EPP protecting Fidesz

Ever since it formed its first majority government in 2010, Viktor Orbán’s Fidesz party has carried out numerous constitutional amendments. These have been deliberately designed to ‘weaken institutions that might have checked’ the government’s plan ‘to impose upon Hungary a wholly new constitutional order using only ideas and votes from Fidesz’ (Bánkuti et al. Reference Bánkuti, Halmai and Scheppele2012: 139; see also Bozóki and Hegedüs Reference Bozóki and Hegedüs2018; Krekó and Enyedi Reference Krekó and Enyedi2018; Müller Reference Müller2013). Most notably, the Fidesz government attacked the Hungarian Constitutional Court, altering the system for appointing constitutional judges. The new system allowed Fidesz to unilaterally nominate judges and let its parliamentary majority elect those judges to the court without requiring support by any other party.

In further steps, the Fidesz government brought the Electoral Commission under its political control. This effectively gave the party the ability to decide on proposals for referenda (and oversee election monitoring). Fidesz also restructured the Media Authority, appointing a former Fidesz MP to a nine-year term as head of that authority. By law, the same person has also been made the chair of the newly created Media Board, which can levy hefty fines on media outlets that fail to achieve ‘balanced’ news coverage. In addition, the Hungarian government banned ‘political advertising during the election campaign in any venue other than in the public broadcast media, which is controlled by the all-Fidesz media board’ (Scheppele Reference Scheppele2013). Furthermore, it greatly expanded the number of so-called Cardinal Laws – laws that can only be changed by a two-thirds majority – so as to limit the capacity of any future government to change the basic rules of public finances, public service provisions, the pension system and so on.

These initiatives were widely regarded as conflicting with the values enshrined in Article 2 TEU, in particular democracy and the rule of law. Like all other EU member states, Hungary formally committed itself to these values by ratifying the treaty. This led the European Commission to bring legal challenges against some of the Fidesz government’s actions. Reacting to this, the leaders of the EPP, of which Fidesz is a member, immediately came to Fidesz’s defence and at least up until the vote of 12 September 2018, where a majority of EPP MEPs voted to finally initiate the Article 7 procedure against Hungary, the party blocked robust EU intervention. In this way, it hampered the efforts of EU institutions to hold the Orbán regime accountable (Kelemen Reference Kelemen2017: 220; Kirchik Reference Kirchik2013).

The EPP’s efforts to obstruct the imposition of sanctions on the Hungarian government have already been documented by scholars (see Introduction). It is still worth mentioning two key examples of obstruction. First, in 2013 the European Parliament’s Committee on Civil Liberties, Justice and Home Affairs issued a report criticizing the erosion of fundamental rights in Hungary – the so-called ‘Tavares Report’ (European Parliament 2013). This report was voted down by a majority of EPP MEPs and EPP vice-chair Manfred Weber publicly dismissed the report as a ‘wish list of the European leftist parties who aim to impose their own political agenda on Hungary’ (EPP Group 2013). Second, in 2014, the Barroso Commission established the Rule of Law Framework, in part as a response to its frustration that the EU lacked adequate tools to counter democratic backsliding. The idea behind the framework was to allow the EU to increase pressure on governments that systematically undermine democracy and the rule of law by issuing a series of warnings prior to triggering Article 7. In 2015, then, when the European Parliament passed a resolution calling on the Commission to launch the Rule of Law Framework procedure against Hungary, only parties on the political left voted in favour of the resolution. The EPP voted against it and its leadership again publicly defended the Orbán government (Kelemen Reference Kelemen2017: 226). In the end, the Juncker Commission decided against launching the procedure.

Recent research has also shown that EPP MEPs were generally ‘less likely to emphasise the issue’ of democratic backsliding in Hungary in the European Parliament, relative to the MEPs of other party groups (Meijers and van der Veer Reference Meijers and van der Veer2019). Whether this can plausibly be interpreted as a way of protecting the Hungarian government from sanctions can only be determined through a close study of the EPP MEPs’ actions, but it neatly fits the picture that the EPP generally supports the Orbán regime. That said, the EPP is internally divided about the Orbán regime, and in recent times these divisions have become increasingly visible.

In 2018, the party’s systematic support for the Hungarian government became an issue of contention in the internal primary that was held to determine who would be the party’s Spitzenkandidat for the 2019 European Parliament election. Especially after the ouster of the Central European University, more and more EPP members grew uncomfortable with Orbán’s attacks on the Union’s core values; some went so far as to demand Fidesz’s expulsion from the EPP. In response, the aspiring Spitzenkandidat Manfred Weber announced that he would endorse the ‘Sargentini Report’ that called for the triggering of Article 7 against Hungary (European Parliament 2018). At the same time, however, Weber and other EPP leaders declared that they would not eject Fidesz from the EPP, even though the Sargentini Report eventually passed. Instead, they ‘suspended’ Fidesz’s membership of the EPP, though the suspension ‘did not apply to … Fidesz MEPs in the European Parliament, and … their seats would still count towards the EPP’s tally’ (Kelemen Reference Kelemen2020: 488).

In April 2020, the leaders of thirteen EPP-member parties signed a letter demanding that Orbán’s party should be ejected from the group (signatories include the Swedish Moderaterna’s Ulf Kristersson, Rutger Ploum from the Dutch Christen-Democratisch Appèl and Kyriakos Mitsotakis of Greece’s Nea Dimokratia). Meanwhile, Orbán is rallying his few unambiguously illiberal allies within the EPP around their shared nationalist-conservative agenda. These are most notably Janez Janša, the Slovenian prime minister and leader of the Slovenian Democratic Party, and Serbian President Aleksandar Vučić, whose Serbian Progressive Party is an ‘associate member’ of the EPP (Bayer Reference Bayer2020). This suggests that the EPP is increasingly polarised about its relationship with Fidesz.

In short, today the EPP is divided about how to deal with Orbán’s party, and even top-level EPP officials are becoming more hesitant about overtly defending the Hungarian government. But this is only a recent development. For the larger part of the post-2010 period, the EPP has systematically protected its Hungarian member party from possible EU-level sanctions. The following statement by the EPP’s former president, Joseph Daul, encapsulates the EPP’s relationship with the Hungarian government: ‘Viktor Orbán … is the “enfant terrible” of the EPP family, but I like him’ (cited in Barbière Reference Barbière2015).

ECR refusing to interfere with PiS

Poland’s democratic backsliding began later than Hungary’s, namely in October 2015, when the ultra-conservative, nationalist PiS-party won a plurality (just under 38 per cent) of the vote, which translated into an absolute majority (51 per cent) of seats in the Polish parliament. Shortly after assuming office, the new government attacked the independence of the country’s high court, the Constitutional Tribunal. The assault on the Constitutional Tribunal caused a constitutional crisis, as the PiS-affiliated President Andrzej Duda, together with the PiS-controlled parliament, attempted to pack the Tribunal with judges who were loyal to the party, passing legislation aimed at weakening the Tribunal’s position (Sadurski Reference Sadurski, Graber, Levinson and Tushnet2018). The Tribunal reacted to this in March 2016 by striking down the reforms intended by the PiS, declaring them unconstitutional. Yet the PiS-government refused to recognise the judgment, maintaining that the court had no authority (for critical discussion, see Venice Commission 2016: 16).

The newly elected government also tried to assert control over public media. New legislation was passed (and signed into law by PiS-loyalist President Duda) depriving the independent Public Broadcasting Council of its authority over public media, instead endowing the treasury minister with the right to decide over the management of public television and radio broadcasters. As these new rules were adopted in January 2016, both the managers and supervisory board members of the country’s public broadcasters were fired, and the PiS immediately tried to replace them with party loyalists (The Guardian 2016). All these initiatives closely resembled the script Fidesz deployed in Hungary, which is no coincidence. Jarosław Kaczyński, the ultra-conservative co-founder of PiS and former prime minister of Poland – who holds no formal role in government but is widely thought to ‘pull the strings’ in Warsaw – has repeatedly expressed admiration for Orbán and emphasised that the Fidesz government is an example to follow (Buckley and Foy Reference Buckley and Foy2016).

Compared with its handling of Hungary’s democratic backsliding, the EU’s response to the illiberal developments in Poland has been quite robust. Reacting to the government’s attacks on the independence of the country’s high court, the Constitutional Tribunal, the European Commission triggered Article 7 in September 2017, just two years after the PiS-government was elected. This was an unprecedented move, indicative of the extent to which the PiS’s ‘judicial reforms’ were seen as a threat to the rule of law. In addition, the European Commission referred Poland to the European Court of Justice for allowing, among other things, ordinary court judges to be subjected to disciplinary investigations depending on the content of their decisions (European Commission v Republic of Poland, C-791/19). Poland has also been found by the Court of Justice to have failed to fulfil its obligations under the second sub-paragraph of Article 19(1) TEU, as well as to have failed to fulfil its obligations under the second and third paragraphs of Article 267 TFEU (European Commission v Republic of Poland, C-791/19; see also Opinion of Advocate General Tanchev in European Commission v Republic of Poland, C‑619/18, ECLI:EU:C:2019:325; Pech and Wachowiec Reference Pech and Wachowiec2020).

Now, the Polish governing party PiS is the largest member of the ECR, a much younger and smaller European party than the EPP. Founded in 2009 and formerly known as Alliance of European Conservatives and Reformists (AECR) (2009–16) and Alliance of Conservatives and Reformists in Europe (ACRE) (2016–19), its membership comprises a quite heterogenous group of parties, although the general ideological trajectory is conservative and EU-sceptic. Until the 2019 European Parliament elections, it included such ideologically different parties as the British Conservatives (who withdrew from the group in January 2020 following the completion of the UK’s withdrawal from the EU), the left-leaning Christian Union from the Netherlands (which joined the EPP in 2019) and the right-wing ‘populist’ Danish People’s Party (which joined the newly founded ‘Identity and Democracy’ group in 2019). Notably, several current ECR members have been classified as ‘right-wing populist,’ ‘radical right’ or ‘authoritarian populist’ in the relevant social-scientific literature, in particular the Sweden Democrats, the relatively new Spanish Vox party and indeed the PiS itself (e.g. Akkerman et al. Reference Akkerman, de Lange and Rooduijn2016; Norris and Inglehart Reference Norris and Inglehart2019: 235; Turnbull-Dugarte et al. Reference Turnbull-Dugarte, Rama and Santana2020; Zulianello Reference Zulianello2020).

How did the ECR react to the ‘reforms’ that its member PiS conducted in Poland after 2015, and the reaction of EU bodies? The first thing to note is that the ECR, by virtue of being a relatively small party, did not have the power to protect the Polish government from EU sanctions in the same way that the EPP could protect the Hungarian government. The ECR indeed has only ‘marginal influence on law-making in the European Parliament’, so it simply lacked the means to exercise a great deal of influence on the relevant parliamentary votes (Kelemen Reference Kelemen2017: 230). Most of its MEPs voted against motions condemning the Polish government (Meijers and van der Veer Reference Meijers and van der Veer2019), but this was not particularly effective in holding off sanctions.

More relevant than the near-absence of effective protective behaviour, I suggest, is the fact that the ECR has not taken any action against the PiS – despite the fact that European institutions (and many legal experts) found that the Polish government’s behaviour constituted a violation of Article 2 values and other treaty-based obligations (Alemanno and Pech Reference Alemanno and Pech2019). Some evidence suggests that this can be at least partly explained by ECR members’ ideological conviction that EU member states should have the right to decide on both their constitutional architecture and ordinary laws without having to heed the EU’s rules and recommendations. For example, when the former Conservative British Prime Minister Theresa May visited Warsaw in December 2017, she emphasised, regarding the PiS-government’s constitutional reforms, that ‘constitutional issues are normally, and should be primarily a matter for the individual country concerned’ (Reuters 2017a). Similarly, reacting to the aforementioned Polish media law, Anders Vistisen, MEP of the Danish People’s Party, remarked that ‘even if this would not be okay in Denmark, it may well be okay in Poland … We [in the ECR group] all agree that the EU should have less influence and nation states more, and so we also have to respect what nation states decide to do’ (Journalisten.dk 2016). At any rate, the point to note is that the ECR took a permissive approach vis-à-vis PiS.

Identifying wrongdoing

Turning now to the concept of complicity, the first thing that must be identified is a particular instance of wrongdoing. For without assuming wrongdoing, it would make little sense to speak of complicity in the first place. That complicity presupposes wrongdoing appears in virtually all discussions of complicity in legal and political theory (e.g. Gardner Reference Gardner2007; Kutz Reference Kutz2000; Lepora and Goodin Reference Lepora and Goodin2015), and it is neatly captured in the Oxford English Dictionary’s definition of complicity as ‘the fact or condition of being involved with others in an activity that is unlawful or morally wrong’.

Importantly, as this latter definition highlights, the wrongdoing through which complicit contributions become, loosely speaking, ‘problematic’ can be a transgression of moral norms or positive laws. Inasmuch as the facticity of law itself has both positive and normative dimensions, however, it would be misleading to keep morality and law entirely separate (for extensive discussion of this point, see Habermas Reference Habermas1992: 45–60, 90–108, 135–51; Habermas Reference Habermas1996: 296–300). As Jürgen Habermas (Reference Habermas1996: 297) puts it, law must be conceived as a ‘functional complement’ (funktionale Ergänzung) to morality that ‘unburdens’ (entlastet) its subjects from the cognitive and motivational challenges of individually passing moral judgements concerning all of their actions: it specifies the general norms that one might individually judge to be morally required (cognitive unburdening), and incentivises rule-following through the threat of coercion (motivational unburdening). This relationship between positive law and morality implies, however, that the legitimacy of law derives from it being ‘consonant with the everyday norms of justice, proportionality, the moral value of human dignity, equal respect for persons, and so on’ (Finlayson Reference Finlayson2019: 85). In other words, when specific laws fail to track morality qua ‘everyday norms of justice’, they will lack legitimacy.

That morality and positive law are connected in this way also means that the political theory literature on complicity can usefully inform our understanding of the phenomenon at stake. The larger part of that literature operates with a non-legal understanding of complicity – that is, it speaks of complicity in moral wrongdoing rather than complicity in breaking the law. This is partly because theorists seek to apply the concept of complicity to cases where there is relatively little positive law to hold on to, such as genocide in so-called ‘failed states’ (e.g. Kutz Reference Kutz2000; the important contribution of Lepora and Goodin Reference Lepora and Goodin2015 is also in many ways about complicity under circumstances where the rule of law cannot be relied on). By making explicit and clarifying more or less widespread intuitions about why complicit contributions to wrongdoing are themselves ‘wrong’, they enhance our understanding of when and why complicit secondary agents can be subject to moral blame, and when legal consequences could potentially be appropriate. This strategy is especially helpful, I suggest, when a contributory action appears intuitively wrong, but what makes it wrong is not entirely obvious – as in the case of the contributions of European political parties to democratic backsliding.

Turning now to the question asked at the outset of this sub-section: What is the particular wrongdoing in our two cases? The wrongdoing is, I suggest, democratic backsliding qua the ‘state-led debilitation or elimination of any of the political institutions that sustain an existing democracy’ (Bermeo Reference Bermeo2016: 5). In the cases with which I am concerned here (see Section II), democratic backsliding qualifies as wrongdoing in virtue of being a violation of particular treaty-based obligations. As already noted, Hungary and Poland violate in particular Article 2 TEU, which establishes that the EU is ‘founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law and respect for human rights’. On what may be called the ‘standard view’ that is defended by leading legal scholars, member states and EU institutions (including the European Court of Justice), these values are ‘absolutely indispensable for the operation of the Union’ (Klamert and Kochenov Reference Klamert, Kochenov, Kellerbauer, Klamert and Tomkin2019: 26). They constitute ‘an objective of the Union, and … a cardinal aim of its institutional framework’ (Hillion Reference Hillion, Closa and Kochenov2016: 62). As such, explains Christophe Hillion (Reference Hillion, Closa and Kochenov2016: 62–63, italics added),

This means, first and foremost, that EU member states cannot simply organise their national judiciaries as they see fit, hiding behind appeals to ‘national sovereignty’ or ‘the will of the people’ (Burchardt Reference Burchardt2019: 96; Halmai Reference Halmai2017; Scheppele and Pech Reference Scheppele and Pech2018b). Nor can they reorganise their democratic institutions in such a way that oppositional forces are systemically constrained – for instance, by new constitutional provisions that commit single-majority governments to pursue particular substantial political aims against their will (Möllers and Schneider Reference Möllers2018: 133–34; Venice Commission 2011: 6–7). These are just some illustrative examples of the implications of the just-mentioned treaty-based obligations.

Of course, the argument from treaty-based obligations is not the only possible argument for why democratic backsliding is a ‘wrong’. It would indeed be odd to say that undermining democracy and the rule of law is wrong only because it constitutes a violation of particular treaty-based commitments that are specific to the EU, as this would imply that actions such as the partisan hijacking of constitutional courts are unproblematic and permissible in countries that did not sign up to the values of Article 2 TEU (such as non-EU countries). More plausibly, democratic backsliding is a ‘wrong’ because it violates values that possess validity and meaning beyond the specific context of the EU. This argument need not be pursued further at this point, however, since Article 2 already gives expression to these norms and thus makes the self-standing normative argument for why democratic backsliding is a wrong unnecessary. (In Section IV, I will return to the important question of whether the standard interpretation of Article 2 values is arbitrary.)

Unpacking complicity

Having identified what makes democratic backsliding a form of wrongdoing, let us now concentrate on the concept of complicity. As already indicated, complicity is essentially about contributing as a ‘secondary agent’ to the unlawful or immoral activity of a ‘primary wrongdoer’. This is less than committing unlawful or immoral acts jointly together with the primary wrongdoer, but more than being a mere bystander who holds no responsibility for the wrongdoing in question – indeed, in most legal orders as well as on most normative accounts of complicity, a complicit secondary agent can be held responsible inasmuch as that agent has made a causal or potentially causal contribution to the primary wrongdoing. I say causal or potentially causal because a complicit contribution can be against the law or morally wrong even if it ultimately turns out that it does not contribute to the success of the primary wrongdoing (e.g. Gardner Reference Gardner2007: 72–75; Lepora and Goodin Reference Lepora and Goodin2015: 61–62; Zakaras Reference Zakaras, Sobel, Vallentyne and Wall2018: 197). Because both law and morality are supposed to guide our actions (see Habermas’s argument mentioned above), we must take into account what consequences might reasonably have been expected at the time of the contributory action when assessing the immorality or illegality of a contributory action, rather than focus just on what actually happened in the end.

Chiara Lepora and Robert E. Goodin (Reference Lepora and Goodin2015) helpfully further unpack the concept of complicity. Specifically, they distinguish a variety of more specific kinds of complicit behaviour by making explicit the moral intuitions implicit in a range of ‘conceptual cousins’. Among these, complicity by connivance and complicity by condoning are the most relevant for our purposes. These differ in two respects from the generic notion of complicity that I have just introduced. On the one hand, neither connivance nor condoning always or necessarily involve a causal or potentially causal contribution to the primary wrongdoing. But when they do involve a causal or potentially causal contribution, they qualify as forms of complicity. On the other hand, they make more specific assumptions about the behaviour of the secondary agent that will be useful for the below application of concepts. This is why I introduce these more specific notions to begin with.

Consider first complicity by connivance, which the Oxford English Dictionary usefully defines as ‘willingness to allow or be secretly involved in an immoral or illegal act’. Connivance so conceived involves such things as shutting one’s eyes to another’s wrongdoing or tacitly assenting to it, perhaps out of sympathy or approval. It does not involve directly partaking in planning wrongdoing. Connivers need not even approve of the wrongdoing, or know any of its details. Nor do they need to do any more than refuse or fail to act in a particular way that would prevent or limit wrongdoing. Indeed, even ‘doing nothing’ can causally contribute to wrongdoing, thus qualifying as a form of complicity: ‘If there was something you could have done to stop [a wrongdoing] and you didn’t, your inaction can properly be counted as a part of the causal chain that allowed the event to occur’ (Lepora and Goodin Reference Lepora and Goodin2015: 45). To be clear, connivance thus conceived presupposes that that secondary agent is actually capable of preventing or limiting wrongdoing; it requires a position of de facto or de jure power.Footnote ¹ This is not the case in complicity by condoning.

Complicity by condoning, then, refers to an agent accepting and pardoning the morally or legally wrong behaviour of another – an action that is separate from the principal wrong committed by the principal wrongdoer (Lepora and Goodin Reference Lepora and Goodin2015: 47). In contrast to the conniver, who simply remains silent in relation to a particular wrongdoing, the condoning agent directly acknowledges the wrongdoing and explicitly pardons it. This can even happen ahead of the wrong occurring – say when an actor announces that they would in principle excuse another actor’s committing a particular wrong before that actor has actually committed the wrong. In doing so, an actor can contribute causally to a wrong by ensuring the future wrongdoer that their acts will not be blamed or sanctioned by at least one other actor. This means that the causal contribution to wrongdoing consists in the encouraging effect of the secondary agent’s insurances. Again, complicity by condoning does not presuppose that the complicit agent is equipped with special de facto or de jure power. But arguably the assurances of the condoner will have a greater encouraging effect on the wrongdoer if the condoner has the capacity to, say, block sanctions.

The EPP’s complicit contribution

It is easy to see how the EPP’s actions qualify as instances of complicity. First, the perhaps most substantial contributions the EPP has made to democratic backsliding in Hungary, namely that it has voted down the Tavares report in 2013 and voted against the resolution to launch the Rule of Law Framework procedure in 2015, may be classed as complicity in the generic sense of a causal (or potentially causal contribution) to the primary wrongdoing of democratic backsliding. As many have argued, the EPP’s opposition to taking action against the Hungarian government indeed had a causal and not merely potentially causal effect on democratic backsliding in Hungary, for it effectively blocked the way for holding the Orbán regime to account (e.g. Bugarič Reference Bugarič, Closa and Kochenov2016: 90; Kelemen Reference Kelemen2017; Kelemen Reference Kelemen2020). That Article 7 was eventually triggered against Hungary in September 2018 confirms that this judgement was correct in terms of counter-factual causality: initiating the sanction procedure was possible only because a majority within the EPP suddenly endorsed it.

Second, complicity by condoning adequately describes many of the EPP leaders’ reactions to the behaviour of the Hungarian government. Recall Weber’s comment that the Tavares report is but a ‘wish list of the European leftist parties who aim to impose their own political agenda on Hungary’ (EPP Group 2013; this was not the only time that Weber made statements to this effect – see Mounk Reference Mounk2018), or Joseph Daul’s cordial remark that the Hungarian prime minister is ‘the “enfant terrible” of the EPP family, but I like him’ (cited in Barbière Reference Barbière2015). Now, it is difficult to prove that such speech-acts encouraged the Hungarian government to continue with their wrongdoing, but it is hardly controversial that they can make a potentially causal contribution, in the sense that it can reasonably be expected that the Hungarian government will be encouraged to carry on with its attacks on democracy and the rule of law as long as EPP leaders publicly signal their sympathy for its political project and are willing to dismiss criticisms as politically motivated. It can reasonably be expected not least because the EPP has demonstrated that it can causally influence collective decisions about possible sanctions, as explained in the previous paragraph. This adds weight to such assurances of support.

Third, EPP leaders not only ‘offer occasional words of support’ to Orbán ‘that help him maintain power domestically’, but also conspicuously ‘turn a blind eye to his misdeeds’, as in complicity by connivance (Kelemen Reference Kelemen2017: 226–27). EPP group-chair Manfred Weber even verbalises this attitude of ‘shutting one’s eyes’ to the Hungarian government’s wrongdoing:

These statements undoubtedly also contain an element of condoning, but at their heart they are about pretending away the attacks on democracy and the rule of law for which Orbán’s government is responsible. Importantly, at least before September 2018, even the EPP-internal opponents of the Orbán regime connived at Hungary’s democratic backsliding. Indeed, as commentators have noted, even the EPP delegations that have repeatedly ‘expressed disquiet or outright hostility at having Orbán’s party in their midst’ mostly adopted ‘a wait-and-see approach’ (De la Baume and Bayer Reference De la Baume and Bayer2018).

Concerning the EPP’s leadership, it seems clear that their connivance had a causal effect on democratic backsliding, in the sense that they could have done something to stop it (e.g. exercised significant pressure on Orbán and threatened him with sanctions) but didn’t. So their connivance no doubt qualifies as complicity by connivance. Concerning the group of Orbán opponents within the EPP, however, things are less clear-cut. If complicity by connivance entails, at a minimum, that a secondary agent refuses or fails to act in a particular way that would prevent or limit wrongdoing, this implies that that secondary agent is actually capable of preventing or limiting wrongdoing. In sharp contrast to the EPP leaders, though, the Orbán opponents hardly wielded enough power within the party. To be sure, they could have resorted to more radical strategies and left the EPP, founding a new centre-right party that was committed to upholding democratic values. But it is doubtful whether this would have had any preventive or limiting effect on democratic backsliding in Hungary. Most plausibly, then, their connivance does not qualify as complicity by connivance.

The ECR’s complicit contribution

On the face of it, the generic notion of complicity would seem to apply to the ECR’s MEPs’ voting against motions condemning the Polish government. However, it is doubtful whether this may be classified as a causal contribution to wrongdoing. As already noted, and unlike the large and powerful EPP, the ECR group was simply too small to block sanctions against Poland (see Kelemen Reference Kelemen2017: 230). For this reason, the ECR’s MEPs’ voting against motions condemning the Polish government could not even have been potentially causal. Much more relevant for our analysis of complicity in democratic backsliding are the ECR’s acts of connivance and condoning. These acts have at the very least had a potentially causal effect on the Polish government’s wrongdoing.

As far as connivance goes, we saw that the ECR mainly refused to act in a way that would prevent or limit the PiS’s wrongdoing at the member state level (Alemanno and Pech Reference Alemanno and Pech2019). Besides not taking actions against their large Polish member party, ECR MEPs also remained more passive in parliament when it came to discussing democratic backsliding in Poland, asking fewer (critical) questions than other party groups (Meijers and van der Veer Reference Meijers and van der Veer2019: 849). Now, it is a real question what exactly the ECR could have done to effectively prevent democratic backsliding in Poland. Sceptics might argue that sanctioning their member party PiS may not have had any notable effect, pointing out that the semi-suspension of Fidesz’s membership by the EPP did not stop the Orbán government’s attacks on democracy and the rule of law. But it would be spurious to conclude that, just because the EPP’s half-hearted suspension of Fidesz – which came after several years of systematic support – appears to have been ineffective, analogous interventions by the ECR would necessarily have failed. It is indeed reasonable to assume that taking action to prevent democratic backsliding might have had some effect, while inaction undoubtedly can be counted as part of the broader causal chain that allowed backsliding to occur. This is because the Polish government has very few EU-level allies (Kelemen Reference Kelemen2017: 230); and those few allies could have exercised considerable pressure on it had they made it aware that it could no longer count on their support if it didn’t change its behaviour.

Representatives of (former) ECR member parties, in particular the British Conservatives and the Danish People’s Party, also condoned the actions of the Polish government. As we saw, officials of those parties suggested that EU member states may legitimately reform their constitutions (or media laws) as they wished, even if they did not necessarily endorse the developments in Poland themselves. In the case of the British Conservatives, it was indeed the prime minister of a major member state who made remarks to that effect (Reuters 2017a). This, I suggest, may have contributed at least in a potentially causal way to democratic backsliding in Poland. For it not only indicated to the Polish government that the ECR would likely remain inactive, but also that the head of state of another member state might be willing to block the Article 7 procedure in the Council and/or oppose other sanctions that the EU at that point in time (i.e. 2017) planned to impose on Poland. Again, it is difficult to prove whether the British prime minister’s condoning statements actually had this effect. But if the relevant test is whether said consequences could reasonably have been expected at the time of the presumed contributory action, I think that at least this one instance of condoning qualifies as complicity by condoning.

To sum up, the EPP and ECR have made a range of secondary contributions to democratic backsliding in Hungary and Poland. Not all of these classify as complicity in the sense of causally or potentially causally contributing to the primary wrongdoing, but there are many instances where a causal or potentially causal contribution can be traced. Besides these relationships of complicity, there is a more complex web of complicit activity that merits analysis. Most notably, the Hungarian Prime Minister Viktor Orbán has repeatedly asserted that Hungary would make sure that the Article 7 procedure against Poland is killed latest when it reaches the Council (e.g. Reuters 2017b; Scheppele Reference Scheppele2016). In this way, Orbán and his government potentially become complicit secondary agents in democratic backsliding in Poland (as suggested indirectly by Scheppele Reference Scheppele2016) and, by supporting Orbán’s party Fidesz, the EPP may likewise assume the role of a secondary agent. Since examining these issues would take us too far away from the topic of the article, however, I will bracket them here.

Possibilities for sanctioning parties

In 2009, the Lisbon Treaty enshrined a ‘capacious set of democratic principles as the EU’s founding commitments’ (Isiksel Reference Isiksel2016: 131). The new treaty provisions attribute an especially important role to political parties at the European level, underlining their role in ‘expressing the will of the citizens of the Union’ (Article 10(4) TEU). More particularly, in EU Regulation No 1141/2014 of the European Parliament and Council of 22 October 2014 on the Statute and Funding of European Political Parties and European Political Foundations, it is emphasised that EU-level parties have a crucial role to play in ‘bridging the gap between the national level and at Union level’, and that they should be ‘encouraged and assisted in their endeavour to provide a strong link between European civil society and the Union institutions, in particular the European Parliament’.

Importantly, Article 3 of Regulation 1141/2014 also specifies ‘conditions for registration’, which must be met by a political alliance that seeks to register as a European political party. One of these conditions, laid out in Article 3(2c), is that a political alliance ‘must observe, in particular in its programme and its activities, the values on which the Union is founded, as expressed in Article 2 TEU, namely respect for human dignity, freedom, democracy, equality, the rule of law and human rights, including the rights of persons belonging to minorities’. The Regulation furthermore establishes, in Article 6, an ‘Authority for European political parties and European political foundations’ (henceforth ‘the Authority’) that is tasked with ‘registering, controlling and imposing sanctions on European political parties … in accordance with this Regulation’ (Article 6(1), emphasis added).

There thus exists an official body that has the authority to sanction European political parties. It can do so when they fail to comply with the registration conditions and requirements explicated, inter alia, in Article 3 of the Regulation. Article 10(3), amended in May 2018 by Regulation 2018/673, explains how this works:

If the Authority decides to de-register the party on the grounds of ‘a manifest and serious breach as regards compliance with the conditions’ set out in Article 3, this enters into force ‘only if no objection is expressed by the European Parliament and the Council within a period of three months of the communication of the decision to the European Parliament and the Council or if, before the expiry of that period, the European Parliament and the Council have informed the Authority that they will not object’ (Article 10(4)). Note that in 2019, a new rule 223a has been introduced, according to which a group of at least 50 citizens may submit to the President of the European Parliament a reasoned request inviting Parliament to request said compliance verification.

This, then, is the main sanctioning mechanism for European political parties that is currently available in the EU. It constitutes a form of what Ulrich Wagrandl (Reference Wagrandl2018: 145) calls ‘transnational democracy gone militant’, meaning an application of the ‘traditional notion of militant democracy to the transnational level’. To be sure, traditionally militant democracy is primarily about banning parties by law, and deregistering a European political party is not the same as outlawing it. But as Bourne and Casal Bértoa (Reference Bourne and Casal Bértoa2017: 225–30) have shown, militant democratic mechanisms can vary regarding the degree to which they exclude a targeted party from the public sphere. These range from dissolution to deregistration, to a withdrawal of certain more specific rights and privileges (Bourne and Casal Bértoa Reference Bourne and Casal Bértoa2017: 226). In the case of Regulation 1141/2014, militant-democratic measures involve only depriving a European political party of its right to formally exist as a party.

As with other rule of law enforcement and democracy protection mechanisms in the EU (Bugarič Reference Bugarič, Closa and Kochenov2016: 85–101; Müller Reference Müller2015: 147–49; Müller Reference Müller2016: 262–63), the procedure for deregistering a European political party is not exactly easy to see through. Further, the failure of recent attempts to set it in motion suggests that it might be quite a toothless instrument (Alemanno and Pech Reference Alemanno and Pech2019). Still, it is the primary available instrument for holding European political parties to account. With regard to lodging the compliance verification request on the basis of which further sanctions will be decided, I would argue that the above analysis of the EPP’s and ECR’s complicit contributions to democratic backsliding can considerably strengthen the case that the two parties’ activities were in conflict with the values laid down in Article 2 TEU. Treating the behaviour of the two parties as instances of complicity in democratic backsliding clarifies the sense in which their actions might be counted as non-compliance with Article 2 values, and explain why even potentially causal contributions can be classified as such.

To be clear, the suggestion I am making is that there is a strong presumptive case that the EPP’s and ECR’s complicit activities discussed in Section III amount to the non-observance of (some) Article 2 values. I say ‘presumptive case’ because these are matters about which the Authority needs to render a final decision (just as the argument that some of the actions of the Hungarian and Polish governments involve violations of Article 2 values is primarily a presumptive case for the violation of Article 2 values, about which the various institutions that are implicated in the Article 7 procedure must decide). As Alemanno and Pech (Reference Alemanno and Pech2019) have indicated, the case can be broken down to two issues. First, commitment to the fundamental value of democracy entails, at a minimum, that one does not encourage, tolerate or condone attempts to undermine democracy (even ones backed by democratic majorities). Second, commitment to the value of the rule of law entails, at a minimum, that one does not encourage, tolerate or condone attempts to undermine the rule of law. This might, of course, not be true on any thinkable definition of democracy or the rule of law, but I submit that it is the most plausible conclusion to draw from the understanding of democracy or the rule of law on which Article 2 TEU builds. I will return to this point in the next sub-section in connection with my discussion of whether this interpretation of democracy and the rule of law is arbitrary.

Before moving on, another issue needs handling. Assuming that the EPP’s and ECR’s complicit contributions to democratic backsliding actually constitute violations of Article 2 values, do we need to distinguish between less serious and more serious cases of complicity when making the case for sanctioning European political parties?Footnote ² It seems that we do not need to make such a distinction. For even if it seems plausible that some of the two parties’ complicit contributions are more serious than others – for instance, because of their differential causal impact on processes of democratic backsliding (Kelemen Reference Kelemen2017; Wolkenstein Reference Wolkenstein2020) – the sanctioning mechanism that I discussed a few paragraphs ago merely requires us to answer a binary question, namely whether or not a European political party observes (to quote the relevant passage again) ‘in its programme and its activities, the values on which the Union is founded, as expressed in Article 2 TEU’. In other words, to sanction European political parties, it need only be shown that such a party does not observe Article 2 values, but not whether some forms of non-observance are ‘graver’ than others.

Is sanctioning European political parties democratically defensible?

The simplest answer to this question is that it is indeed defensible, since EU member states have democratically decided to ‘bind themselves and follow EU rules’ (Müller Reference Müller2015: 144), which include Regulation 1141/2014. Yet militant democracy is a controversial doctrine. Many have argued that, even if militant democratic mechanisms were put into place by democratic means, dissolving or deregistering political parties is always democratically suspect (for an up-to-date overview treatment, see Müller Reference Müller2016: 251–53). The criticism is often framed in terms of a ‘paradox’: A democracy cannot defend itself by (ostensibly) anti-democratic means without undermining itself. Pioneering this line of criticism was Hans Kelsen (Reference Kelsen, Jestaedt and Lepsius1932: 237), who famously argued that, ‘If [democracy] remains true to itself, it has to tolerate also a movement that aims at the destruction of democracy.’ Taking this criticism (which has also been levelled at the EU’s mechanisms of rule of law enforcement and democratic defence – e.g. von Achenbach Reference von Achenbach2018) seriously, I want to discuss a more specific objection to the sanctioning of European political parties.

The objection on which I want to focus is a version of what, to my mind, is the most powerful argument that can be made against militant-democratic mechanisms: that any militant-democratic mechanism amounts to ‘a means for those empowered to make the relevant decisions to arbitrarily exclude an indeterminately expansive range of political competitors from the democratic game, thereby restricting the democratic nature of the regime’ (Invernizzi Accetti and Zuckerman Reference Invernizzi Accetti and Zuckerman2017: 183–84, emphasis added). The arbitrariness in all of this arises from the fact that the decision of who to exclude from the possibility of participating in democratic procedures amounts to a decision over the boundaries of the political community that, by definition, cannot be taken in a democratic fashion. On the face of it, this is a compelling argument that could easily be used – and has, in somewhat less systematic form, been used (Halmai Reference Halmai2017; Scheppele Reference Scheppele2019; Scheppele and Pech Reference Scheppele and Pech2018b) – by illiberal governments that seek to defend themselves against legal and political criticism and possible sanctions.

The first thing that must be noted in response is that, in the case of Regulation 1141/2014, it is hardly the case that those who are ‘empowered’ – this would here be the Authority – could exclude an ‘indeterminately expansive range of political competitors’. On the contrary, the Regulation specifies quite clearly what sort of behaviour may lead to deregistration, namely European political parties’ violation of the conditions laid down in Article 3(1) and (2). The fact that the Authority cannot decide unilaterally but must ask a committee of six ‘independent eminent persons’ for their opinion, and that the European Parliament and the Council have a right to object should the Authority eventually decide to deregister a European party, further reduces the discretion of the Authority to deploy sanctions in an ‘indeterminately expansive’ fashion. Thus, the Authority will not be able to use militant democracy ‘indiscriminately’.

Notice too that the deregistering of European political parties by no means entails an absolute ‘exclusion from the democratic game’. It is certainly true that deregistration would mean a big blow to any European party, seriously undercutting its organisational basis and reducing its capacity to make its voice heard in the European Parliament; however, this does not mean that the individual national parties that comprise the European party are disenfranchised in the sense of being once and for all excluded from parliament. Nor are they banned from establishing new European parties that can seek to obtain the status of being formally registered (for an analogous argument, see Wagrandl Reference Wagrandl2018: 150–51). So not only is the Authority not able to exclude an indeterminately expansive range of political actors from EU-level democracy, it also cannot exclude them wholly from the European ‘political community’.

These replies go some way towards reducing the force of the arbitrariness objection, but the core of the objection has not been addressed yet: the arbitrariness of the normative basis for deregistering European political parties, namely the values of Article 2 TEU. Now, what exactly is the objection here? It is that an interpretation of Article 2-values according to which the EPP and ECR presumptively violate those values – to recall: the argument was that commitment to the fundamental value of democracy entails that one must not encourage, tolerate or condone attempts to undermine democracy, and that commitment to the value of the rule of law entails that one must not encourage, tolerate or condone attempts to undermine the rule of law – is based on an arbitrary judgment by those ‘empowered.’ Put differently, the decision of what exactly constitutes a violation of Article 2 values ‘is necessarily an exceptional (i.e. ultimately political) decision, which cannot be subsumed into any prior [democratically established] norm, and must therefore be established arbitrarily by whoever has the power to enforce it’ (Invernizzi Accetti and Zuckerman Reference Invernizzi Accetti and Zuckerman2017: 186).

It is tempting to think that, if one were to look for a prior norm that could help determine the substance of Article 2 values, the Copenhagen Criteria for accession might be the most intuitive place to find them. Yet, as Dimitry Kochenov (Reference Kochenov2014: 166) convincingly argues, this is not a promising route, since ‘the Commission failed to shape the substance behind Article 2 TEU … in the pre-accession context’ (and this is not even to raise the difficult question of whether we could in this case plausibly speak of a democratically established norm). An alternative response would be a qualified embrace of the political and, as it were, ‘arbitrary’ nature of the decision. This is essentially what Jan-Werner Müller (Reference Müller2015) suggests in a widely-discussed article.

Müller’s argument is that the interpretation of Article 2 TEU, according to which commitment to democracy and the rule of law fundamentally involves remaining vigilant about and, if necessary, opposing and fighting attempts to subvert democracy and the rule of law has a normatively relevant historical pedigree. According to Müller, the meaning of the values laid out in Article 2, in particular the meaning of the values of democracy and the rule of law, is importantly shaped by the ‘Distrust of unrestrained popular sovereignty, and … unconstrained parliamentary sovereignty’ that was ‘in the very DNA of post-war European politics’ (Müller Reference Müller2015: 152; see also Müller Reference Müller2011). More specifically, distrust of unconstrained popular and parliamentary sovereignty gave rise to a conception of ‘constrained democracy’ (Müller Reference Müller2011) that from the 1950s onwards was widely endorsed on both the political left and right. Democracy and the rule of law were seen as mutually supporting each other, with the law placing constraints on democratic majorities in order to protect individual (and some collective) rights, as well as the integrity of the broader institutional architecture. This conception of democracy was, at its root, about restraining those actors that encourage, tolerate or condone the attempts of others to undermine democracy and the rule of law, rather than tolerating their actions as an expression of the ‘vagaries of democratic politics’.

Importantly, Müller’s argument does not boil down to the simplistic and naive claim that history has ‘decided’ on a single model of organising democracy, and that we should accept this decision. Rather, it settles for the more modest suggestion that there is a presumptive case that the correct meaning of democracy and the rule of law in Article 2 TEU can be gleaned from an analysis of how these concepts were understood in the (more recent) past, not least by those who constructed the EU. If one accepts that some level of arbitrariness is unavoidable in the interpretation of political values, one might indeed be content with this reasoning. However, it remains the case that Müller appeals primarily to the normative force of history, without supplying additional reasons for why we should think that a certain historically evolved understanding of democracy and the rule of law is worth preserving. For anyone who – for whatever reasons – remains unconvinced that ‘constrained democracy’ is desirable, or simply unwilling to defer to the forces of history alone, Müller’s argument will indeed seem spurious.

A stronger argument would highlight that the above-mentioned interpretation of Article 2 values is the one that is most consistent with the fundamental internal logic of modern constitutional democracy, according to which private and public autonomy mutually presuppose each other in theory and reciprocally reinforce each other in practice (Habermas Reference Habermas1992: 109–65; Habermas Reference Habermas1996: 301–3). Without basic norms that enshrine the equal moral worth of individuals, it is not possible to realise democratic self-legislation, let alone explain why democracy is valuable in the first place. From this, we cannot infer specific guidelines concerning how much power exactly constitutional courts should have vis-à-vis parliaments, or how exactly basic constitutional rights ought to be framed (see Habermas Reference Habermas1992: 151–65). What we can infer is that hijacking strong constitutional courts in the name of a (supposed) majority in order to limit both private (e.g. by limiting the freedoms of minorities) and public autonomy (e.g. by limiting the freedoms of the political opposition) – which, as Blokker (Reference Blokker2019) and Castillo-Ortiz (Reference Castillo-Ortiz2019: 56–71) amongst others have demonstrated, is essentially what the Hungarian and Polish governments have done – is at odds with democracy and the rule of law. And we can also safely infer that the EPP’s and ECR’s encouraging, tolerating or condoning of the Hungarian and Polish governments’ actions – in short, their complicity in democratic backsliding – amounts to a non-observance of the values of democracy and the rule of law.

Does this resolve the problem of arbitrariness? Have we now established that the above-advanced interpretation of the values specified in Article 2 is not arbitrary? If ‘not arbitrary’ means that it can be justified by taking recourse to a prior, democratically established norm, then the answer is no. The tradition of modern constitutional democracy with whose fundamental internal logic the standard interpretation of Article 2 coheres has its origin not in democratic procedures but in democratic revolutions as well as in the European post-war ‘emergency regime’ (White Reference White2015: 312–13) in which new democratic constitutions were written (Müller Reference Müller2011: 126–30, 146–50). Yet the eminently political arbitrariness at the root of this tradition might not be indefensible, at least if we accept the Habermasian proposition that the citizens of the past would have decided on a similar constitution of democracy to that which they eventually got, had they decided on these in an inclusive, democratic procedure (Habermas Reference Habermas1992). Ultimately, a level of arbitrariness is unavoidable when it comes to deciding on the correct interpretation of democracy and the rule of law, but I hope that the arguments advanced somewhat temper the force of the objection.