph (2)—‘negotiates and concludes’—must rather be seen as attribution of competence to theEU63 and not as a substantive precondition for withdrawal.64

The wording of Article 50(3) is also problematic. It provides that EU law shall ceaseto apply to the state in question from the date the withdrawal agreement enters intoforce.65 If this paragraph were to be narrowly interpreted as ‘prohibiting interimprovisions in a withdrawal implementation agreement, the drafters of such an agree-ment would have a hard time finding practical solutions regarding ongoing legalrelationships, such as cases pending before the Community courts’.66

On a more general note, Article 50—in particular paragraph (1)—is couched in verybroad terms, thus making interpretation essential. Yet, when interpreting this norm theeffect of the principle of sincere cooperation, as enshrined in Article 4(3) TEU, needs tobe taken into account. From this principle there follows a duty to abstain from

57 Nor does it lay down the contours to be dealt with in the agreement.58 To be precise, it does not even require such an agreement at all.59 How much money has to be transferred back to the EU/Members State? How is this to be done? When

must this be done?60 See Friel, op cit n 21 supra, at 427.61 As well as the compensation mechanism to be employed.62 Article 50(2) of the revised TEU.63 And not to the individual Member States.64 See, eg, Heintschel von Heinegg, op cit n 17 supra, at 256–257.65 See Art 50(3) of the revised TEU; see also J. Herbst, ‘Observations on the Right to Withdraw from the

European Union: Who are the “Masters of the Treaties”?’, (2005) 6 German Law Journal 1757.66 ibid, at 1757. Though it should not be forgotten that the two-year period can be extended; see Art 50(3).



measures that could jeopardise the attainment of the objectives of the EU. Henceproblems might arise with regard to the question of how to balance the right towithdraw and the principle of sincere cooperation when interpreting this vague provi-sion. Regrettably, Article 50 does not provide any guidance on this important aspect.

f) Undesirable Consequences: A Europe à la carte?A unilateral right to withdraw as envisaged by Article 50 could lead to some sort ofregressive, gradual disintegration: For example, if a state withdraws from the EU, butnevertheless still participates—by way of association—in some EU politics, this mayeventually lead to a situation of partial membership. This might result in a sort ofcherry picking approach (‘Europe à la carte’), whereby the withdrawing state picks andchooses its rights and duties.67 This could not only create tension between the differentstates68 but also a lack of transparency, since each withdrawing state would negotiate itsown catalogue of rights and obligations.

C Conclusion

Article 50 is not only confusing but also deficient in many respects.69 In particular, theprocedural provisions lack clarity and are incomplete. ‘During the difficult period thatwould surround any potential withdrawal, one would expect a rule that would provideclear and unambiguous answers.’70 Regrettably, the contrary is the case. Article 50 isremarkably deficient in this respect. Moreover, an unfettered right to withdraw—asenvisaged by Article 50—is open to abuse, in particular by large and economicallypowerful states.

What is needed instead is a provision that subjects withdrawal to clear and stringentconditions. For instance, withdrawal should only be possible under exceptional cir-cumstances which would need to be clearly defined.71 This would minimise the risk ofabuse: states would thus no longer be able to (ab)use the right to withdraw as a

67 This fear was also expressed by Brok et al on behalf of the EEP Convention Group: A right of a memberstate to pick and choose the rights and duties stemming from union membership (single market yes,rest, no) must be explicitly excluded, available at www.european-convention.eu.int/amendments.asp?content=46&lang=EN.

68 ie between the withdrawing and the remaining states.69 Some even claim that Art 50 would make the ECJ the final arbiter over national constitutional law. Their

argumentation is as follows. According to Art 50, a state’s ‘right to withdraw is subject to the requirementthat it be in accordance with its own constitutional provisions. Since Article 50 would be justiciable by theECJ, this insertion—it is claimed—has catapulted that court into the role of final arbiter over a significantissue of national constitutional law—something not previously attempted within Treaty provisions.Hence, if a dispute arises regarding the validity of the national decision to withdraw, the question as towhether or not that decision was constitutionally made would fall ultimately to the ECJ’: Friel, op cit n 21supra, at 425. This criticism is not convincing though. There is no risk of the ECJ becoming a final arbiterover national constitutional law. Article 50 only requires the ECJ to verify whether the appropriatenational institution has acted, ie made the decision to withdraw. Any further substantive legal examinationwould not be undertaken by the ECJ, since this is exclusively a matter of national constitutional law; see,eg, J. Wichard, in C. Calliess and M. Ruffert (eds), Verfassung der Europäischen Union (Beck, 2006), at637.

70 Friel, op cit n 21 supra, at 426.71 Anne van Lancker (MEP) during the discussions of the Convention meeting of 25 April 2003, Democratic

Life and Membership of the Union, available at http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+PRESS+BI-20030429-1+0+DOC+XML+V0//EN&language=EN#top. For instance, in caseof a fundamental change of circumstances as specified by the concept of ‘clausula rebus sic stantibus’.



bargaining tool, since they could no longer threaten to leave whenever it suited them.72

Moreover, withdrawal should be made contingent upon the successful conclusion of adetailed withdrawal agreement. This would guarantee an orderly process of with-drawal. Such an agreement would also specify the legal obligations of the withdrawingstate vis-à-vis the EU, thereby avoiding the risk that parties may find themselves in alegal vacuum. From a dogmatic perspective, these restrictions could be derived from abroad interpretation of the principle of sincere cooperation as enshrined in Article 4(3)TEU.

IV Will States Make Use of this New Right? The Political and EconomicAspects of Withdrawal

Although the focus of this article has been on the legal implications of the new right towithdraw, the following section also briefly considers economic and political aspects.The main objective of this section is to show that while withdrawal may now be legallypossible, states are unlikely to make use of this new right for economic and politicalreasons.

A Economic Aspects

Economists have for a long time argued in favour of a free entry option, as well as acorresponding free exit option, ie an unfettered, unilateral right to withdraw.73 Eco-nomic integration areas, it is claimed, cannot be definitive but need to be flexible: forinstance, both market developments, as well as other economic changes, might neces-sitate treaty modifications, including changes in the composition of the parties to it.74

From an economic perspective, the underlying rationale of a treaty, such as the EUTreaty, is to make all parties better off.75 Ideally, they should create so-called ‘win-winsituations’. However, if this objective can no longer be attained, then the respectivestates must not be denied the right to withdraw.76

Yet, this argument not only completely disregards the political dimension of Euro-pean integration, but is also unconvincing for other reasons. First of all, this theoryneglects the costs of withdrawal. The withdrawing state would incur huge administra-tive and economic costs in the process of reversing the impact of Europeanisation andit would, for instance, have to reinstate tariffs and customs duties. The situationbecomes even more complex and costly if the withdrawing state is a member of theEuro zone, since it would have to reinstate a national currency. Moreover, withdrawalmight also lead to a tax increase by the withdrawing state in order to substitute forformer financial contributions by the EU.77 Transition costs might further includeso-called ‘uncertainty costs’ which arise as economic actors are forced to lower their

72 But only in clearly defined circumstances as mentioned above.73 See, eg, W. Schäfer, ‘Withdrawal Legitimized? On the Proposal by the Constitutional Convention for the

Right of Secession from the EU’, (2003) 38 Intereconomics 182.74 See J. Buchanan, ‘Europe’s Constitutional Opportunity’, in F. Vibert (ed.), Europe’s Constitutional Future

(Institute of Economic Affairs, 1990), at 6 et seq.75 See R. Vaubel, in G. Kreis (ed.), Der Beitrag der Wissenschaften zur künftigen Verfassung der EU’, (2003)

66 Baseler Schriften zur europäischen Integration 48.76 See Schäfer, op cit n 73 supra, at 182–186.77 See S. Lechner, A Right of Withdrawal in the Constitution?, Conference Paper, IMAD 2nd International

Conference 2005—Institutions and Policies for the New Europe (Siena, Italy, 27–29 October 2005), at 4.



expectations regarding future conditions and to consider risks that may arise out ofcomplex political changes.78

In addition, citizens of the withdrawing state would lose access to the single market.79

They would thus no longer benefit from the four fundamental freedoms, ie free move-ment of goods, persons, services and capital. If the withdrawing Member State wantsto avoid this and negotiates a withdrawal agreement along the lines of the EuropeanEconomic Area Agreement, it will face another problem: it will have to comply withmany EU law obligations,80 while at the same time being excluded from the decision-making process at EU level. In this scenario, the withdrawing state would in manyrespects ‘remain a member of the EU but without any influence over the decisions madeat the European level’.81 In other words, the withdrawing state would run into ‘theproblem of regulation without representation’.82 Whether this is really the intention ofthe withdrawing state is more than questionable.

The withdrawing state might also incur the wrath of the remaining Member Statesfor the following politico-economic reasons: the larger the number of states whichparticipate in the internal market, the greater is its utility because of ‘cost reductions,realization of economies of scale and easier market access’.83 The withdrawing state willtherefore be blamed for reducing these gains. It risks becoming a pariah state. Thismight, in turn, seriously affect its decision on whether or not to withdraw.

In short, the degree of economic integration is far too advanced to allow for a swiftand cheap disentanglement.

B Political Aspects

The political costs for the withdrawing state are manifold as well: for example, it wouldlose influence over the future direction of the EU, while at the same time remainingdependent on it.84 Moreover, the withdrawal of a Member State—particularly of alarge and economically powerful state—might significantly reduce the relevance of theEuropean project and, consequently, the Union’s political clout in the global arena.85 Inthis way, the withdrawing state is likely to gamble away the goodwill of the remainingMember States, and this, in turn, might adversely influence later negotiations onmatters unrelated to the EU.86

Last but not least, reversing the impact of ‘Europeanisation’ will prove difficult.Many scholars have pointed to the impact of EU membership on the policies and polityof a state: ‘Europeanisation encompasses the penetration of European rules, directivesand norms into otherwise differentiated domestic spheres’.87 As a consequence of these

78 ibid, at 4.79 Depending, of course, on the specific content of the withdrawal agreement.80 Since it will continue to be bound by the acquis. It will also remain heavily dependent on the remaining

Member States in many other respects.81 S. Berglund, ‘Prison or Voluntary Cooperation? The Possibility of Withdrawal from the European Union’,

(2006) 29 Scandinavian Political Studies 162.82 N. Pain and G. Young, ‘The Macroeconomic Impact of UK Withdrawal from the EU’, (2004) 21

Economic Modelling 398.83 Lechner, op cit n 77 supra, at 6.84 At least to a certain extent, depending of course on the content of the withdrawal agreement; see Berglund,

op cit n 81 supra, at 162.85 ibid, at 162.86 ibid, at 162.87 P. Mair, ‘The Europeanization Dimension’, (2004) 11 Journal of European Public Policy 341.



interferences with ‘the political systems of the Member States, domestic change andadaptation take place. Changes due to European influence occur on two differentlevels: institutional adaptation and adaptation of policies and policy processes’.88 Suchchanges to the political structure of a Member State indicate that ‘a change on the mostfundamental level has taken place, and the effects of such a change are thus among themost far reaching. There are indications that even if the constitution of a state is notformally changed in order to adapt to EU membership, the constitutional practices ofMember States are in fact altered’.89 Withdrawal would significantly disrupt thecomplex mechanisms created by the process of Europeanisation in the withdrawingstate.90 This is particularly true for the old Member States, whose political and consti-tutional structures have been shaped by EU membership for many decades.91

C Conclusion

In conclusion, it can be argued that while withdrawal may now be legally possible,political and economic considerations will nevertheless rule out withdrawal as a real-istic option.

V Conclusion—Should I Stay or Should I Go?

Article 50 of the revised EU Treaty is not a masterpiece of legal draftsmanship.92 Inparticular, the procedural provisions lack clarity and are incomplete. ‘During thedifficult period that would surround any hypothetical withdrawal, one would expect arule that would provide clear and unambiguous answers.’93 Regrettably, the contrary isthe case. Article 50 is remarkably deficient in this respect. Moreover an unfettered rightto withdraw—as envisaged by Article 50—is open to abuse, in particular by large andeconomically powerful states.

What is needed instead is an Article that subjects withdrawal to clear and stringentconditions. For instance, withdrawal should only be possible under exceptional cir-cumstances that would need to be clearly defined.94 This would minimise the risk ofabuse: states would thus no longer be able to (ab)use the right to withdraw as abargaining tool, since they could no longer threaten to leave whenever it suited them.95

While withdrawal may now be legally possible, states are unlikely to make use of thisnew right for economic and political reasons. Withdrawal would be a very expensiveadventure: the withdrawing state would incur huge economic costs in the process ofreversing the impact of Europeanisation. For example, it would have to reinstate tariffsand customs duties. Moreover, citizens of the withdrawing state would no longer enjoy

88 B. Sittermann, Europeanisation—A Step Forward in Understanding Europe?, Working Paper Nachwuchs-gruppe Europäische Zivilgesellschaft, Universität Münster (2006), at 6.

89 Berglund, op cit n 81 supra, at 156.90 ibid, at 157.91 ibid, 157.92 See n 69 supra.93 Friel, op cit n 21 supra, at 426.94 See n 71 supra.95 But only in clearly defined circumstances as mentioned above.



access to the single market96 and hence would not benefit from the four fundamentalfreedoms anymore.97

In short, it can be concluded that even if the right to withdraw has now entered intoforce, economic and political considerations will nevertheless rule out withdrawal as arealistic option.

96 This depends, of course, on the content of the withdrawal agreement, provided such an agreement isconcluded at all.

97 If the withdrawing Member State wants to avoid this and negotiates a withdrawal agreement along thelines of the European Economic Area Agreement it faces another problem: it will have to maintain mostof the acquis.



‘should i stay or should i go ?’—a critical analysis of the right to withdraw from the eu

Documents