Skip to main content

The Need for an EU Expulsion Mechanism: Democratic Backsliding and the Failure of Article 7


What should the EU do about the fact that some Member States are backsliding on their commitments to democracy, supposedly a fundamental value of the EU? The Treaty provisions under Article 7 TEU are widely criticized for being ineffective in preventing such developments. Are they legitimate? I argue that the ultimate sanction of Article 7 TEU falls into a performative contradiction, which undermines its ability to coherently defend fundamental values. Instead, expulsion from the EU is the appropriate, coherent and legitimate final political sanction for democratic and rule of law backsliding by a Member State. The argument has the following steps: In Part 1, I argue that the current Article 7 framework for responding to democratic and rule of law backsliding in the EU is normatively problematic, in that the mechanism currently in the Treaty undermines the values it purports to defend; in other words, it falls into a performative contradiction. It is undemocratic to deprive Member States of their right to vote in the Council while holding them subject to Council decisions. However, Part 2 studies relevant philosophical arguments from an adjacent literature on criminal disenfranchisement, concluding that allowing backsliding Member States to keep their voting rights in the Council also taints the democratic character of Council decision-making. In Part 3, I consider the resulting paradox in light of the literature on militant democracy. Could militant democracy justify Article 7? I argue not; even if we accept the hypothetical justifiability of militant measures, they are not legitimate here since a democratically acceptable alternative exists that would safeguard the democratic character and legitimacy of Council decision-making: expulsion from the Union. I also address a central objection to an expulsion mechanism—that it would require treaty change and is therefore practically impossible.


Over the last five or six years, discussions over what reaction the EU ought to have to its Member States backsliding on core democratic and rule of law commitments have taken an increasingly urgent tone (e.g. Kelemen 2017; Kochenov and Pech 2016; Müller 2015; Pech and Scheppele 2017; Niklewicz 2017; Oliver and Stefanelli 2016; Bárd 2018; von Bogdandy and Spieker 2019; Gora and de Wilde 2020; Bellamy and Kröger 2021). The legal context for some of these responses are defined in Article 7 of the Treaty on European Union (‘Article 7’). This article develops a new and original critique of Article 7 based on it being normatively incoherent in that it is itself in conflict with EU fundamental values. I argue that expulsion from the EU rather than disenfranchisement in the Council is the appropriate, coherent and legitimate final political sanction for democratic and rule of law backsliding by an EU Member State.

The two states most often in discussion in relation to Article 7 are Hungary and Poland. At the time of writing, rule of law proceedings are underway against these two states for violating EU fundamental values.Footnote 1 However, both countries have indicated that they would support the other in these proceedings, so it seems unlikely that they will result in Article 7 sanctions given unanimity requirements (see Part 1, below. Also: Niklewicz 2017; cf. Kochenov 2017). Unsurprisingly, this has resulted in much criticism of Article 7.

Commentators have focused much of their ire on the unanimity requirement of Article 7.2, but have also criticized the very slow speed of the proceedings, and their politicized nature (e.g. Kochenov and Pech 2016; Niklewicz 2017; cf. Sedelmeier 2017). For instance, while concerns with Hungarian democratic backsliding have been around at least since Fidesz won the 2010 general election in Hungary, and have intensified progressively after their re-election in 2014 (Oliver and Stefanelli 2016) and 2018, the first concrete steps towards rule of law proceedings against Hungary under Article 7 were made in Autumn 2018. Procedures against Poland were quicker to start following the constitutional crisis of 2015, but anti-democratic actions by the PiS party have intensified since ‘dialogue’ with the European Commission started (Kovács and Scheppele 2018). Regarding the politicization of the proceedings, it has been argued that Fidesz, though recently a defector from the EPP group, was protected for a long time at the EU level by their EPP membership given the EPP dominance in the Council and in Parliament (Kelemen 2017, p. 220).

Alongside criticisms of the unanimity requirement of Article 7.2 and of the slow speed and politicized nature of Article 7 procedures, commentators have also criticized the sanctions laid out in Article 7.3. Since stripping a Member State’s votes in the Council is symbolically weighty, observers think it is unlikely ever to be used to sanction a Member State (e.g. Oliver and Stefanelli 2016; Sedelmeier 2017; cf. Kochenov 2017). I have previously argued that Article 7 is normatively suspect qua democratically illegitimate, in that it breaks with the ‘All Subjected Principle’ of democratic theory (Theuns 2020). Some also note that the sanction amounts to a sort of ‘moral quarantine’ rather than directly addressing the fact of backsliding in a Member State (Müller 2015, p. 144, emphasis in the original). Proposals for better using other existing mechanismsFootnote 2 or developing alternative sanctions mechanisms abound (indeed, this paper is in part a contribution to this literature), and tend to focus on swifter, milder political sanctions, on depoliticizing the procedure by giving a stronger place to the courts, or complementing political with economic sanctions (e.g. Kochenov and Pech 2016; Blauberger and Kelemen 2017; Pech and Scheppele 2017; von Bogdandy and Spieker 2019; Theuns 2020; Bellamy and Kröger 2021).

This paper makes two main related claims.

  1. 1.

    Disenfranchisement in the Council is in conflict with EU fundamental values and ought to be abandoned.

  2. 2.

    Expulsion from the Union is the appropriate, legitimate, normatively coherent final political sanction for Member States in breach of the values referred to in Article 2.

The paper is structured as follows: Part 1 argues that part of Article 7 falls into a performative contradiction, which undermines its ability to adequately express the EU’s commitment to Article 2 values. Part 2 takes a contrasting tack, arguing that doing nothing were a Member State to seriously flaunt fundamental values would taint the democratic character of Council decision-making. In other words, there seems to be a paradox between the anti-democratic nature of disenfranchisement in the Council, and the anti-democratic nature of permitting a Member State in serious and persistent breach of fundamental values to vote in the Council. In Part 3, I consider this paradox in light of the literature on militant democracy, which theorizes precisely the justifiability of acting contrary to democratic values in order to defend democracy in times of crisis. Yet, I conclude that the appropriate sanction in such extreme circumstances would be expulsion from the Union rather than disenfranchisement in the Council. I also address a central objection to an expulsion mechanism—that it would require treaty change and is therefore practically impossible.

The Need for Expressive Consistency in Defending EU Fundamental Values and the Performative Contradiction of the Current Article 7 Framework

This Part examines the normative coherence of the rule of law framework as it is currently formulated in Article 7. I argue that one of the key purposes of Article 7 is to express commitment to the fundamental values listed in Article 2, and that it fails in this purpose by itself undermining the values of equality and democracy.Footnote 3 In other words, it falls into a performative contradiction.

The Purpose of Article 2 and Article 7

Article 7 holds that:

  1. 1.

    On a reasoned proposal by one third of the Member States, by the European Parliament or by the European Commission, the Council, acting by a majority of four fifths of its members after obtaining the consent of the European Parliament, may determine that there is a clear risk of a serious breach by a Member State of the values referred to in Article 2 […]

  2. 2.

    The European Council, acting by unanimity on a proposal by one third of the Member States or by the Commission and after obtaining the consent of the European Parliament, may determine the existence of a serious and persistent breach by a Member State of the values referred to in Article 2 […]

  3. 3.

    Where a determination under paragraph 2 has been made, the Council, acting by a qualified majority, may decide to suspend certain of the rights deriving from the application of the Treaties to the Member State in question, including the voting rights of the representative of the government of that Member State in the Council […]

  4. 4.

    The obligations of the Member State in question under this Treaty shall in any case continue to be binding on that State […]

An interesting aspect of Article 7 is that at every stage of the process the modality ‘may’ is invoked (‘the Council… may determine/decide’). This guarantees a margin of manoeuvre even once the European Council has determined a serious and persistent breach; there is no legal ‘duty’ for the Council to sanction a backsliding state, and no role for the ECJ (Sadurski 2010, p. 394). In other words, it is politics ‘all the way down’. This gives us an important clue that one of the purposes of Article 7 is centrally political—to communicate something—rather than being geared solely at the rectification of a legal wrong. Article 7, in this sense, is not only the ‘enforcer’ of Article 2 values, but also their ‘loudspeaker’, functioning to underline the EU’s supposedly unequivocal and deep commitment to the fundamental values listed in Article 2.Footnote 4

It is also not immediately clear from the wording of Article 7 what its purpose is. Yet it is crucial to interrogate the purpose of the Article if we are to take a position on whether Article 7 is normatively coherent (i.e. consistent with its own aims). As a sanctions mechanism, it is useful to look to the extant theories of justifying penal sanctions for illumination. With this in mind, several possible purposes or functions of Article 7 can be imagined.

  • A dissuasive purpose. On this line of thought, the existence of weighty sanctions themselves could motivate Member States to respect Article 2 values. This purpose would be similar to the justification of criminal law sanctions that grounds these in their deterrent function.

  • A punitive purpose. Here the focus is on avenging the supposed wrong that transgressing Article 2 values may constitute. This purpose would be structurally similar to justifications of criminal sanctions that emphasize their retributive function.

  • An expressive purpose. Here the purpose would be to express the EU’s continuing commitment to, and the importance of, Article 2 values. This function tracks the denunciative rationale of penal sanctions.

  • A reformative purpose, where the aim is centrally to change the behaviour of the Member State putatively in violation of Article 2 values. This purpose would track justifications of penal sanctions focused on rehabilitation.

  • An inoculative purpose. If the main intention is to prevent governments backsliding on democratic values to influence Council decisions (by stripping them of a vote), we might say the purpose is inoculative. This purpose is similar to justifications of penal sanctions focused on incapacitating the offender.

Likely elements of each type of justification for a sanctions-based response can be identified in Article 7 in some shape or form. The normative contradiction I wish to highlight, however, focuses on the intersection of the expressive and the reformative purposes, although the importance of the inoculative purpose will return in Part 2 when I discuss the corruption of Council decision-making by autocratic members.

The expressive function of Article 7 can be understood along two lines. First, Article 7 has the function of underlining the normative importance of the values listed in Article 2, for the European Union. It is this sort of purpose Sadurski has in mind when he writes that sanctions ‘should be seen as a general ideological declaration setting the limits of what is politically acceptable within the Union’ (2010, p. 405).Footnote 5 Regardless of whether certain states in fact are (at risk of) breaching Article 2 values, the specification of a weighty political sanction for such a breach serves to bring the Member States (closer) together around a shared perspective of which values are of fundamental political importance. This could be called the ‘declarative’ expressivist function of Article 7. Second, Article 7 can have the function of expressing to Member States the costs and the weightiness of breaching fundamental values. This second perspective focuses on the ‘instructive’ aspect and is geared towards pushing a backsliding Member State to reform practices that lead them to be in breach of those values (Poama and Theuns 2019, p. 797). These perspectives are related to one another in that in order to instruct Member States regarding their breach of fundamental values these values must be clearly expressed—the reformative purpose of Article 7 rests in part on the clarity and coherence of the expressive purpose.

Article 7 Understood as a Performative Contradiction

For both the expressive and the reformative functions of Article 7 to be successful, it must minimally be the case that the procedure described in Article 7 does not itself undermine the values it purports to express. Yet Article 7 fails to live up to the values of democracy and equality, and thus falls into a performative contradiction, or so I will argue.

A performative contradiction, in its simplest form, is ‘when the content of a particular statement is in conflict with its utterance’ (ibid., p. 800), like when saying ‘I do not exist’ (Hintikka 1962, p. 32). In the context of the expressive purpose of a rule or law, the relevant performative contradiction is when a rule or law intended to express commitment to certain values itself undermines those values. The claim I defend here is that stripping a Member State in serious and persistent breach of EU fundamental values of their right to vote in the Council itself undermines the EU fundamental values of democracy and equality. As such, Article 7 is in a performative contradiction with the fundamental values listed in Article 2: it cannot adequately express them and, consequently, is hampered in both its declarative and instructive functions.

The approach I take here focuses on performative consistency and normative coherence. As such, it operates on an ‘internal’ rather than an ‘external’ mode of critique. This is related to methods of ‘immanent critique’ whereby a minimal normative standard of a political practice is that it does not violate its own normative premises (Nicolaïdis 2013; Theuns 2017). There are many ways immanent critique has been interpreted, but the analysis here consists of the critical normative evaluation of a practice grounded not on an external set of normative values but on justificatory standards drawn internally from legal and political praxis.Footnote 6 There are two elements to such an immanent critique: first, one must isolate the evaluative standards from the ‘evolving aspirations, tensions and contradictions within this world observed’ (Nicolaïdis 2013, p. 357). Second, one must evaluate the extent to which the practice lives up to its normative commitments.

Following this two-step approach, then, one must first ask oneself what, minimally, a commitment to the fundamental value of democracy and equality requires in the context of EU decision-making. The values of democracy and equality are notably hard to pin down (Mos 2020). However, not ‘everything goes’. Minimally, respecting democracy and equality requires that each agent in a political process has a formally equal stake in that process, that no agents’ interests are arbitrarily weighted more than any other and that agents are able themselves to determine how their vote will be cast (Christiano 2008, pp. 75 ff.). Valuing democracy and equality in this way does not require that agents’ interests or views are judged to be equally valuable, but rather that their views, filtered through a political procedure that accords them formally neutral weighting, are treated equally.Footnote 7

What does this mean for a supranational and multilevel polity such as the European Union? At the level of Council decision-making it means, minimally, that each state subjected to an authoritative rule co-determined by Council voting ought to have an proportionally equal stake in determining that rule. This approach to democratic legitimacy overlaps with what is known as the all-subjected principle, whereby all those permanently subject to a rule ought to have an equal stake in authorizing it (Theuns 2020; cf. Beckman 2014). The idea here is that it is necessary and sufficient to be a permanent subject of a particular rule in order to be authorized to have a democratic stake in its formulation. It is enough, under such a principle, to be formally bound by a rule or decision, even if one is not otherwise heavily affected by that rule or decision.Footnote 8

While I am partisan in favour of a version of the all-subjected principle in democratic theory (Theuns 2021), the task here is not to defend it. Rather, as is more appropriate to the ‘immanent’ method we have adopted, it is sufficient to note two things. First, it is the standard that best corresponds to the juridical framework of the EU legal order—only those states that are full members of the European Union are ordinarily and automatically bound by EU law.Footnote 9 Second, it allows us to clearly identify which persons or agents are and which are not legally subject to EU law. Once jurisdictional boundaries have been drawn, as they have in the EU, it is not a technically complex or politically controversial decision to determine the corresponding ‘demos’. This is especially true of Council decisions: all Member States are subject to the rules and decisions co-legislated in the Council, within the boundaries of the scope of its authority and prerogative stipulated in the EU treaties. We must therefore interpret the demands of equality in democratic decision-making (via the Article 2 fundamental value of democracy) to apply, immanently, between EU Member States in those supranational processes that collectively bind them.Footnote 10 To sum up, the fundamental values of democracy and equality, as interpreted through the lens of the EU legal and political order, require that states subject to EU law—EU Member States—participate as equals in EU law-making and governance.Footnote 11

Does the sanctions procedure in Article 7 live up to this demand of democratic equality? Recall that the final political sanction outlined in Article 7.3 includes the suspension of ‘the voting rights of the representative of the government of that Member State in the Council’. Importantly, a Member State sanctioned in this way would lose their right to vote in the Council while still being subject to the decisions made in the Council (given Article 7.4). But this violates the demand that all Member States subject to EU legislation participate as equals in EU law-making. Indeed, Wojceich Sadurski has noted that disenfranchisement under Article 7 is ‘functionally almost equivalent to temporary expulsion, with the crucial difference that the obligations of the Member State in question remain in place’ (2010, p. 390, emphasis in original). The sanctions procedure in Article 7 thus undermines the values of democracy and equality stipulated in Article 2, and falls into a performative contradiction. This contradiction undermines the normative coherence of Article 7 and, consequently, the possibility of Article 7 adequately declaring a commitment to EU fundamental values and instructing backsliding states in returning to the democratic fold.

The anti-democratic character of the Article 7 procedure is one horn of a fundamental dilemma regarding what the EU ought to do about democratic backsliding. As I argue in Part 2, doing nothing about backsliding states voting in the Council also undermines democracy. To let autocrats participate in EU law-making and governance undermines the democratic legitimacy of these processes, but to disenfranchise them while holding them subject to EU law breaks with the values of democracy and equality. Before addressing this paradox, it is worth considering an important objection to the first horn of the dilemma.

One may think that disenfranchising a backslidden state would not violate democratic principles as, ex hypothesi, the backslidden state’s government would no longer be a legitimate democratic representative of its people.Footnote 12 How could it be democratically problematic to disenfranchise an autocratic member? Autocratic governments have no democratic claim to represent their peoples in supranational democratic procedures.

This objection goes astray as it focuses on the wrong aspect of the problem. Indeed, as I argue in the next section, an autocratic Member State cannot be included in a legislative process committed to a democratic ideal.Footnote 13 Autocrats, by definition, cannot be legitimate democratic representatives and empowering autocrats does not further democratic ends. However, ruling over such disenfranchised governments (and their citizens) by holding them subjected to the decisions of a body in whom they have no democratic representation is no solution. It is subjection without equal representation that is the crux of the objection I have made to the procedure outlined in Article 7. Granted, one may consider that this objection is tempered somewhat by the continued representation that the citizens of a disenfranchised state would enjoy in the European Parliament (Bellamy and Kröger 2021). However, the fundamental legislative importance of the Council, the citizens of a disenfranchised state would be subject to EU laws without having an equal/proportional stake in their authorship. As such, the Article 7 procedure is at odds with the fundamental values of democracy and equality—the performative contradiction described above.

Do Votes by Backsliding States Corrupt Council Decision-Making?

Stripping a backsliding Member State of their vote in the Council seems analogous, in some senses, to stripping criminals of their voting rights. In both cases a breach of a legal norm results in a sanction stripping the offender of their democratic voting rights. This analogy is useful since criminal disenfranchisement has been extensively treated in political and legal theory, while there are hardly any normative analyses of stripping an EU Member State of its vote in the Council (cf. Müller 2015, p. 150; Theuns 2020; Scherz 2021). I will use the theoretical literature on criminal disenfranchisement to tease out a paradox; on the one hand, we have seen in Part 1 that Article 7 is deficient from the perspective of democratic legitimacy in that the sanctions mechanism it lays out falls into a performative contradiction with the values it purports to defend. However, in this Part I will show that doing nothing is also in conflict with democratic norms.

While I will argue some arguments in favour of criminal disenfranchisement further illustrate the flaws of Article 7, a ‘republican’ argument in favour of criminal disenfranchisement draws attention to the costs of permitting a backsliding state to continue to participate in EU legislative processes. As well as a normative paradox, this leaves us with a practical dilemma: disenfranchising a backsliding Member State while continuing to hold them subject to Council decision-making is democratically illegitimate, but their continuing to vote also undermines the democratic legitimacy of the EU. A classic catch-22, or so it seems.

Attempts to justify criminal disenfranchisement take different forms. One version argues that criminals break a social compact with their criminal acts, and should thus be stripped of the political rights associated with upstanding citizenship (e.g. Ewald 2002). Another emphasizes the deterrent function of criminal disenfranchisement (e.g. Cholbi 2002; Bennett 2016). A third argument, sometimes called the ‘republican’ argument, holds that there is a need to keep electoral politics ‘pure’. On this version, allowing serious felons to vote would ‘pollute’ elections with their supposedly malevolent views and nefarious interests (e.g. Manfredi 1998). Are these arguments convincing? If so, can analogous arguments be made to support the disenfranchisement of Member States backsliding on EU fundamental values?

In democratic theory, the above strategies for justifying criminal disenfranchisement have attracted widespread criticism. In response to the ‘social compact’ argument, critics have pointed out that criminals remain subject to the laws and decisions of their state, so any social compact they have supposedly violated remains in place after that act of violation—indeed it is on that very logic that their penal sanctions are justified (e.g. Schall 2006; Beckman 2009). Given that Article 7.3 also explicitly notes that sanctioned Member States’ obligations under the Treaty of European Union will continue after their disenfranchisement, it seems that a justification for Article 7 grounded on a ‘social compact’ argument would fail for the same reason—the sanction presumes that any supposed social compact remains in place subsequent to backsliding.

Responding to the second type of justification of criminal disenfranchisement, focusing on its deterrent effect, critics have pointed out that there is no convincing evidence that the empirical assumptions undergirding this logic are correct (e.g. López-Guerra 2014). There is no measurable drop in crime rates corresponding to the installation of criminal disenfranchisement policies, nor a rise associated with such policies being abandoned. While it would take careful empirical research to secure the analogous conclusion in the context of EU, the mere fact that the high-profile cases of democratic backsliding in EU Member States have taken place after the entry into force of the Treaty of Lisbon in 2009 (and Article 7) suggests that the deterrent function of Article 7 is weak.

The last, ‘republican’, justificatory strategy for criminal disenfranchisement is more interesting for our purposes. In the context of criminal disenfranchisement, the notion that criminals ‘pollute’ elections rests, as critics have pointed out, on speculative and empirically unwarranted assumptions about criminals’ voting behaviour. It presumes that criminals vote with ‘nefarious’ motivations, rather than in line with their legitimate interests and judgements (Reiman 2005; Sigler 2014; López-Guerra 2014; Poama and Theuns 2019). Furthermore, criminals individually have very little voting power (like all citizens taken individually), thus reducing the likelihood of them ‘corrupting’ aggregate results, even if we were to accept the (dubious) claim that a significant proportion of them vote ‘malevolently’ when they can. Yet, these rebuttals seem to have less purchase on an analogous justification for stripping backsliding Member States’ votes in the Council with a view to inoculating Council decisions from backsliders’ anti-democratic influence.

There are several reasons why such a ‘republican’ argument is more convincing in the context of Article 7. Member States potentially facing sanctions under Article 7 would have been found to be in serious and persistent breach of Article 2 fundamental values. It is not a stretch to imagine that an EU Member State willing to undermine, for instance, judicial or academic independence, or flaunt the basic human rights of refugees or ethnic minorities (to cite some of the charges against Poland and Hungary) at the level of national law would seek also to undermine these values at the European level. Whereas most criminal acts have little if anything directly to do with the criminals’ political orientation, the actions of backsliding Member States in serious and persistent breach of EU fundamental values are squarely in the realm of law and politics.

Also in the supposed lack of electoral power there are pertinent differences that ought to give us pause. While each individual felon has little aggregate effect on a national electoral outcome, Member States have substantial voting power in the Council. In any of its configurations, the Council ordinarily has 27 voting members. Many votes are taken by consensus—even when the voting procedure does not mandate it. Individual votes are thus very powerful in the Council. In sum, neither of the objections made to the republican justification for criminal disenfranchisement seem to stick against an analogous republican argument for stripping EU Member States backsliding on democratic values of their right to vote in the Council. Notwithstanding the argument in Part 1, there seems to be an important republican case for ‘inoculating’ Council decision-making from the anti-democratic influence of backsliding states.

There is one further important difference with criminal disenfranchisement which also lends itself to the conclusion that states violating fundamental values ought not to participate in Council law-making; at the national level one way the democratic character of a procedure is ascertained is by measuring the degree to which all permanent subjects have a formally equal share of political power. But the democratic character of the Council is different. Granted, there are voting procedures, with every EU Member State wielding a neutrally weighted vote through their government ministers. But the Council has no direct democratic mandate. The Council’s democratic legitimacy is derivative—it depends on the democratic legitimacy of the governments making up its members. So it follows that if a backsliding Member State has dropped below some relevant threshold of democratic legitimacyFootnote 14 the derivative democratic legitimacy of the Council would suffer.

In sum, while arguments analogous to the ‘social compact’ and ‘deterrent’ arguments for criminal disenfranchisement cannot justify stripping a backsliding state of their right to vote in the Council, the republican argument for criminal disenfranchisement, while dissuasive in that context, lends heft to stripping Member States in serious and persistent breach of the EU fundamental values of their vote in the Council.

Militant Democracy, Banishment and Expulsion

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.


An EU expulsion mechanism is the only solution to the paradox of the illegitimacy of disenfranchising a Member State via Article 7 and the illegitimacy of having backsliding states continue to participate in EU legislative procedures. Article 7 is currently itself in conflict with the EU fundamental values of equality and democracy. This conflict cannot be justified via recourse to a militant measure such as disenfranchisement in the Council, since, even if we accept the hypothetical justifiability of militant measures, a democratically acceptable alternative exists that would safeguard the democratic character and legitimacy of Council decision-making: expulsion from the Union. Disenfranchisement in the Council should thus be replaced with expulsion from the Union as the ultimate political sanction for Member States flagrantly violating EU fundamental values.

This is neither to say that the ‘existential threat’ threshold discussed in Part 3 has now been met in the cases of Hungary and Poland, nor to say that expulsion from the Union ought, currently, to be pursued against them. I take no position on these questions here. Yet given the severity of developments in these states, alternative sanctions to the ‘ultimate’ political sanction currently described in Article 7.3 ought to be explored urgently, and ought to be supplemented by non-sanctions responses in support of pro-democratic actors (Theuns 2020). These do not pose the same threat to EU fundamental values as stripping a Member State of their voting rights in the Council. Indeed, an important conclusion to draw from the necessity of expulsion as a final political sanction is that more varied, scaleable sanctions ought to be used by EU actors committed to fundamental values to try to ensure that recourse to expulsion never becomes necessary.

I have argued for an important normative disanalogy between the illegitimate banishment or denaturalization of an individual from a state and the expulsion of a state from an association. Yet there is also an important point of similarity, at least in the context of the EU. Expelling a Member State from the Union against its wishes (and the wishes of its citizenry) would have grave effects on the rights and freedoms of its citizens, and on the rights and freedoms of other citizens of the European Union (Olsen and Rostbøll 2017). It ought never be considered lightly. However, the commitment to exclude seriously anti-democratic governments from the Union is vital to ensure that its democratic character and legitimacy can be preserved in the face of the all-too-persistent threat of autocratic politics.

That the last resort in terms of a political response to democratic and rule of law backsliding must be the expulsion of a Member State from the EU is undoubtedly a radical conclusion. Its radical nature seems both tempered and justified, however, when one considers that some EU Member States may continue to sink into increasingly anti-democratic politics. Optimists hope that budding autocrats will back down, or be overturned at the ballot box, perhaps after economic sanctions under rule of law budget conditionality start to bite. I share this hope. But what if they are not? Or what if they resist a peaceful transition of power? However unpalatable, EU Member States cannot both permit a frankly autocratic state to continue to be a member of the Union and at the same time commit to the fundamental values of Article 2.


  1. The Article 7 procedure against Poland was initiated when the Commission submitted a reasoned proposal to the Council that there was a clear risk of a serious breach of the rule of law in December 2017. The Article 7 procedure against Hungary was initiated when the European Parliament submitted a reasoned proposal to the Council in September 2018, paying special attention to the values of democracy, the rule of law and human rights.

  2. Such as the use of the new Conditionality Regulation, adopted on 16 December 2020, which permits the EU to cut off funds to member states backsliding on the rule of law when the sound financial management of the EU budget and/or the EU’s financial interests are at stake. Or for instance the use of ‘systemic’ infringement actions brought by the EU Commission under Article 258 TFEU or by member states under Article 259 TFEU (Scheppele et al. 2021).

  3. I use the term ‘purpose’ in contrast to ‘intention’ deliberately as a way of flagging the ‘purposive’ position in jurisprudence whereby legal interpretation seeks to identify the purposes of statutes based on an idealized standard of what reasonable legislators could be supposed to pursue (see e.g. Hart and Sacks 1994; cf. Posner 1985) rather than what a body of legislating persons actually intended—the ‘intentionalist’ view—or what a legal text means ‘objectively’ (for a discussion see Greenberg 2021).

  4. Article 2 reads: ‘The Union is founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law and respect for human rights, including the rights of persons belonging to minorities. These values are common to the Member States in a society in which pluralism, non-discrimination, tolerance, justice, solidarity and equality between women and men prevail’.

  5. Sadurski was referencing the short-lived unilateral sanctions taken by 14 member states against Austria in January 2000 for the inclusion of the far-right FPÖ in the Austrian government, however, not the Article 7 sanction.

  6. Immanent critique has as an advantage that any observer drawn or committed to the internal values and norms of a particular legal and political environment may feel implicated by the normative analysis. However, this quality also serves to delimit its normative value: where a legal or political environment is abhorrent, an immanent critique of its normative coherence is worthless.

  7. Whereas in national democracies this is ordinarily cast as ‘one person one vote’, that standard does not translate easily to a supranational polity such as the EU, where some member states are much more populous. The current EU Treaties aim to neutrally settle respective voting weights through balancing majorities of Member States and majorities of European citizens. Weighted voting, consensus norms and representation at the EU level all deserve more careful normative scrutiny, but such issues cannot be addressed here.

  8. So, for instance, childless retired persons’ judgements over appropriate standards for governing childcare are legitimately included in a democratic decision-making procedure upholding the all-subjected principle. As are the judgements of blind persons on the appropriate limits to roadside advertising.

  9. ‘Ordinarily’ given the possibility in some cases of opt-outs or exemptions, now a field of study in their own right as ‘differentiated integration’; ‘automatically’ given the direct effect of EU law.

  10. This is not to say that EU law does not constrain the choices of non-EU Member States. Think for instance of those EEA non-EU Member States that are, in the parlance that has become popular in light of the Brexit debate, ‘rule takers’. Or of those states in the EU’s ‘neighbourhood’ who are encouraged to adopt large swathes of the EU acquis communautaire in order to enter into privileged trading relationships with the EU (Theuns 2019). Yet, while such dependencies may violate a pro-democratic ethos, they do not in themselves violate the EU’s internal commitment to equality and democracy.

  11. For a contrasting analysis of EU fundamental values that emphasizes their inherent ambiguity see Mos (2020).

  12. I thank Nikolas Kirby and an anonymous reviewer for raising this objection. See also Schertz (2021).

  13. Again, there is a boundary-setting problem here over when states fall below the relevant democratic standards: ‘States that are sufficiently democratic ought… to have an equal stake in democratic rule-making and rule-authorization (in this context, an equal vote in the Council) while states that are insufficiently democratic cannot legitimately be bound to the Council’s decisions’ (Theuns 2020, p. 147).

  14. Determining that threshold will be complicated both in principle and practice (Theuns 2020, pp. 147–148).

  15. The ideal/non-ideal theory distinction comes from the work of John Rawls. For an overview of its use, see Zala et al. (2020, pp. 37–43).

  16. In other words, I am not endorsing the militant democratic position. Rather, the aim is to show that even if militant democracy is correct, and anti-democratic measures are justifiable under ‘existential’ circumstances, Article 7 is still illegitimate.

  17. I will not address here whether the threat must be to the system as a whole, or to the (more limited) enjoyment of democratic civil and political rights for some, as nothing in my argument turns on whether we read the existential threat condition strictly or loosely. For an interesting discussion see Kirshner (2014, pp. 50–55).

  18. Militant democratic theory takes democratic legitimacy and democratic equality as fundamental pro tanto values that can only be thwarted for their own sake—to ensure that a polity maintains its character as democratic in light of an existential threat. Theorists who prioritize a different fundamental value such as, for instance, egalitarian distributive justice or republican non-domination, might balk at the normative costs of expelling an autocratic member state. I cannot resolve disagreement on such bedrock normative principles here, but note that my key claim—that guaranteeing the democratic legitimacy of the EU requires the possibility of recourse to an expulsion mechanism—is not vulnerable to such disagreements.

  19. Wir können nicht akzeptieren, dass die Grundwerte der Europäischen Union massiv verletzt werden. Wer wie Ungarn Zäune gegen Kriegsflüchtlinge baut oder wer die Pressefreiheit und die Unabhängigkeit der Justiz verletzt, der sollte vorübergehend oder notfalls für immer aus der EU ausgeschlossen werden’ (my translation): Accessed 22 September 2021.

  20. Kun je een begroting maken via een intergouvernementeel verdrag of kun je nu een Europese Unie oprichten zonder Hongarije en Polen?’ (my translation): Accessed 22 September 2021.

  21. This view presupposes a vision of the EU polity focused more on the voluntarism of Articles 49 and 50 regarding, respectively, accession to and withdrawal from the Union, than on the EU as an ‘ever closer union’ from the preamble to the TEU and Article 1. In other words, it presupposes a rather more inter-governmental Union than a federalist or post-national one (Bellamy 2019, pp. 83–89).

  22. If expulsion would undermine Article 2 values anyway then there would be no reason to posit it as a better (qua non-militant) measure vis-à-vis disenfranchisement in the Council.

  23. One might also hope that acknowledging eventual recourse to an expulsion mechanism may make its use unnecessary, given widespread support for EU membership in, for instance, Poland and Hungary.

  24. Given this view of the EU it could be asked whether Member States could not legitimately volunteer to be subjected without representation. One may wonder whether a Member State sanctioned under Article 7 but deciding not to trigger Article 50 does not ‘voluntarily’ remain subject to the Treaties given their option to withdraw. However, the voluntarism of the EU as an association (i.e. Member States are members voluntarily and can leave) does not mean they could legitimately disenfranchise themselves, or accept their subjection. Just as vote-selling may be voluntary but cannot be democratic, so too for subjection to EU law without the right to vote in the Council.

  25. See, centrally, Article 20 of the Treaty on the Functioning of the European Union which makes EU citizenship derivative from the citizenship of a Member State. For discussion see van den Brink and Kochenov (2019).

  26. In, for example, the Kaur, Grzelczyk and Zambrano decisions of the Court of Justice of the European Union (see ibid., pp. 1372–1373).

  27. Notwithstanding Christophe Hillion’s controversial suggestion that the ‘continued and deliberate defiance of the core principles of [EU] membership’ by backsliding states amounts to those states notifying the Council of their intention to withdraw from the EU under Article 50 TEU (Hillion 2020; cf. Scholtes 2020; Hofmann 2021).

  28. This extends beyond Council law-making to, for instance, the appointment of European Commissioners (though the fact that these must be approved by the EP acts as a check, as we recently saw with the rejection of Hungarian nominee Laszlo Trocsanyi in September 2019).


  • Arendt, Hannah. 1951. The burden of our time [later title: The origins of totalitarianism]. London: Secker & Warburg.

    Google Scholar 

  • Athanassiou, Phoebus. 2009. Withdrawal and expulsion from the EU and EMU: Some reflections. ECB Legal Working Paper 10: 3–48. Accessed 22 Sept 2021.

  • Bárd, Petra. 2018. The rule of law and academic freedom or the lack of it in Hungary. European Political Science 19: 1–10.

    Google Scholar 

  • Beckman, Ludvig. 2009. The frontiers of democracy: The right to vote and its limits. Cham: Springer.

    Google Scholar 

  • Beckman, Ludvig. 2014. The subjects of collectively binding decisions: Democratic inclusion and extraterritorial law. Ratio Juris 27 (2): 252–270.

    Google Scholar 

  • Bellamy, Richard. 2019. A republican Europe of states: Cosmopolitanism, intergovernmentalism and democracy in the EU. Oxford: Oxford University Press.

    Google Scholar 

  • Bellamy, Richard, and Sandra Kröger. 2021. Countering democratic backsliding by EU Member States: Constitutional pluralism and ‘value’ differentiated integration. Swiss Political Science Review (early view).

  • Bennett, Christopher. 2016. Penal disenfranchisement. Criminal Law and Philosophy 10 (3): 411–425.

    Google Scholar 

  • Blauberger, Michael, and R. Daniel Kelemen. 2017. Can courts rescue national democracy? Judicial safeguards against democratic backsliding in the EU. Journal of European Public Policy 24 (3): 321–336.

    Google Scholar 

  • Chamon, Merijn. 2020. Re-establishing the EU: Dissolution, withdrawal or succession. EU Law Live 32: 2–6.

    Google Scholar 

  • Chamon, Merijn, and Tom Theuns. 2021. Resisting membership fatalism: Dissociation through enhanced cooperation or collective withdrawal. Verfassungsblog. Accessed 11 Oct 2021.

  • Cholbi, Michael J. 2002. A Felon’s right to vote. Law and Philosophy 21 (4/5): 543–565.

    Google Scholar 

  • Christiano, Thomas. 2008. The constitution of equality. Oxford: Oxford University Press.

    Google Scholar 

  • Ewald, Alec. 2002. Civil death: The ideological paradox of criminal disenfranchisement law in the United States. Wisconsin Law Review 5: 1045–1132.

    Google Scholar 

  • Gora, Anna, and Pieter de Wilde. 2020. The essence of democratic backsliding in the European Union: Deliberation and rule of law. Journal of European Public Policy 1–21 (online first).

  • Greenberg, Mark. 2021. Legal interpretation. Stanford Encyclopedia of Philosophy. Accessed 11 Oct 2021.

  • Hart, Henry M., and Albert M. Sacks. 1994. The legal process: Basic problems in the making and application of law. Foundation Press.

    Google Scholar 

  • Hausteiner, Eva Marlene. 2020. Can federations expel member states? On the political theory of expulsion. Journal of International Political Theory 16 (1): 47–67.

    Google Scholar 

  • Hillion, Christophe. 2020. Poland and Hungary are withdrawing from the EU. Verfassungsblog. Accessed 22 Sept 2021.

  • Hintikka, Jakko. 1962. Cogito, ergo sum: Inference or performance? Philosophical Review 71 (1): 3–32.

    Google Scholar 

  • Hofmann, Herwig. 2021. Sealed, stamped and delivered the publication of the Polish Constitutional Court’s judgment on EU law primacy as notification of intent to withdraw under art. 50 TEU? Verfassungsblog. Accessed 20 Oct 2021.

  • Invernizzi Accetti, Carlo, and Ian Zuckerman. 2017. What’s wrong with militant democracy? Political Studies 65 (1): 182–199.

    Google Scholar 

  • Kelemen, R. Daniel. 2017. Europe’s other democratic deficit: National authoritarianism in Europe’s Democratic Union. Government and Opposition 52 (2): 211–238.

    Google Scholar 

  • Kirshner, Alexander. 2014. A theory of militant democracy: The ethics of combatting political extremism. New Haven: Yale University Press.

    Google Scholar 

  • Kochenov, Dimitry, and Laurent Pech. 2016. Better late than never? On the European Commission’s rule of law framework and its first activation. JCMS: Journal of Common Market Studies 54 (5): 1062–1074.

    Google Scholar 

  • Kochenov, Dimitry. 2017. Busting the myths nuclear: A commentary on article 7 TEU, EUI Working Paper No. LAW 2017/10. Accessed 22 Sept 2021.

  • Kovács, Kriszta, and Kim Lane Scheppele. 2018. The fragility of an independent judiciary: Lessons from Hungary and Poland–and the European Union. Communist and Post-Communist Studies 51 (3): 189–200.

    Google Scholar 

  • Loewenstein, Karl. 1937. Militant democracy and fundamental rights, I. American Political Science Review 31 (3): 417–432.

    Google Scholar 

  • López-Guerra, Claudio. 2014. Democracy and disenfranchisement: The morality of electoral exclusions. Oxford: Oxford University Press.

    Google Scholar 

  • Manfredi, Christopher P. 1998. Judicial review and criminal disenfranchisement in the United States and Canada. The Review of Politics 60 (2): 277–306.

    Google Scholar 

  • Mos, Martijn. 2020. Ambiguity and interpretive politics in the crisis of European values: Evidence from Hungary. East European Politics 36 (2): 267–287.

    Google Scholar 

  • Müller, Jan-Werner. 2015. Should the EU protect democracy and the rule of law inside member states? European Law Journal 21 (2): 141–160.

    Google Scholar 

  • Nicolaïdis, Kalypso. 2013. European democracy and its crisis. Journal of Common Market Studies 51 (2): 351–369.

    Google Scholar 

  • Niklewicz, Konrad. 2017. Safeguarding the rule of law within the EU: Lessons from the Polish experience. European View 16 (2): 281–291.

    Google Scholar 

  • Oliver, Peter, and Justine Stefanelli. 2016. Strengthening the rule of law in the EU: The Council’s inaction. JCMS: Journal of Common Market Studies 54 (5): 1075–1084.

    Google Scholar 

  • Olsen, Tore Vincents, and Christian F. Rostbøll. 2017. Why withdrawal from the European Union is undemocratic. International Theory 9 (3): 436–465.

    Google Scholar 

  • Olsen, Tore Vincents. 2019. Liberal democratic sanctions in the EU. In Militant democracy and its critics: Populism, parties, extremism, eds. Anthoula Malkopoulou and Alexander Kirshner. Edinburgh: Edinburgh University Press.

    Google Scholar 

  • Patberg, Markus. 2019. Can disintegration be democratic? The European Union between legitimate change and regression. Political Studies 68 (3): 582–599.

    Google Scholar 

  • Patberg, Markus. 2021. The democratic ambivalence of EU disintegration: A mapping of costs and benefits. Swiss Political Science Review (early view).

  • Pech, Laurent, and Kim Lane Scheppele. 2017. Illiberalism within: Rule of law backsliding in the EU. Cambridge Yearbook of European Legal Studies 19: 3–47.

    Google Scholar 

  • Poama, Andrei, and Tom Theuns. 2019. Making offenders vote: Democratic expressivism and compulsory criminal voting. American Political Science Review 113 (3): 796–809.

    Google Scholar 

  • Posner, Richard A. 1985. The federal courts: Crisis and reform. Cambridge, MA: Harvard University Press.

    Google Scholar 

  • Reiman, Jeffrey. 2005. Liberal and republican arguments against the disenfranchisement of Felons. Criminal Justice Ethics 24 (1): 3–18.

    Google Scholar 

  • Sadurski, Wojceich. 2010. Adding bite to a bark: The story of article 7, EU enlargement, and Jorg Haider. Columbia Journal of European Law 16 (3): 385–426.

    Google Scholar 

  • Schall, Jason. 2006. The consistency of felon disenfranchisement with citizenship theory. Harvard Black Letter Law Journal 22: 53–93.

    Google Scholar 

  • Scheppele, Kim Lane, Dimitry Kochenov, and Barbara Grabowska-Moroz. 2021. EU values are law, after all: Enforcing EU values through systemic infringement actions by the European Commission and the Member States of the European Union. Yearbook of European Law 39: 3–121.

    Google Scholar 

  • Scherz, Antoinette. 2021. How should the EU respond to democratic backsliding? A normative assessment of expulsion and suspension of voting rights from the perspective of multilateral democracy [Paper Presentation]. ECPR early career workshop on democratic backsliding. Rome, Italy.

  • Scholtes, Julian. 2020. On doctrinal contortions and legal fetishes: A sceptical response to Christophe Hillion, Verfassungsblog. Accessed 22 Sept 2021.

  • Sedelmeier, Ulrich. 2017. Political safeguards against democratic backsliding in the EU: The limits of material sanctions and the scope of social pressure. Journal of European Public Policy 24 (3): 337–351.

    Google Scholar 

  • Sigler, Mary. 2014. Defensible disenfranchisement. Iowa Law Review 99: 1725–1744.

    Google Scholar 

  • Theuns, Tom. 2017. Promoting democracy through economic conditionality in the ENP: A normative critique. Journal of European Integration 39 (3): 287–302.

    Google Scholar 

  • Theuns, Tom. 2019. The legitimacy of free trade agreements as tools of EU democracy promotion. Cambridge Review of International Affairs 32 (1): 3–21.

    Google Scholar 

  • Theuns, Tom. 2020. Containing populism at the cost of democracy? Political vs. economic responses to democratic backsliding in the EU. Global Justice: Theory, Practice, Rhetoric 12 (2): 141–160.

    Google Scholar 

  • Theuns, Tom. 2021. Pluralist democracy and non-ideal democratic legitimacy: Against functional and global solutions to the boundary problem in democratic theory. Democratic Theory 8 (1): 23–49.

    Google Scholar 

  • van den Brink, Martijn, and Dimitry Kochenov. 2019. Against Associate EU citizenship. JCMS Journal of Common Market Studies 57 (6): 1366–1382.

    Google Scholar 

  • von Bogdandy, Armin, and Luke Dimitrios Spieker. 2019. Countering the judicial silencing of critics: Article 2 TEU values, reverse Solange, and the responsibilities of National Judges. European Constitutional Law Review 15 (3): 391–426.

    Google Scholar 

  • Zala, Miklos, Simon Rippon, Tom Theuns, Sem de Maagt, and Bert van den Brink. 2020. From political philosophy to messy empirical reality. In Justice and vulnerability in Europe, ed. Trudie Knijn and Dorota Lepianka. Cheltenham Glos: Edward Elgar Publishing.

Download references


This article has been a fairly long time in the making and contains ideas salvaged from abandoned papers as well as things written initially for other papers (especially Theuns 2020). As such, my list of debts is long. I would like to thank the participants of the GJN/TPR/MWP Global Justice and Populism workshop held at the EUI in June 2019, those at Injustice and Illegitimacy in Europe panel of the 2019 CES conference, the 2021 ECPR Political Theory Seminar and the 2021 ECPR Joint Sessions workshop on militant democracy. I would particularly like to thank Gabriele Badano, Ludvig Beckman, Jan Pieter Beetz, Jelena Belic, Richard Bellamy, Angela Bourne, Dimitrios Efthymiou, Franca Feisel, Maurits de Jongh, Nikolas Kirby, Sandra Kröger, Glyn Morgan, Alasia Nuti, Tore Vincents Olsen, Andrei Poama, Andrea Sangiovanni, Antoinette Scherz, Nicole Scicluna, Martijn van den Brink, Frank Vandenbroucke, Philippe Van Parijs, Fabio Wolkenstein and Miklos Zala, as well as the editors and anonymous reviewers at Res Publica. Few of them endorse my position here—mistakes that remain are made in spite of their generous interventions. I also gratefully acknowledge support from the NWO Veni grant VI.Veni.201R.061 since February 2021.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Tom Theuns.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Theuns, T. The Need for an EU Expulsion Mechanism: Democratic Backsliding and the Failure of Article 7. Res Publica (2022).

Download citation

  • Accepted:

  • Published:

  • DOI:


  • Article 7
  • Democratic backsliding
  • Democratic theory
  • European Union
  • Expulsion
  • Legitimacy
  • Militant democracy
  • Rule of law