Introduction

How should we respond to populism? In recent years, a cottage industry of texts has emerged that seek to answer this question, with political theorists, political scientists and pundits alike putting forward potential ‘solutions’ to the supposed ‘problem’ of populism. Yet underlying many of these initiatives—consciously or not—is a conflation between populism and extremism, or populism and authoritarianism, with the well-established literature on combatting these more extreme forms of politics often informing the potential ‘solutions’ to the far more democratically ambiguous phenomenon of populism. Our argument in this article is that while such literatures might be historically informative, populism is not the same thing as these phenomena and as such, it requires different ‘solutions’ in terms of how to respond to it from the perspective of public authorities in liberal democratic contexts, which is our focus in this article. The conflation between these concepts, we argue, has important ramifications in terms of the democratic legitimacy of initiatives opposing populism—tolerant, intolerant or otherwise. This is particularly the case for populist parties as they emerge on the political scene.Footnote 1

We make this argument over three sections. First, we provide an overview of the ways in which populism is sometimes conflated with or portrayed as an automatic precursor to extremism, authoritarianism or totalitarianism in a significant amount of contemporary academic literature. We critique these conceptual conflations and assumptions, and outline the central differences between populism and these phenomena. In doing so, we consider populism’s liminal relationship with liberal democracy, and draw on the political theory literature that argues that populism sits ‘within’—not ‘outside’—democracy. Second, we turn to influential responses to populist parties, as laid out in the introductory article to this special issue by Bourne—the ‘intolerant’ (which we take to be synonymous with ‘militant democracy’, since the former is grounded on the same principles, justifications and practices as the latter) and the ‘tolerant’ response—that have been put forward in the name of how public authorities can protect democracy, and track how conceptual conflations between populism and anti-democratic politics have informed them.Footnote 2 We conclude by pointing towards some basic characteristics that public initiatives opposing populist parties should have to avoid these pitfalls, thus pointing out some potential avenues forward that do not merely fit the ‘intolerant’ and ‘tolerant’ approaches discussed in this special issue.

Populism (and not authoritarianism or extremism)

Populism is a difficult phenomenon to grapple with, both conceptually and normatively. Conceptually, there is continual debate about how to define populism, although this has become less contested over time. As a result, there is now something of a core agreement across the central approaches in the literature that populism is a political phenomenon that revolves around a central divide between ‘the people’ and ‘the elite’, and moreover, that it can appear across the ideological spectrum (Moffitt 2020).Footnote 3 Normatively, populism has what has been labelled as an ‘ambivalent’ relationship to democracy, which makes it hard to discern whether it is a clear threat or corrective to democracy (Rovira Kaltwasser 2012)—but which in turn makes it hard to practically target populists as outright anti-democrats in the mode of extremists or authoritarians. This, however, has not stopped an increasing tendency in the academic and popular literatures on populism to conflate, or at the very least, blur the lines between the phenomenon and its anti-democratic cousins. This takes two central forms.

The first is the literature that treats populism as synonymous with only radical right variants of the phenomenon. Here, the features of the populist radical right party family—nativism, authoritarianism and populism (Mudde and Rovira Kaltwasser 2017, 83, pp. 95–96)—are treated as features of populism in general. This is in many regards a result of the recent success of populist right parties as compared to populist left parties in Europe, with the former incorrectly coming to stand for the phenomenon in general in many analyses (on this, see Stavrakakis et al. 2017). The results of this are evident in the common conflation between nativism and populism in the literature (on this, see De Cleen 2017), or the frequent assertions that populism is a racist and xenophobic form of politics, even though left-wing populist parties in several countries are amongst the most strident anti-racist voices in their legislatures (Katsambekis and Kioupkiolis 2019) or have constructed multi-ethnic alliances (Madrid 2012).

Some authors go a step further and run populism together with extremism or authoritarianism. Noting this tendency, van Kessel contends that ‘in the European context, populism is habitually associated with xenophobic politics and parties of the extreme or radical right (and therefore considered to be dangerous)’ (van Kessel 2015, p. 2). Practically, this is seen in the literature in terms of case selection, where parties that are clearly extremist, with explicitly anti-democratic goals and an open embrace of violence, are instead wrongly labelled as ‘populist’ (see Eatwell 2017, pp. 376–379). Examples of this include the Greek Golden Dawn, an openly neo-Nazi party (Norris 2019), and the neo-fascist British National Party (Hogan and Haltinner 2015).Footnote 4 Such misclassifications associate populism with violence, anti-democratic politics and Nazism—and arguably can go the other way as well, unwittingly ‘whitewashing’ these genuinely dangerous neo-fascist or extremist parties by instead seeing them as arguably less-dangerous ‘populist parties’.

The second literature in this vein is that which presents populism as a form of ‘proto’-fascism or latent authoritarianism. Here, although populism is seen as distinct from extremism, fascism or totalitarianism, it is nonetheless seen as something of a steppingstone on the way to the establishment of an anti-democratic regime (see Urbinati 2019). Most importantly, there is a distinct directionality at play here—rather than populism having the potential to work as a radically democratic phenomenon (Mouffe 2018) or democratic ‘corrective’ (Rovira Kaltwasser 2012), here it moves inexorably in an anti-democratic direction. For example, Mounk has claimed that populism is ‘a prelude to autocratic rule’ and argued that ‘it is easy for illiberal rulers to make the transition from populism to dictatorship’ (Mounk 2018, p. 35), while Müller has similarly claimed that if populists get into positions of power, they will essentially become authoritarians: they will ‘engage in occupying the state, mass clientelism and corruption, and the suppression of anything like a critical civil society’ (Müller 2016a, p. 102). Some are even more explicit than this in noting populism’s allegedly unescapable march towards anti-democratic forms: Abts and Rummens (2007, p. 421) claim that populism is ‘proto-totalitarian’, while Finchelstein argues that ‘[d]ictatorship is one of the foundations of modern populism’ (2016, p. 229).

This tendency became increasingly pronounced during the presidency of Donald Trump, when this case—one in which populism indeed did lead towards anti-democratic ends—started to function as representative of populism in toto in public debates. However, universalizing one case ignores the fact that many populist leaders and parties have managed to avoid becoming totalitarians, extremists, or authoritarians, and have functioned within democratic party systems, both outside and inside the halls of power (in the European context, we can think here of examples as ideologically varied as Podemos, the 5 Star Movement or the Norwegian Progress Party). Moreover, the empirical literature shows that some populist parties have moderated over time (Akkerman et al. 2016; Capaul and Ewert 2021). Thus, this directionality is not automatic or inexorable, and should not be taken as a general rule. This is not to say that populists cannot become authoritarian: they clearly can, but when this point is reached, we argue that they should be definitionally labelled as such—as authoritarians—rather than populists, as they have clearly crossed a line in terms of their relationship to democracy.

Why is any of this conceptual slippage problematic? Quite simply, because populism, extremism, authoritarianism, fascism and totalitarianism are all different things—and these labels matter. Although we acknowledge that these are all contested concepts, to run these together, treat them synonymously, or even to assume that there is a clear and automatic directionality at play between them risks overlooking the specificities of populism. First, they are different types of political phenomena: where there is debate about whether populism is an ideology, strategy or mode of discourse, authoritarianism and totalitarianism are often understood as regime types or political systems (Levitsky and Way 2010; Roberts 2020), while extremism does not fit any of these categories, and is rather a quality of adherence to an ideology (Backes 2011; Lucardie 2014). Second, they have different political subjects: for example, while populism valorises ‘the people’, fascism idealises the figure of the ‘new man’ as the redeemer of the nation—as Eatwell notes, ‘it is inconceivable that a populist could espouse the Hitlerjugend slogan: ‘You are nothing, your nation (Volk) is everything” (2017, p. 381).

Third, populism has a far more ambiguous relationship to democracy than these phenomena. Totalitarianism, for example, is explicitly defined by its anti-democratic characteristics and its opposition to equality (Lefort 1986); likewise, fascism rejects the ‘decadence’ of democracy and seeks to replace it with an ultra-nationalist ‘new order’ in which the state and society are one (Griffin 2018). In contrast to this, Albertazzi and McDonnell (2008, pp. 3–4) note that populism relies on a notion of equality (at least among ‘the people’) as opposed to the strict hierarchy of totalitarianism, and that populists tend to see community rather than the strong state as the locus of politics. The same goes for the label of ‘extremism’, which also (at least in the contemporary context) is set against democracy and notions of popular sovereignty. For example, Mudde, in using the term to delineate the different subtypes of the contemporary right, notes that the ‘extreme right’ ‘rejects the essence of democracy, that is, popular sovereignty and majority rule’, whereas the ‘radical right’ (a category that he places many populist right parties under) ‘accepts the essence of democracy, but opposes fundamental elements of liberal democracy, most notably minority rights, rule of law, and separation of powers’ (Mudde 2019, p. 7). Similarly, Lucardie, in his work on extremism, notes that extremists openly reject the existing democratic political system and its dominant values, and seek to put forward a far-reaching alternative in its place (Lucardie 2014, pp. 11–31). In other words, extremists wish to overturn the existing system, and replace it with something completely different—we can think here of fascists, Islamists or anarchists. One may not like populists, but this is not the kind of political transformation that they tend to advocate: they do not launch a wholesale opposition to democracy as such.

Populism & (liberal) democracy

In line with this argument, there is something of a growing consensus in the academic literature that populism is a democratic phenomenon, just not a liberal democratic one. This is argued in a number of different ways: Mudde maintains that ‘populism is pro-democracy, but anti-liberal democracy’ (Mudde 2019, p. 8); Krastev notes that ‘populism is anti-liberal but it is not anti-democratic’ (Krastev 2007, p. 60), while Mounk (2018) claims that populism is a form of ‘illiberal democracy’ that should be interpreted as a reaction to the growth of ‘undemocratic liberalism’. In this vein, Pappas (2019) goes so far as to define populism as ‘democratic illiberalism’, which he contrasts, like the above authors, with (democratic) liberalism, but also with non-democratic illiberalism.

These formulations hinge on the distinction between a generic understandings of democracy versus liberal democracy. In the former, which is often practically rendered as majoritarian democracy, the sovereign voice of ‘the people’ is the primary—and indeed only—political force that matters. In the latter, the worst tendencies of majoritarian democracy must be balanced by institutional and systemic protections of individual rights and minorities as a way of ensuring pluralism. It is worth noting that there are democratic theorists who work on populism, such as Müller (2016a), Urbinati (2019) and Rummens (2017) who doubt if we can speak of ‘real’ democracy—even in a minimal sense—without its liberal protections intact.

As a result, such theorists contend that populism’s illiberalism automatically renders it a threat to democracy. They point to several of its serious failings in this regard, arguing that it does not leave space for plural understandings of the good; nor of multiple characterizations of who ‘the people’ are (Müller 2016a). They contend that when in power, populist parties reject the notion of a legitimate opposition, who are seen to be in thrall to ‘the elite’ (Rummens 2017). They also point to populism’s problematic faith in the leader as the vessel of ‘the people’s’ will and its tendency to construct power relations between ‘the people’ and the leader in a strictly vertical manner, and how this can lead down the path of authoritarian leadership (Weyland 2020). And in terms of institutions that we tend to associate with the proper functioning of liberal democracy as a political system—a free press, an independent judiciary, monitory bodies—they argue that many populists have a poor record when in power, tending to block, stack, ignore or impinge on any of these that happen to get in their way (Arato and Cohen 2021).

Yet there are other scholars who push back against this characterization, and argue that populism can be a democratically inclusive phenomenon. Mudde and Rovira Kaltwasser (2017, ch. 5), for example, have argued that populism can engage social sectors that have previously been ignored in political processes, and have tracked how populists can play an important role in processes of democratization, in contexts as distinct as Bolivia and Thailand. Others are doubtful of depictions of populism as anti-pluralist, instead seeing it as a phenomenon that can open space for egalitarian praxis (Frank 2017) or grassroots democratic experimentation (Grattan 2016). Radical democrats such as Mouffe (2018) and Laclau (2005) go a step further, claiming that contemporary liberal democracy displaces the essential role of dissensus, conflict and emotions in democratic politics, as well as the constituent power of ‘the people’. Here, they see populism as able to challenge the hegemonic status quo, rehabilitate democratic dissent and deliver power back to ‘the people’.

With this in mind, where does populism actually sit vis-à-vis liberal democracy, and what makes it such a vexed relationship? Conceptually speaking, we contend that part of the difficulty of pinning down populism’s democratic credentials results from it being a phenomenon that does not clearly exist in opposition to or on the ‘outside’ of democracy: rather, its development, rise and functioning is inextricably tied to the development of liberal democracy. As Urbinati argues, ‘[c]ontemporary populism is not the product of some malevolent force but of the very model of democracy, representative and constitutional, that stabilized our societies after World War II’ (Urbinati 2019, p. 124). Its ‘interiority’ in terms of its relationship to democracy is thus what makes it so vexing in terms of working out how we should respond to it as a political phenomenon. Here, Arditi (2007) uses the spatial metaphor of seeing populism as sitting at the ‘internal periphery of democracy’, which he conceptualizes as

  • a hazy territory that indicates the outermost limit of an inside and the beginning of the outside of a system, a grey area where the distinction between inside and outside is an effect of polemic. Populism can remain within the bounds of democracy, but also reach the point where they enter into conflict and go their own separate ways (Arditi 2007, 87).

Urbinati echoes this spatial metaphor, claiming that ‘populism can stretch constitutional democracy toward its extreme borders and open the door to authoritarian solutions and even dictatorship; the paradox is that, should this regime change take place, populism would be unseated’ (2019, p. 112), as it would not actually be populism anymore. The salience of this metaphor lies in the fact that it acknowledges a) the dangers of populism in terms of its potential nefarious effects on liberal democracy and b) the fact that populism is ultimately of democracy, and that these two things can be true at the same time. Populism’s liminality in this regard is what makes it so challenging in terms of determining how to respond to it in its party form: such actors certainly require a response when they seem to be breaching democratic boundaries (which given populism’s liminal spatial relationship to democracy, populists are set up in an opportunistic position to do), but this needs to be a considered one that takes populism’s democratic ambiguity into account.

Responding to populist parties: intolerant and tolerant approaches

Bearing in mind populism’s double-sided relation to democracy, we can now turn to the question: how should public authorities respond to populist parties? The extant literature (and practice) on the question offers a wide array of answers. They include on one hand suggestions to ban, isolate, censor and impose on them international sanctions, and on the other hand proposals to either ignore or engage with them, cooperate or even incorporate them into government. These strategies, in line with the introductory article of this special issue, can roughly be divided into two broad approaches: intolerant and tolerant.Footnote 5 Our goal here is to highlight the implications and suitability of choosing one or the other approach as a basis for formulating public responses specifically to populist parties. We do not examine the effectiveness of each approach in containing populism (a realm that calls for empirical analysis), nor the legitimacy of tolerance or intolerance in general, but their legitimacy in relation to populism. In that sense, we evaluate tolerance and intolerance towards populism from the viewpoint of the conceptual mismatch between, on one hand, the nature of democratic threats as envisaged by both approaches and, on the other, the nature of populism as we understand it. Indeed, what becomes clear when examining these approaches, in line with the previous discussion, is how older work on responding to extremism or fascism has influenced the contemporary literature—and the limits inherent in doing so. In this section, we outline these two approaches, provide illustrative examples of each from the literature on populism, and advance a critique of their ill-suitedness in dealing specifically with populism.

The intolerant approach to populism and why it is unjustified

Historically, the intolerant approach—better known through the paradigm of ‘militant democracy’—was developed as a reaction to fascism. Coined by Karl Loewenstein, who had just escaped Nazi Germany, it laid out a normative responsibility for democracies to restrict the civil rights and freedoms of those who were committed to abolishing democracy through nonviolent means (Loewenstein 1937b). These restrictions were justified on the grounds that fascism was not an ideology per se in Loewenstein’s view, but a ‘technique’ based on emotionalism, opportunism and manipulation of the rules (Loewenstein 1937a, p. 423); as a result, the ‘rational’ practices of existing constitutional politics were ill-equipped to resist or address it. Hence, militant democracy was conceived as a set of tools to be used by public authorities against parties that were, on one hand, opposed to democracy and, on the other, assumed to be lacking ideological principles. Contrast this to the populist parties of our day. Not only are they hard to definitively label as anti-democrats, but they often combine their populism with a clearly identifiable ideological viewpoint, whether of the left or the right. The populist principle can thus neither be qualified as ‘non-ideological’ or illegitimate in the mode that Loewenstein viewed fascism, nor as automatically anti-democratic.

Until the dawn of the twenty-first century, militant democracy developed into a widely shared constitutional principle that was used to block both fascist successor parties and communist parties in Europe. Yet, as new challenges to liberal democracy arose, the problem of identifying those who ‘seek to undermine or abolish the free democratic basic order’ (as stipulated in the archetypical militant democratic constitution, the German Basic Law, Art.21) has become a daunting one. This challenge produced two separate tendencies. The first has seen militant democrats search for stricter, more minimalist or self-limiting criteria (Kirshner 2014; Rijpkema 2018). The second tendency is to expand the scope of application of militant democracy to a variety of actors, such as anti-secular parties (Sajó, 2012), or parties whose leaders ‘do not disavow the language of democracy’ and retain some democratic institutions (Müller 2016b, p. 262). This is where many scholars are tempted to conflate populism with extremism and use experiences of fighting the latter without accounting for the important differences between these phenomena.

Indeed, most attempts to sketch out initiatives opposing populism draw uncritically from the literature on anti-extremism and militant democracy. For example, Rovira Kaltwasser and Taggart’s special issue on ‘Dealing with Populists in Government’ explicitly looks ‘for characterizations of responses to political extremism in general’ (Rovira Kaltwasser and Taggart 2016, p. 209) in constructing their analytical framework for understanding how to respond to populism. Even if they do not identify as militant democrats (with Rovira Kaltwasser 2017, 2019) explicitly arguing against such approaches in other work), they draw on the work of Capoccia (2005), who examined responses to extremism in inter-war Europe, developing a typology that covered strategies including militancy, incorporation, purging and education; and the work of Downs (2012), whose work on reactions to ‘pariah parties’ draws together (and to some extent, fails to delineate between) extremist, radical, populist and nationalist parties. However, there is no consideration of whether drawing on reactions to extremism may or may not be a suitable source of inspiration for conceptualizing reactions to populist parties.

An even more explicit militant democratic approach towards populism is put forward in the work of Abts and Rummens (Abts and Rummens 2007; Rummens and Abts 2010; Rummens 2017). Against the idea that populists should be treated as legitimate political contenders, they argue that ‘populists are no longer ordinary adversaries, but political enemies who hold an incompatible view of the symbolic structure of the locus of power itself’, and that ‘populist parties, to the extent that they are inimical to democracy, should be revealed as such, treated accordingly and, if necessary, isolated from power’ (Abts and Rummens 2007, p. 422). They also classify parties that are generally regarded as populist in the academic literatureFootnote 6—the Vlaams Belang/Blok in Belgium, the Front National in France, New Democracy in Sweden, and the Freedom Party of Austria—as ‘political extremists’ alongside the likes of the neo-Nazi National Democratic Party of Germany and the political wing of the Basque separatist group ETA, Batasuna, in Spain (Rummens and Abts 2010). Overall, they advocate strategies including isolation of populist parties in the form of a cordon sanitaire, or in the last instance, party bans. Interestingly, in earlier work, Müller also explicitly stated that populists are suitable targets of militant measures (Müller 2012, pp. 1268–1269), although he has now moved to recommending milder responses (Müller 2016a, pp. 75–99).

There are normative and practical issues with utilizing a militant democratic approach when opposing populism. As we have argued, populists are not extremists or fascists, and thus strategies to deal with them are likely to be different and require more nuance than dealing with those who wish to overhaul or destroy the democratic system. Indeed, militant democracy was conceived as a defence of democracy against explicitly anti-democratic parties. This meant that the process of identifying and excluding such parties was more or less straightforward: party manifestos, public speeches and activities of key members were screened to detect direct or indirect calls for overturning the liberal democratic regime. Such calls are unlikely to be found in modern-day populist discourse, which often emphasizes opening up the political class and giving voice to hitherto ‘unheard’ social groups and their political demands (Laclau 2005). Nor does any populist party today, to our knowledge, maintain party militia, illicitly produce firearms or support wearing political uniforms, which were some of the typical warning signs for Loewenstein (1937a).

To be sure, a populist party’s anti-democratic credentials may be clandestine or nascent but nevertheless real, which for some scholars means that we simply need to update and sharpen our militant antennas in order to trace them (Rijpkema 2018). But two problems persist. Banning populist parties requires turning a blind eye to definitions of populism as not opposed to democracy. In other words, if we accept that populism challenges only aspects of liberal constitutionalism but not aspects of democracy as such, then it is hard to justify their banning in the name of democracy. Even if we assume that democracy and liberal constitutionalism cannot be analytically or practically separated, banning is still problematic because it cancels out the ‘inclusive’, ‘productive’, ‘democratic’ side of populism (Mouffe 2018). If banning is supposed to protect democracy from collapse, is a political party that can potentially restore some of its defunct aspects really worthy of a ban? In short, repressing populist parties cannot be justified unless it relies on a partial, incomplete account of what populism is. To ‘qualify’ for such repression, a party must turn from a genuinely populist into an anti-democratic one.

A similar conclusion can be drawn if we review general normative objections to militant democracy from the viewpoint of populism. Militant democracy is accused of being anti-pluralist, in that it silences groups that may have other legitimate interests beyond their fundamental questions about democracy (Kirshner 2014, 86–106). This is certainly the case for populist parties, whose agendas not only go beyond issues relating directly to the form of government, but also tend to include core systemic demands—for example, the expansion of the use of referendums, or the increased voice of citizens in political decision-making—that are democratically legitimate. As put by Rovira Kaltwasser, ‘fighting populist forces with ‘hard’ measures (e.g. banning them or cancelling the political rights of their leaders) is particularly problematic, because populists normally do have some right in claiming that it is crucial to ‘control the controllers’ and prevent elites from disempowering the many’ (Rovira Kaltwasser Rovira Kaltwasser 2017, 497).

In addition, militant intolerance is criticized for undermining the majority will and opening the door to expert rule and counter-majoritarian institutions (Malkopoulou 2020). But these practices not only put into question the regime’s democratic legitimacy; they can actually help fuel the fire of populism. Insofar as populists stand for majoritarianism and against elite power, deploying counter-majoritarian tools against them may confirm the justifiability of the populist cause and increase polarization (Rovira Kaltwasser 2017, 500). In sum, excluding populists in a militant democratic fashion forecloses what could be an inclusive and pluralistic debate over crucial issues, such as the norms and practices of democracy itself. Therefore, it is only reasonable to assume that the right way to address populism is with toleration. The next section examines this presupposition.

The trap of tolerating populism

At the opposite end of the militant approach, we find the paradigm of tolerant democracy, which relies on the unconditional primacy of democratic procedures. In principle, the tolerant approach rejects repression and supports unrestricted inclusion of all political parties, even those that promote unacceptable, even anti-democratic views. A democracy should be value-neutral, in order to preserve negative freedom, and operate through clearly established procedures and formal rights that treat every citizen equally (Kelsen 2006). No exceptions to this equal treatment are justifiable without endangering the very nature of democracy, and certainly no discrimination on the basis of political ideas, statements, intentions or actions. As Bourne notes in the introduction of this special issue, this approach sees populists as relatively ‘normal’ political actors—or at least worthy of ordinary treatment (see also Malkopoulou, forthcoming).

That said, contrary to its reputation for being passive and self-complacent (a critique launched by Loewenstein himself), the tolerant approach does foresee remedies for dealing with extremists, and even some limited type of restrictions. What distinguishes these interventions, however, is that they do not punish opposition to substantive democratic values as militant democrats do but aim on one hand at protecting the legality and integrity of procedures and on the other at preventing direct physical harm against individuals (Malkopoulou 2021). The aim here is not to protect a certain set of values associated with liberal democracy, but to redress harmful activity regardless of its political motives. On these grounds, the tolerant approach raises two main barriers against democide. The first is a set of institutional checks and balances, such as a clear division of power which, among other measures, establishes an independent judiciary that can perform judicial reviewFootnote 7 (Ginsburg and Huq 2018). Similar checks and balances, such as majoritarian or high-threshold proportional electoral systems, presidentialism and federalism, work in the same spirit: to prevent concentration and abuse of political power regardless of the ideological motives of such abuse.

However, populists in government seem to have established a blueprint for how to navigate these procedural barriers. They often rely on their electoral mandates and use constitutional or legal methods to dismantle the constitutional system they inherit. Scheppele has aptly described this phenomenon as ‘autocratic legalism’ (Scheppele 2018), while Weyland has referred to a similar phenomenon in Latin America as ‘discriminatory legalism’ (Weyland 2013). Take the (in)famous example of Hungary. Since coming to power in 2010, Viktor Orbán and his Fidesz party have deployed formally legal procedures to limit the Constitutional Court, hijack bottom-up referendums, pass a new constitution, compromise accountability institutions, and assert control over the media, the courts, the electoral commission and the presidency by hiring loyalists in key positions (Bánkuti et al. 2012). These measures have effectively weakened judicial independence and the separation of powers. Because elections continue to take place (and are rigged in nonviolent and technical ways that are not always obvious) and ironically ‘nothing illegal is going on’, populist governments that follow Hungary’s lead can maintain a veneer of democracy and legality (Scheppele 2018, 547, 579). Thus, it is by now painfully clear that procedural checks and balances are often unsuitable against populist parties—particularly those in power—because the latter rarely launch frontal attacks on democracy.

The second tool that tolerant public authorities commonly use to prevent democratic subversion is the sanctioning of illegal activities perpetrated by individual party members. This allows curbing political violence and other acts that pose an ‘imminent danger of direct harm’ against other groups or individuals. Incitement to violence and hate speech may fall into this category, but only if it can be established that they are linked to criminal conduct or direct harm to victims (Issacharoff 2007, 1416).Footnote 8 This criminal law avenue is appropriate for extremist parties, who tend to be involved in underground activities and acts of violence and whose culpability is therefore easier to establish. For example, there was ample evidence of grave crimes committed by cadres of the Greek neo-Nazi party Golden Dawn, which was convicted as a criminal organization under such a tolerant scheme of democratic defence (Malkopoulou 2021). This is usually not the case for populist parties. Unlike fascism or perhaps some variants of anarchism, there is little that links populism with a principled endorsement of violence. In addition, populist parties have at times systematically removed from their ranks individuals involved in violent incidents, precisely to dispel any perception that they operate at the margins of legality. This can be seen in the cases of the Sweden Democrats, who broke with their youth league due to the latter’s links with prominent neo-fascists (Jungar 2016), or Marine Le Pen’s dédiabolization strategy for the Front National/National Rally, which has sought to sever ties with the former party’s extremist past (Ivaldi 2016).

Furthermore, as Sajó pointedly remarks, ‘populist rhetoric does not fall into the net of hate speech’ and is therefore difficult to regulate (Sajó 2019, 202): emotional outbursts or displays of political outrage and social contempt, common among populists, are equally protected by the right to free speech. Populist right proposals to exclude migrants or misrepresentations that dehumanize them, even if they rely on hate, do ‘not count as discriminatory in law’ (Sajó 2019, 202). Here too, the tolerant approach does little to address the challenges raised by the special nature of populism. Due to their apparent conformity with the law, populist parties are considered to be ‘ordinary’ political actors, whose voters need to be engaged rather than cordoned off. As long as populists stay within the boundaries of the law, the argument goes, there is an obligation to treat them as legitimate opponents (Mouffe 2018).

How to do this—and how far democrats should go in terms of entertaining the grievances of populist parties—differs throughout this approach. Müller notes that listening to the social-cultural and material grievances of populist sympathizers in a serious way is a necessary task for liberals, who ought to remain committed to universal suffrage and the associated belief in the ability of the masses to vote responsibly (Müller 2016b, 17). Badano and Nuti argue that, although supporters of right-wing populism actually are unreasonable (in a Rawlsian sense), there is a duty to discursively engage with them; this involves both a duty to talk but also ‘a duty to listen to them’ (Badano and Nuti 2018, 161). Rovira Kaltwasser similarly calls for ‘engaging in an honest dialogue’ (Rovira Kaltwasser 2017, 502) with populists.

But the logic behind this permissibility is not always clear. Will engaging populists not increase democracy’s vulnerability? Rosenblum has tried to counter these worries by highlighting the moderating potential of inclusion. For her, the reason for maximal inclusion is a belief that ‘electoral political competition, like any strong institutional practice, is formative’ (Rosenblum 2010, 290). Participation teaches citizens how to act democratically, to argue for the public good, ‘to cast their demands in terms that apply beyond themselves’ (Rosenblum 2010, 290), to compromise and to accept majority decisions. These presumptions may hold for ethnic and religious parties that lie at the centre of Rosenblum’s concerns; but less so for populist parties. Ethnic and religious parties may challenge Western societies’ liberal and moral values, but not necessarily their social and political structures, as populism sometimes does; further, they are usually minority parties and not as combative or power-seeking as populists. Hence, the risks associated with integrating populist parties may be different and probably larger than those of integrating ethnic or religious parties.

Other concerns have to do with the scope, assumptions and side-effects of populist parties’ unconditional inclusion. While it all sounds rather nice, the ‘solutions’ of listening or engaging in dialogue with populists remain rather vague: what happens after you listen and talk with them? What if they continue to advocate illiberal and discriminatory views? Are liberal democrats expected to adopt some version of these views in order to demonstrate a commitment to true compromise? There is a risk that the principle of toleration goes as far as prescribing accommodation, if not endorsement of populist views in the name of inclusion. At the same time, ‘talking with’ populists may be just a decorative appeal, reproducing a political divide or even further polarization. Besides, the ‘unreasonable’ may not be willing to engage with ‘us’, especially if they are not involved in setting the terms of this deliberative engagement.Footnote 9 There is indeed nothing unreasonable in protesting against the potentially patronizing character of such deliberations.

In sum, although not nearly as democratically or normatively problematic as the ‘solutions’ offered by the militant democratic approach to populism, significant issues around the tolerant approach remain. On one hand, populist parties manage to circumvent procedural checks and balances and are difficult to censure on criminal grounds, as these tools were conceived with different actors in mind—actors that launch more frontal attacks on democracy. On the other hand, calls for inclusion are unclear about the terms and scope of engagement and, here too, such calls have been conceived with different actors in mind—religious and ethnic minorities. It seems then that the tolerant approach as it currently stands is also not optimally geared to the specific problems posed by populism.

Conclusions

The lessons we draw from our observations are primarily negative. Because populism is not extremism, public authorities should not respond to it as if it was. This problem arises when scholars of populism and of democratic defence do not take enough care in distinguishing between populist and anti-democratic parties. As a result, they tend to apply or at least consider the same types of responses to both phenomena. Yet, as we show, populists are neither necessarily anti-democrats nor violent extremists per se; most commonly they tend to operate within the confines of legality, making it hard to hold them accountable for breaches of the rule of law stricto sensu. And while engaging with populists is a reasonable endeavour, the purposes and limits of this engagement are rather unclear.

Taken together, these insights can inspire some positive guidelines for responding to populist parties. Our basic premise, drawn from this article, is that initiatives ought to be adjusted to the peculiar nature of populism. They should be triggered by specific moves that are typical of populist parties, such as court packing (if in power) or the spread of disinformation—some of which are outlined in the other articles of this special issue—and set appropriate rules for relevant institutions and the authorities that regulate them. What form could such initiatives take? How could public authorities successfully contain populism’s most egregious characteristics, yet salvage its productive side?

‘Good’ responses may involve a carrot and stick approach, encouraging certain manifestations of populism and discouraging others. Yet, not all such mixed approaches pass the test of actually engaging with the precise characteristics of populism. Van Spanje and Graaf (2018), for example, propose an intra-partisan strategy of ‘parroting the pariah’, which engages with the policy demands of voters of pariah parties, yet at the same time keeps such actors at bay by excluding them from cooperation in government. Rummens and Abts (2010) suggest a ‘concentric model’ of conditional inclusion and exclusion, according to which problematic actors should be accorded sufficient leeway to speak and act when they operate informally at the periphery of the public sphere, but tolerance towards them should gradually decrease the closer they get to the centres of power. While we find these mixed approaches promising, they are nevertheless unable to live up to the challenge of salvaging the democratic side of populism—and moreover, they often fall back into the trap described earlier of running populism together with extremism or radical politics. How so?

As we have argued throughout this paper, public authorities must steer clear from excluding populists from policy-making processes (even if this exclusion is conditional), because such a move tends to cancel out the ‘corrective’ potential of populism—its ability to bring out democracy’s shortcomings and force us to face them. Instead, not only do parroting and concentric strategies condone exclusion, but they over-politicize the process of drawing the line between what is tolerated and what not, which adds an extra layer of arbitrariness to the mix. In the parroting-model case, adopting some of the populists’ policy demands leaves it up to competing parties to decide which of these demands would be popular to adopt, with the impending and very real risk of amplifying and expediting the populists’ less-than-legitimate (e.g. nativist) key demands; moreover, in the long term, this may result in a ‘mainstreaming’ of populism even if populist actors are kept at bay. In the concentric-model case, tolerating minor but not major populist parties creates an opportunistic incentive for other political forces to shift towards silencing populists and rearranging power relations when the latter threaten the political prospects of the former. In short, intolerant methods, whether conditional or not, miss the target. Thus, carrot and stick approaches adapted to populism should primarily draw from the repertoire of the ‘tolerant approach’ and enforce only those prohibitions that do not involve restrictions on political rights or exceptions on regular procedures. But, as mentioned above, this line of response is also likely to misfire.

Thus, based on our understanding of the nature of populism presented in this article, we find that public-authority responses to populism should move beyond the binary of toleration or non-toleration, which is far too legalistic and political, and be redirected towards a closer engagement with the social problems that light the populist fire. Responses ought to account for the fact that populism has emerged due to a combination of circumstances, whose prime ingredient is social injustices. To live up to the challenge of containing some and retaining other sides of populism, public authorities must capture the social dynamics that animate populist parties through a range of policies that rehabilitate the social and egalitarian face of democracy.Footnote 10 The broader aim is not to eliminate democracy’s vulnerability altogether, but to turn it into a productive force that, taking its cues from existing shortcomings and imperfections, moves democracy into a new (and better) direction.