The arguments developed over the two previous Parts leave us with a paradox. Disenfranchising autocratic Member States in the Council via Article 7 is anti-democratic, but so is letting them vote in the Council. Militant democratic theory offers a way of thinking through such paradoxes, as it theorizes precisely whether and when democratic polities are justified in acting anti-democratically in order to protect their democratic character. Militant democratic views—holding that acting anti-democratically is sometimes justified to protect democracy—are controversial (Invernizzi Accetti and Zuckerman 2017). Yet, crucially, even if we suppose, arguendo, that some form of militant democracy is acceptable in some cases (that is, if we assume that anti-democratic measures may be justified when democratic institutions are under an existential threat), Article 7 is still illegitimate. This is because there is a coherent measure that is not anti-democratic that the EU could turn to in such circumstances—the expulsion of the offending state from the EU.
The theory of militant democracy holds that, under exceptional circumstances, democratic states are normatively justified in acting anti-democratically in order to preserve democracy. Militant democratic theory is a particularly suitable theoretical terrain for exploring the above paradox as it explicitly recognises that there is a democratic cost to militant democratic policies. It was formulated first in the 1930s by Karl Loewenstein (1937). Loewenstein’s foil was European and especially German fascism. He noted that fascists were eager to use the procedures and rights of the liberal-democratic state in Germany to their advantage (ibid., pp. 423–424). When powerful anti-democratic actors undermine and discredit democratic procedures and institutions in this fashion, democracy must fight back. ‘If democracy believes in the superiority of its absolute values over the opportunistic platitudes of fascism, it must live up to the demands of the hour, and every possible effort must be made to rescue it, even at the risk and cost of violating fundamental principles’, including the ‘suspension of fundamental rights’ (ibid., p. 432, my emphasis).
The theory of militant democracy is a ‘non-ideal theory’;Footnote 15 the ideal, for democratic states, is not to be ‘at war’ against anti-democratic actors, and thus to remain within the ‘normal’ confines of the rule of law and the democratic constitution. Whether or not democrats agree with the militant democratic view that militant policies can sometimes be justified in the face of serious anti-democratic pressure, it is clear (though too often overlooked) that any such measures come at a cost to democracy; this is what Kirshner calls the ‘paradox of militant democracy’ (2014, p. 2). In other words, even those sympathetic to militant democracy think that there is a pro tanto democratic cost to militant measures. The difference between the pro- and anti-militant positions lies in the view of militant democrats that anti-democratic measures are sometimes all-things-considered instrumentally justified despite these costs.
I do not wish here to tackle the justifications for militant democracy at the most fundamental level—whether or not militant democratic measures can ever be justified.Footnote 16 Rather, let us assume, for the sake of argument, that such measures can be justified democratically, but only when the sorts of conditions described by Loewenstein are in play and at least two conditions are met. First, there is an existential threat to democracy.Footnote 17 Second, militant measures are necessary—if there are non-militant measures that could pacify the autocratic enemy at democracy’s gates, then the cost of sacrificing democratic fundamentals is too high. Thus, to return to the normative questions specific to our case, we must ask: a) whether EU Member States backsliding on democratic values (could) constitute a fundamental threat to the EU’s democratic character, and, b) whether a militant response such as disenfranchisement in the Council would be appropriate, or if there are non-militant alternatives.
To take the first question: surely backsliding can constitute an existential threat to the EU’s nature as a political community of states sharing a commitment to democracy, equality, the rule of law and so on. For a start, a Member State being in serious and persistent breach of these values would, it seem, itself constitute a threat to the EU being able to self-affirm its democratic character and, in the absence of a reaction from the EU, its deep-seated commitment to fundamental values (Kelemen 2017). Further, and as argued above, it follows from the indirect democratic legitimation of Council decisions that there is a threshold of the domestic democratic illegitimacy of EU governments that would lead us to conclude that Council-voting is no longer even a derivatively democratic procedure.
The last question is whether the sanction described in Article 7.3—disenfranchisement in the Council—would be the appropriate militant response, or if there are other, non-militant measures that could be invoked that would ‘defend’ democracy. Because even if the threat posed by backsliding Member States is sufficient to demand urgent action, anti-democratic action would only be warranted (on the militant democratic view I set out above) if there are no alternative avenues that do not undermine democratic values and could be reasonably expected to safeguard the democratic nature of the institution or procedures under threat.Footnote 18 And, at least in principle, such possibilities do exist in the context of the EU and European integration: the expulsion of a Member State.
The Need for an EU Expulsion Mechanism
An expulsion mechanism has rarely been seriously considered in the literature (cf. Athanassiou 2009, pp. 32–36; Hausteiner 2020, pp. 15–16; Müller 2015, pp. 145–150; Olsen 2019; Patberg 2019, 2021; Bellamy and Kröger 2021, p. 632; Scherz 2021). In 2016, Jean Asselborn, Minister of Foreign Affairs for Luxembourg, caused a stir when he came out in favour of a potential recourse to expulsion: ‘We cannot accept that the fundamental values of the European Union are flagrantly violated. Those who, like Hungary, build fences against refugees fleeing war or violate the freedom of the press and the independence of the judiciary, should be temporarily or, if necessary, permanently expelled from the EU’.Footnote 19 More recently, when discussing Hungarian and Polish opposition to a conditionality mechanism for the coronavirus recovery fund in Parliament, Dutch Prime Minister Mark Rutte wondered ‘Can you make a budget via an intergovernmental agreement, or can you found an EU without Hungary and Poland?’Footnote 20
One reason that the normative case for an EU expulsion mechanism is largely absent in normative theorizing may be that democratic theorists usually oppose the domestic analogues to expulsion: banishment and ‘denaturalization’. Banishment and denaturalization, the argument goes, violate the fundamental rights of the banished, for instance by making them stateless. All citizens of democratic states have the right to remain in their state (albeit in prison) and everyone has the right to a nationality. This is part of the ‘right to have rights’, to use Hannah Arendt’s famous formulation (1951, p. 294). Such a right is non-contingent. No one, on this view, is utterly beyond the pale—at least in terms of their right to political membership.
However, in contrast to citizens’ inalienable right to have rights, states do not have a fundamental right to trans-, inter- or supra-national associations with other states. Such associations are voluntary.Footnote 21 States would not have their fundamental rights violated by being excluded from such associations. Thus, and consequently, a state’s eventual expulsion from such an association is not at odds with their fundamental civic and political rights. Nor, as long as the process of exclusion happened within the appropriate legal framework, would the expulsion from the European Union of a Member State in serious and persistent breach of EU fundamental values, especially when such a breach constitutes an ‘existential threat’ to the democratic character and legitimacy of EU rules and Council decisions (cf. Hausteiner 2020, pp. 59–62; Athanassiou 2009, pp. 32–36; Scherz 2021). Indeed, expelling a Member State that had become frankly autocratic may be the only way to preserve the democratic character of the European Union.
Democratic Costs to Excluding Autocratic Members
Before we conclude in favour of an expulsion mechanism as the appropriate final response to a Member State becoming frankly autocratic, we must consider the democratic costs to expelling a backsliding Member State—particularly with regard to the loss of rights and freedoms of the citizens of that state (Olsen and Rostbøll 2017; Patberg 2021; Scherz 2021). The citizens of such a state would be deprived, following expulsion, of their EU citizenship. Consequently, they would lose many important freedoms such as the freedom to work and travel in other Member States. They would no longer be protected by the stringencies of EU labour, employment and competition law, nor could they have their trans-national interests represented in the European Parliament. These are just a smattering of examples of the deep and far-reaching ways in which citizens of an expelled state would be disadvantaged. Worse, many citizens in such a state would not have supported—and may have actively opposed—their government’s autocratization. With them in mind, their state’s expulsion from the EU seems harsh, lamentable, unjust (Olsen 2019, pp. 159–162; Patberg 2021; Scherz 2021). And even those citizens that supported their autocratic government, for instance with their votes, could not be held fully responsible for such choices given the unequal democratic playing field in a state backsliding on democracy and the rule of law (e.g. the absence of freedom of speech, a free press, independent courts and so on).
As a means of furthering the EU’s commitment to the Article 2 values, the expulsion of an autocratic state may also pose questions. The values of democracy and equality do not apply only to the interaction of Member States and EU processes and institutions, but also in relation to EU citizens. After all, Article 9 of the Treaty on European Union states that the EU ‘shall observe the principle of the equality of its citizens’, while Article 10 guarantees EU citizens the ‘the right to participate in the democratic life of the Union’. Would the expulsion of a frankly autocratic Member State not violate these rights, leading to a different sort of performative contradiction? Further, there is currently no Treaty mechanism to expel a Member State; the only way for a member to leave the Union is via the Article 50 procedure. Loewenstein argued for extra-legal militant responses in the context of his militant democratic theory. However, such a move would be difficult to justify in this context, since the rule of law is, just as much as democracy and equality, a fundamental value protected under Article 2.Footnote 22 Given that expulsion is not mentioned in the Treaties, could an expulsion mechanism proceed in accordance with the rule of law?
The democratic costs of expelling a state that had backslid too far on democracy and the rule of law cannot be taken lightly. Indeed, while an EU expulsion mechanism as a final political sanction of democratic and rule of law backsliding is necessary, it is equally, if not more urgent to use lesser, scaleable, targeted responses such as economic sanctions that would hopefully prevent expulsion from ever being necessary (see e.g. Kochenov and Pech 2016; Theuns 2020; Scheppele et al. 2021).Footnote 23 Yet, contrasting the expulsion of an EU Member State with the denaturalization or banishment of an individual citizen helps show that no fundamental rights are violated in the former case (in contrast to the latter). As argued above, the EU is a voluntarist association and (member) states have no unqualified right to membership.Footnote 24 By extension, and in keeping with European lawFootnote 25 and ECJ jurisprudence,Footnote 26 EU citizens have a derivative and not a direct right to participation in Union democratic life. Their state’s exit from the Union would not be undemocratic in itself, and would not violate Articles 9 and 10 TEU (cf. Olsen and Rostbøll 2017).
An EU expulsion mechanism is also consistent with the rule of law. It is true that the EU Treaties do not address the expulsion of a Member State from the EU.Footnote 27 If the argument in this article is correct, then a proposal to add such a mechanism through a revision to the Treaties would be warranted normatively. However, surely (at least) Member States currently under review for democratic and rule of law backsliding would resist such changes, and their consent would be required. It seems that an expulsion mechanism may be desirable from the perspective of democratic theory, but is unrealistic. There is another possibility though, evoked by Rutte (cited above): refounding a new European Union with stronger democratic and rule of law protections and without recalcitrant members of the current EU (Chamon 2020; Chamon and Theuns 2021). This would require EU Member States committed to democracy and the rule of law to collectively withdraw from the EU via the Article 50 procedure and refound a new European Union. Autocratic Member States would be left with a useless husk of the former EU. Such a procedure to withdraw and refound requires no treaty change and violates no fundamental values.
While the eventual expulsion of an EU Member State through a collective withdraw and refounding procedure may seem outlandish, consider the alternative. If current trends of democratic backsliding continue, and a Member State becomes frankly autocratic, could pro-democratic states and citizens countenance being bound to it through continued supranational union? No. Supranational union with an autocratic state would taint all EU law and governance.Footnote 28 Doing nothing about backsliding makes all EU Member States committed to fundamental values complicit (Theuns 2020). Being open to the eventual recourse to an expulsion mechanism is essential if the European Union is to guarantee its commitment to the fundamental values of Article 2.