Non-ideal circumstances and the duties of citizens and civil servants to protect democracy

Non-ideal theory concerns what to do when there is a lack of compliance with the values, principles, and norms that our ideals prescribe. Backward-looking non-ideal theory concerns itself with ‘past’ non-compliance, for example, questions regarding punishment and reparation for historical injustices. Forward-looking non-ideal theory concerns itself with the improvement of ongoing non-compliance (Chahboun Reference Chahboun, Hänel and Müller2024). The present article is mostly forward-looking. It is, therefore, naturally oriented towards the consequences of actions and their effectiveness.

The following argument about the rights and duties of citizens and civil servants under different degrees of (non)compliance rests on 12 assumptions (1)–(12) that, for reasons of space, cannot be fully justified here. They can be grouped into five sets (I–V):

1. I. (1) The democratic ideal can be appropriately interpreted in terms of the principles and norms of the democratic constitutional state, the main purpose of which is to protect and develop the rights that simultaneously secure citizens’ private and public autonomy (Habermas Reference Habermas1996). (2) Non-democratic organisations and parties are defined by their pursuit of political agendas that conflict with this understanding of the democratic ideal regardless of whether they are conscious of and/or explicit about the conflict.

2. II. (3) Democratic backsliding is understood as a movement backwards in relation to the realisation of public and private autonomy through democratic rights. Democratic backsliding means that democratic rights are either obstructed or abrogated (Wolkenstein Reference Wolkenstein2023). Democratic backsliding is seen as a process within democracy at the end point of which democracy ceases to exist. It forms part of a wider process of autocratisation (Lührmann and Lindberg Reference Lührmann and Lindberg2019). (4) Democratic backsliding and autocratisation processes are often incremental and predominantly result from legal reforms of policies and institutions (Lührmann and Lindberg Reference Lührmann and Lindberg2019; Pirro and Stanley Reference Pirro and Stanley2022).

3. III. (5) There is a critical point where democratic institutions have become so undermined that it is not possible to restore democracy via democratic means, in particular elections (Rijpkema Reference Rijpkema2018, Reference Rijpkema, Malkopoulou and Kirshner2019). This point of no democratic return to democracy (PNDRD) is the point where legitimate government power, understood as the right to rule, ceases to exist. (6) It is also the point where citizens’ right to democratic self-government is taken away from them and (7) the point at which citizens no longer have a general obligation to follow the law; that is, it is where political obligation dissolves (Peter Reference Peter and Nodelman2023).

4. IV. (8) Citizens are permitted to defend their own rights, depending on how well public institutions for the protection of rights function. (9) In their defence, citizens may sometimes violate the rights of others (Coons and Weber Reference Coons and Weber2016). (10) People are entitled to defend their right to democratic self-government with coercive means regardless of whether the would-be usurper is foreign or domestic (Frowe Reference Frowe2014a).

5. V. (11) Citizens may also have a natural duty to defend democratic rights – a duty whose exact nature depends on the costs involved both in terms of possible repercussions and the potential futility of the efforts (Rawls Reference Rawls1999, pp. 293–294). (12) Civil servants have both a stronger obligation to follow the law and a stronger role-specific responsibility to defend rights than do ordinary citizens (Valentini Reference Valentini2021).

We can distinguish between ideal circumstances on the one hand and circumstances, which are non-ideal to different degrees, on the other. Schematically,

• • Under ideal circumstances, all citizens comply with democratic principles and norms.

• • Under minimally non-ideal circumstances, some citizens in civil society, including some organised in civil society organisations and parties, are at least ambiguous in their commitment to those principles. They create the concern that they will not be fully committed to these principles should they gain power.

• • Under moderately non-ideal circumstances, non-democratic parties have gained power and initiated a process of democratic backsliding, which partially undermines democratic institutions and leads to a circumscription of citizens’ democratic rights. However, the threshold between a democratic and a non-democratic regime has not yet been passed. The regime therefore still has political legitimacy – that is, the right to rule; and citizens have a prima facie general obligation to obey the law – that is, a political obligation.

• • Under critically non-ideal circumstances, non-democratic parties in government have transgressed this threshold (PNDRD), or there is an imminent danger that they will do so. This has removed or will remove citizens’ democratic rights and their right to democratic self-government. It will also make the government non-legitimate, ie without the right to rule.

• • Under extreme non-ideal circumstances, the regime has turned into an autocracy without any (democratic) legitimacy.

These different conditions can be illustrated by different categorisations in democracy indices as depicted in Figure 1.

Figure 1.

Ideal and non-ideal circumstances.

Note: The figure is an illustration of different levels of non-ideal circumstances and not intended as an exact categorisation of the regimes in the various countries. The divergent categorisations by the two indices (reproduced from Economist 2024 and Nord et al. Reference Nord, Lundstedt, Altman, Angiolillo, Borella, Fernandes, Gastaldi, Ana, Natsika and Lindberg2024) emphasize the difficulty of setting precise thresholds for when a regime ceases to be a democracy.

Figure 1. Long description

The table has five rows and six columns. The columns are labeled Circumstances, Political institutions, Country, Democracy Index (Economist 2023), Liberal Democracy Index (V-Dem 2023), and Party. The rows are grouped into different levels of democratic circumstances: Ideal, Minimally non-ideal, Moderately non-ideal, Critically non-ideal, and Extremely non-ideal. Row 1: Ideal, Democracy,, , , . Row 2: Minimally non-ideal, Democracy, Germany, 8.80 Full democracy, 0.81 Liberal Democracy, AfD. Row 3: Minimally non-ideal, Democracy, Italy, 7.69 Flawed democracy, 0.76 Liberal Democracy, Lega. Row 4: Moderately non-ideal, Democratic backsliding, Poland, 7.18 Flawed democracy, 0.44 Electoral Democracy, PiS. Row 5: Critically non-ideal, Democratic backsliding, Hungary, 6.72 Flawed democracy, 0.32 Electoral Autocracy, Fidesz. Row 6: Extremely non-ideal, Autocracy, Russia, 2.22 Authoritarian, 0.06 Electoral Autocracy, Putin. Row 7: Extremely non-ideal, Tyranny, North Korea, 1.08 Authoritarian, 0.01 Closed Autocracy, Kim.

Unpacking the implications of the assumptions set out above, the following sections will discuss citizens’ rights and duties under different circumstances, going from ideal to critically non-ideal. They leave out extreme non-ideal circumstances since, in a certain sense, they delimit the scope of a theory of the defence of democracy and instead point towards a theory of revolution.

The role of citizens and civil servants under ideal and minimally non-ideal circumstances

The protection of citizens’ rights is the collective responsibility of all citizens. However, civil servants, who constitute the state and maintain the legal order, arguably have a stronger responsibility than other citizens (cf. Habermas Reference Habermas1996 ch. 1). Moreover, some civil servants have an even stronger responsibility than others, such as the police, public prosecutors, and court judges, who are tasked with ensuring that the system as a whole protects the rights of citizens.

Civil servants are expected to fulfil their responsibility for realising the democratic rights of citizens on the basis of valid laws and to perform their respective functions with transparency, truthfulness, and professionalism unaffected by social and political pressure. They are expected to be loyal to the law, which is produced according to the proper procedures and consistent with the principles and values of their country’s constitution. Their work both reflects the collective autonomy of citizens and is based on the need to secure the continuous exercise of this autonomy (Bauer Reference Bauer2024; Yesilkagit, Bauer, Peters et al. Reference Yesilkagit, Bauer, Peters and Pierre2024). However, what should citizens and civil servants do when they are faced with different degrees of non-compliance with democratic principles? How should they react, and what are the normative constraints on their reactions?

The first challenge is the creation of non-democratic organisations and parties aiming to take control of society and its institutions. Populist parties are one case in point (cf. Pappas Reference Pappas2016). Increasing support in Germany for Alternative für Deutschland (AfD), a right-wing populist party, is a concrete example.

Public authorities have the main responsibility for the protection of citizens’ rights against any immediate violations by such organisations. Indeed, in some national contexts like Germany, the protection of democratic rights is reinforced by measures of militant democracy. Public authorities are to intervene if there is an imminent danger that non-democratic organisations will undermine or topple democratic institutions (Kirshner Reference Kirshner2014; Müller Reference Müller2016; Thiel Reference Thiel2016; Tyulkina Reference Tyulkina2015; Vinx Reference Vinx2020).

Without a right to use force and with no current violation of or imminent threat to their democratic rights, citizens would have to respect the rights of their political opponents. They therefore must identify those means that can reasonably be presumed to be most effective without violating the normative constraints set by democratic rights themselves. Ideally, they would try to convince their opponents and their opponents’ supporters that they are wrong, using the ‘force of the better argument’ (Habermas Reference Habermas1993). More realistically, they can employ their rights strategically as power resources to reduce the organisational capacity and appeal of their opponents.

Civil servants, meanwhile, are expected to remain neutral unless their specific task is to monitor society for potential future threats to democracy, such as the security services. Citizens can direct the attention of the security services to non-democratic actors in society. It is also commonly accepted that policing entails making connections and collaborating with stakeholders in society. Nonetheless, there are limits to how organised such collaboration can be without turning civil servants into political actors. The major worry is that the state apparatus could be used by some political actors against other political actors who are in fact legitimate. Indeed, one of the complaints against militant democracy provisions and institutional mechanisms is that they can be abused by those in power to oppress opponents and keep them out of the democratic arena (Invernizzi Accetti and Zuckerman Reference Invernizzi Accetti and Zuckerman2017).

As mentioned, the presence and growth of AfD in Germany exemplifies minimally non-ideal circumstances. The AfD’s rhetoric and policies indicate an uncertain commitment to the principles and norms of democracy, including individual rights (Heitmeyer Reference Heitmeyer2024). So far, the combined efforts of civil society, mainstream political parties, and public authorities have not resulted in diminished support for the AfD. AfD gained 21% of the national vote in the February 2025 election, and in recent polls (April 2026) was the largest party in Germany with 26% of the vote (Politico 2026).

The popularity of AfD could indicate that responses to it have not been effective. Moreover, we might observe that some of Germany’s established mainstream parties have been influenced by the AfD’s success and have co-opted some of its rhetoric and policies (cf. Abou-Chadi and Krause Reference Abou-Chadi and Krause2020). However, that is not the whole story. First, a cordon sanitaire against the AfD has generally been upheld; without it, the AfD would have achieved much more influence on policy (Schroeder, Ziblatt and Bochert Reference Schroeder, Ziblatt and Bochert2025). Second, AfD’s official categorisation as an extreme right party by the security services is bound to have negative effects on Germans’ perception of the AfD as a normal party (cf. Zajak, Meuth and Best Reference Zajak, Meuth and Best2025). Third, it is unclear that public protests against the party do not work against normalisation in general.

In addition, the effectiveness of initiatives is not measured only in terms of electoral support. Laumond’s contribution to this special issue reveals how initiatives by individual citizens that have been backed by elected officials and, to some extent, civil servants have had negative effects on the organisational resources and capacities of the AfD. Citizens have used their property rights in a coordinated manner to exclude the AfD from holding meetings in private venues while maximally exploiting existing regulations to bar them from holding meetings in public facilities. There are signs that these types of citizens’ initiatives are copied and repeated across different territories, creating an even greater impact (Laumond this issue). Such efforts reduce the organisational resources of the AfD and negatively affect the party’s ability to recruit new members. Similarly, church organisations’ exclusion of the AfD from the German Kirchentage since 2019 has prevented the AfD from participating in mainstream political events, which could have legitimised them as a normal party and given them a stage before a wider audience.

The fact that initiatives against the AfD have not stopped them from reaching a high level of popular support is a problem from the perspective of improving non-ideal circumstances as quickly as possible. The German case clearly indicates the need for more knowledge on how to effectively counteract such high levels of popular support for an authoritarian populist party. The Polish case and especially the Hungarian case demonstrate the risk of handing over government power to such parties. Still, under minimally non-ideal circumstances, where all citizens’ rights remain intact, civil society actors are not under a strong requirement to ensure that their responses are effective in that regard. Indeed, rights imply the freedom to make wrong or ineffective choices (Waldron Reference Waldron1981). Actions that do not violate any rights do not have to be effective in the same manner as those that do.

In sum, under minimally non-ideal circumstances, citizens have a right, and arguably a duty, to mobilise in order to minimise the influence and growth of non-democratic parties. Their activities must remain within the constraints of all people’s democratic rights and be as effective as possible. In this context, citizens have good reasons to build alliances both with other citizens in civil society and with public authorities (civil servants) to pursue this goal. Nonetheless, to avoid vigilantism and to maintain the rule of law, the main responsibility for protecting citizens’ rights lies with public authorities. This implies monitoring problematic organisations and parties and can, to some extent, involve collaboration with private citizens. That said, the boundaries between public authorities and private actors should be kept distinct to secure against abuse, whereby some parties are systematically excluded to the benefit of others.

Moderately non-ideal circumstances: when sliding between legitimate rule and autocracy

We now turn to moderately non-ideal circumstances. In these situations, some degree of democratic backsliding has taken place. However, it is not clear that it has progressed so far that the government has no right to rule, and therefore also not clear that civil servants and citizens do not have an obligation to follow the law. An example of this is Poland under the government of the Law and Justice Party (PiS) (2015–2023). Generally, under such circumstances, citizens may experience little responsiveness to or accommodation of their criticisms. Consequently, they may quickly have to give up on convincing their opponents to change course. Instead, they may turn to forms of political action that are more strategic and demonstrative to pressure their opponents, this time in ways that may involve some violation of rights. Applying such pressure becomes more appropriate the less the government is open to genuine dialogue. A classic example is civil disobedience campaigns combining strategic pressure with dialogue (King Jr. Reference King1992). Civil disobedience has conventionally been regarded as appropriate in ‘nearly just societies’ (Rawls Reference Rawls1999, p. 309). Here citizens have a right, and again arguably a (natural) duty, to improve on justice and democracy by drawing attention to unjust and undemocratic laws by temporarily breaking the law in public, non-violent ways while accepting the legal consequences (Scheuerman Reference Scheuerman2018). Court cases can bring further attention to the principles at stake, namely the violations of rights and freedoms. Accepting legal consequences, should they fall to the activists’ disfavour, also implies that citizens compensate for their transgression of the law post hoc. It emphasises that their interest is not in undermining the law but in improving it in terms of democracy and justice (Scheuerman Reference Scheuerman2018).

Civil disobedience involves violating the rights of others. However, it is primarily a democratic means in the sense that its purpose is to change opponents’ minds rather than force them to change their course of action (Scheuerman Reference Scheuerman2018). Nonetheless, given that it is likely to involve some (moral) costs to the targets of civil obedience, to third parties (innocent bystanders), and to the civilly disobedient themselves, the presumption of the latter should be that civil disobedience and strategic forms of political action will be an effective means to restore democracy.

Civil servants have a special responsibility to protect citizens’ democratic rights, but are caught in a dilemma because of their strong responsibility, even under moderately critical non-ideal circumstances, to follow the law. To live up to their responsibility to protect democracy while observing the law, civil servants would do well to maintain contacts or build alliances with vigilant citizens who critically monitor the government to detect risks of further backsliding. They may also facilitate certain citizens’ initiatives, for example by showing leniency towards civilly disobedient activists. Police, public prosecutors, and judges ought to consider that civilly disobedient citizens are not ‘common criminals’ but are acting with the public interest in mind. Arguably, this ought to have some implications for how harshly police handle disobedient citizens, for the kind and size of the legal consequences that public prosecutors seek, and for the verdicts that judges render (Dworkin Reference Dworkin2013; Habermas Reference Habermas1985).

Under less-than-ideal circumstances, civil servants tasked with ensuring that the system as a whole respects democratic rights should publicly and persistently criticise attempts to undermine them (cf. Yesilkagit, Bauer, Peters et al. Reference Yesilkagit, Bauer, Peters and Pierre2024). However, members of the public administration, more widely conceived, also need to defend democratic and constitutional order. The civil servants should see themselves as ‘the servant and guardian of legal and professional rules and a constitutional order, not of the rulers’. At some point, this involves refusing to follow orders and working to obstruct reforms that dismantle rights and institutional accountability.

When refusing to follow the law and the demands of political leadership, it is increasingly important that the actions undertaken by citizens and bureaucrats are effective, not only to maintain and restore democracy, but also to avoid violating legal expectations and the rights of other citizens more than necessary.

Once the process of democratic backsliding gains momentum and pushes institutions closer to the PNDRD, the demand to respect formal legal procedures and (legal) directives from political leadership, as well as the democratic rights of members of the governing populist party, becomes less strict. Civil servants might have to deny service and obedience to the incumbent government to be able to serve the democratic order. Citizens are generally justified in responding in kind to the violation of their democratic rights (Schedler Reference Schedler2021). However, when they do, they should act in ways that they presume will be effective in achieving their goals while assuming democratic responsibility for not escalating the conflict unnecessarily (Kirshner Reference Kirshner, Malkopoulou and Kirshner2019; Kirshner Reference Kirshner2014; Schedler Reference Schedler2021).

Thus, under moderately non-ideal circumstances – the grey area between minimal democratic backsliding and the PNDRD – the duties of citizens and civil servants to respect the rights of those wanting to obstruct or abrogate their rights become less strict. As the PNDRD is approached, the need to respect the rights of opponents who are actively violating one’s democratic rights is progressively diminished, while the requirement to be effective becomes increasingly strict.

In Poland under the Law and Justice (PiS) government (2015–2023), there were elements of democratic backsliding, manifested in government control over public media, systematic legal prosecution and libel lawsuits deployed against journalists, political favouritism of PiS loyalists in the recruitment of civil servants, and the pursuit of illiberal – or at least conservative – interpretations of women’s rights. Similarly, at the level of municipalities, there were campaigns against LGBTQ people. Equally important, the PiS government systematically worked to bring courts and judges under political control, thereby undermining the rule of law. However, despite all of this, Poland had not unequivocally passed the PNDRD (Nord Reference Nord, Lundstedt, Altman, Angiolillo, Borella, Fernandes, Gastaldi, Ana, Natsika and Lindberg2024, p. 17). This also would seem confirmed by the 2023 election in favour of PiS’s opposition.

In her contribution to this special issue, Moroska-Bonkiewicz (Reference Moroska-Bonkiewicz2026) analyses how the many events organised by civil society actors in conjunction with opposition parties and civil servants at both the national and the European levels slowly undermined the moral resources and the legitimacy of the incumbent government. This negatively affected the latter’s attempts to restrict abortion rights and the rights to protest and to bring the courts under government control. There was a high degree of cooperation among judges, lawyers, and civil society activists. When the government wanted to prohibit counterprotests to government rallies, protesters resorted to civil disobedience. They cleverly exploited the chance that courts would pass verdicts in favour of their protests and against government regulations for violating basic democratic rights. Similarly, lawyers initiated lawsuits against appointment procedures for new judges to the Supreme Court. The cases highlighted government violations of central principles and norms of liberal democracy. Their combined efforts also brought the EU into the process. The CJEU passed a verdict to the effect that the Polish government had to change legislation to secure the independence of the Polish courts (Moroska-Bonkiewicz Reference Moroska-Bonkiewicz2026).

The activities of Polish oppositional forces in society and as part of the political and legal system were fully within the scope of legitimate forms of action. Their initial protests were fully in line with democratic rights. Later they resorted to civil disobedience, in part to create the conditions for making the case in a courtroom that the government was violating key principles of the Polish constitution. The civility that characterised their protests allowed them to maintain the moral high ground and may well have been conducive to undermining the legitimacy of the PiS government. According to Moroska-Bonkiewicz (Reference Moroska-Bonkiewicz2026), the effectiveness of their actions was not due to single events but resulted from the linkage of a myriad of events across space and time. They not only repeatedly exposed the government’s violation of fundamental principles, but they also motivated subsequent actions to oust the populist government and reverse its illiberal policies. In the end, the actions of civil society, activist judges, and the verdict issued by the CJEU provided a basis for the opposition parties to make the 2023 election about ‘saving democracy’, the constitution, and the rule of law.

Arguably, the Polish case is one of effective and therefore also legitimate action to restore democracy. However, it also points to the difficulties facing activists working against populist governments under conditions of democratic backsliding. These difficulties consist in determining both proportionate responses to the current level of government backsliding and which course of action is effective, since not every act of resistance is guaranteed to be. These difficulties are further emphasised by the Hungarian case and the responses to the Fidesz populist regime.

Critically non-ideal circumstances: the point of democratic return to democracy

Among the three cases studied in this special issue, the Orban regime (2010–2026) comes closest to being a case of critical non-ideal circumstances. The Fidesz party controlled the legislature, the courts, the executive, and the media, and worked assiduously to secure its own re-election through gerrymandering and election rules. After winning its first election in 2010, the Fidesz party effectively placed loyal people in key positions in legal and political institutions. It also exploited its pivotal role in the European People’s Party to delay EU action against its illiberal reforms (Herman, Hoerner and Lacey Reference Herman, Hoerner and Lacey2021). Many observers thought that the country had passed the PNDRD (cf. Figure 1 and Enyedi and Mikola Reference Enyedi and Mikola2024).

Critically non-ideal circumstances arise if democratic backsliding has brought the political community close to the PNDRD and there is an immediate risk that it will be pushed beyond this point, or indeed if it has already been passed. In this situation, the pressure for citizens to apply effective measures is heightened.

Critically non-ideal circumstances also change the role of civil servants. Past the threshold of the PNDRD, the preconditions for legitimate government and, with them, the right to rule disappear. The government then de facto turns into ‘a band of robbers’ applying coercion against other citizens, who would be justified in responding in kind (Peter Reference Peter and Nodelman2023; Schedler Reference Schedler2021).

Depending on both (1) the kind and degree of attack on their rights and (2) the quality of the political and legal institutions, the actions of citizens can justifiably go beyond the constraints of democratic rights (Frowe Reference Frowe2014b). The quality of the political and legal institutions makes a difference in relation to whether citizens have recourse to public authorities for the protection and vindication of their rights. Often, the kind and degree of attack on rights mirrors the quality of the political and legal institutions. The point of distinguishing between the two is that one part of the state, typically the executive power, may be violating the rights of citizens, while other parts, typically courts, maintain sufficient independence and power to protect them against violations. In such circumstances, citizens do not need to defend their own rights (see Arato and Cohen (Reference Arato and Cohen2021) on the distinction between populists being ‘in’ government and being the government).

Where citizens do have to defend their own rights against attacks from the government because they have no effective remedy elsewhere, the effectiveness of their measures is crucial for their justification (Frowe Reference Frowe2014b; Statman Reference Statman2008). According to the ethics of self-defence, people whose important rights are under attack and who have no recourse to public authorities or others for their protection are justified in protecting themselves with responses that violate the rights of their attackers. This applies, for example, to citizens whose democratic rights are on the verge of disappearing. However, their responses must be necessary and proportionate to the violation of their own rights and take into consideration the costs to innocent bystanders (Coons and Weber Reference Coons and Weber2016; Frowe Reference Frowe2014b).

Responses that do not violate anyone’s rights should be favoured whenever they are effective in this regard, and this implies that rights-violating actions would not be necessary. However, if such responses are not effective, then the situation calls for rights-violating actions that are. This means that those carrying out these actions must have a justified presumption that they will be effective. If they are not, they cannot be judged necessary, proportionate, or in line with the requirement of inflicting the lesser evil on third parties (Frowe Reference Frowe2014a).

Thus, under critically non-ideal circumstances, citizens would be justified in (pre-emptively) defending their democratic rights with actions that are presumed effective (Alexander and Ferzan Reference Alexander and Ferzan2011; Olsen Reference Olsen2022). Given the moral costs involved in rights-violating actions, the requirement that responses are presumed effective becomes stricter than with actions that respect the rights of their targets.

Of course, one objection might be that the ethics of self-defence only pertain to ‘vital interests’ and that democratic rights do not meet this criterion. It is therefore not justified to violate the rights of others in defence of democratic rights. There is some truth to this. Democratic rights are not on par with, say, the right to physical integrity. On the other hand, it is commonly assumed that citizens can use violent means to defend their democratic self-government even against a non-violent usurper (Frowe Reference Frowe2014b).

A further objection could be that democratic backsliding is not analogous to having one’s government taken over by a foreign power. Populist governments are supported by a democratic majority, and any democratic backsliding that they initiate is therefore democratically legitimated. It reflects the majority’s aggregated use of their individual democratic rights and expresses the people’s self-government. This argument is not without weight, but it ultimately fails. It rests on the faulty premise that democratic rights entail the (Hohfeldian) power to remove democratic rights from other citizens as well as oneself (cf. Vinx Reference Vinx2020).

A third objection is that it is wrong to demand that citizens’ responses to the violation of their rights should be effective. Citizens should be allowed to respond (in kind) to violations of their status as free and equal members of society (Frowe Reference Frowe2014b; Statman Reference Statman2008). This objection must be partly granted. However, it pertains to backward-looking non-ideal theory and to a different normative interest than the forward-looking one pursued here. The two interests can be combined by giving priority to effective responses and only entertaining (futile) retaliation when it is not possible to improve the situation.

Turning to civil servants, their roles and responsibilities become unclear when governments remove the effective protection of democratic rights. When the preconditions for producing valid laws no longer obtain, there is no valid law to follow for civil servants (cf. Michelman Reference Michelman1987).

One possible corollary of this is that there are no civil servants anymore because there is no valid law to specify their tasks. It mirrors the loss of political legitimacy. A competing interpretation is that civil servants are tied to the preexisting law and to defending the democratic constitutional state from internal and external attacks. Against this background, judges would be right to resist attempts by the legislature and the executive to bring the judiciary under their control.

Under critical non-ideal circumstances, citizens and civil servants are justified in joining forces to restore and protect democracy, even with the use of public resources. The effectiveness of their measures is important, especially if they violate fundamental political and civil rights (eg using physical force).

Considerations of effectiveness might require civil servants to remain in place and to make a division of labour between themselves and citizens. This implies trying to isolate their ‘own institutions’ from government attempts to control them. An even more complicated situation is created if the PNDRD has not been passed but is still only an imminent risk. The stronger demand on civil servants to stay loyal to the law could be an argument in favour of them assuming a less active role, leaving the main initiative to citizens in civil society. On the other hand, the stronger role-specific responsibility of civil servants to protect democracy, along with their interest as democratic citizens to protect their democratic rights, would trump their duty to stay loyal to the law.

Under critical non-ideal circumstances, citizens and civil servants would eventually be justified in defending their democratic rights and attempting to restore democratic institutions even with means that do not respect the democratic rights of their opponents. Indeed, they may have a duty to do so. Civil servants may in fact have a stronger duty than ordinary citizens in this regard. Part of this common duty is to make sure that the means employed are effective. A lack of effectiveness could eventually nullify the duty if its exercise is futile and it implies unacceptable costs to agents in terms of personal sacrifice and regime repercussions. This implies that citizens and civil servants ought to build alliances with each other (and other actors) that can make their respective efforts more effective and thereby improve the conditions for exercising their duty.

As mentioned, the Orban regime (2010–2026) was a case of critical non-ideal circumstances. Many observers had stopped classifying it as a democracy (cf. Figure 1 and Enyedi and Mikola Reference Enyedi and Mikola2024), effectively claiming that it had passed the PNDRD. Even if the opposition’s victory in the April 2026 election arguably relativises this conclusion, passing the PNDRDFootnote ¹ would have opened a different set of responses to the regime, less civil and more coercive than the ones employed in Poland. However, despite the tangible circumscription of their democratic rights, Hungarian citizens and civil servants rarely resorted to activities that went beyond civil disobedience. Absent potential alliance partners inside institutions, responses to Fidesz mainly came from parties and non-governmental organisations (NGOs) in conjunction with EU institutions and international organisations. Before the rise of the Tisza Party, winner of the 2026 election, oppositional political parties’ cooperation against Fidesz was without any clear success (Tyszka Reference Tyszka and Bourne2024). The same was true of civil disobedience campaigns, at least if their success is evaluated on the basis of immediate goals (Tyszka and Olsen Reference Tyszka and Olsen2026).

The combined studies of Poland and Hungary demonstrate the difficulty of identifying universally effective responses to populist parties in government. The analysis of anti-populists in Poland points to the effectiveness of staging a multiplicity of events linked together by alliances of citizens and civil servants, stretching over space and time in a manner that sustains momentum for further protest action. Responses were mainly non-violent and initially predominantly within the limits of the law. Only later were civil disobedience campaigns employed to expose the government’s violation of constitutional principles.

The Polish case also reveals the importance of alliances across the civil society and state divide, where civil society organisations, professional organisations, and civil servants work together to resist developments and expose the violation of constitutional norms. In comparison, Hungary long looked like the negative case. Considering the recent election results and the benefit of hindsight, this assessment is now more complicated. Still, it is reasonable to claim that the conditions for action for both government and opponents were in fact significantly different in the two countries. In Poland there was much more interaction between people outside the system and between citizens and civil servants than in Hungary. Hungarian civil society also seemed to have been less successful at bringing in the EU as an external actor. Moreover, government capabilities and strategies in the two countries were different. Due to a more centralised control of state institutions, the Fidesz government was able to react more resolutely against opposition initiatives than the PiS government. Anti-populists were not facing identical opponents in terms of capacity.

An important lesson from Poland and Hungary is thus that oppositional actors need to prepare themselves for early intervention when populists with shallow commitments to democracy gain power. They should create alliances between civil society actors, political parties, and groups inside the state as quickly as possible, as well as reach out to outside organisations such as the EU.

Legitimate responses to populist parties

As suggested, effectiveness plays an essential role in justifying responses to populist parties in and out of government, at least in a forward-looking non-ideal theory concerned with reducing the risks to democratic rights and institutions or counteracting democratic backsliding and processes of autocratisation. However, it does not play the same role in all circumstances. Under minimally non-ideal circumstances, efforts should be effective, but this demand is not strict if citizens and civil servants act within their rights (and those of their opponents). As the AfD case demonstrates, applying rights strategically can lead to the effective exclusion of authoritarian populist parties from civil society forums and to the obstruction of their activities and expansion without violating anyone’s rights. This takes place alongside other activities, such as public protests, which are not necessarily effective in any direct and tangible manner but can be instrumental for activists’ morale and mutual confirmation that they are involved in a collective endeavour, boosting and maintaining mobilisation. Effectiveness becomes more important when actions violate the rights of others. This applies, for example, to moderately non-ideal circumstances where actions of civil disobedience are used as a last resort when rights-respecting actions have been exhausted, and to critically non-ideal circumstances where citizens are fighting for their right to democratic self-government. The effectiveness requirement places a constraint on using coercive measures, in particular those involving violence. Violent measures are likely to be regarded as extreme and unjustified by potential supporters and alliance partners (cf. Chenoweth Reference Chenoweth2021). We might have seen this in the context of Poland, where activists seem to have been conscious of not being perceived as extreme. Extremism would deter people from joining the cause, both those who serve in state organisations and ordinary citizens. It would also make it much more difficult to build alliances with civil servants and professional organisations inside the state and at the European level. Non-violent protests are more likely to be effective, and they make it more difficult for populist-led governments to frame the opposition as illegitimate and anti-democratic.

The priority of effectiveness does not mean that each action should be effective in and of itself. First, the sheer presence of protest may moderate populist governments in terms of how far they will go. This is the issue concerning the counterfactual. Even if protest measures and similar tactics are not immediately effective, they may prevent the government from implementing even more illiberal policies and institutional reforms. Second, the Polish case reveals that effectiveness may arise as the aggregate effect of a series of minor events and responses that together make populist governments delay illiberal policies and roll back certain problematic reform agendas.

Third, as the Hungarian case suggests, there are some cases in which actions by citizens – in this instance, civil disobedience – do not lead to any of the immediately intended results, but rather the opposite. While this may discourage citizens from undertaking such actions – indeed, on the face of it, people should refrain from undertaking such actions in light of the effective remedies requirement – it should be noted that they may have positive longer-term effects through the boosting of morale and public support. The defeat of Fidesz in the 2026 election is a likely example of this.