Is There a Crisis of Democracy in Europe?

This paper discusses the current prospects of democracy in Europe from four perspectives: the birdʼs-eye view of long-term trends; the perspective of citizens’ support of democratic principles and their dissatisfaction with the way democracy works in their own countries; the voters’ perspective, which points to the rise of populist challengers in reaction to rising democratic dissatisfaction; and the elites’ perspective of populists in power. Overall, there is reason for concern, but no reason to dramatize. The long-term trends point to the expansion of democracy; the citizens’ support for democracy is still massive in Europe. At the same time, democratic dissatisfaction is widespread, giving rise to the surge of populist challengers from the left and the right. However, even if they gain power, populists meet with a large number of constraints that stabilize democracy.


Introduction
From the vantage point of the Chinese elites, not only free market capitalism but also Western-style democracy has lost its attraction, and they suggest that "90 percent of democracies created after the fall of the Soviet Union have now failed" (Wolf 2018). There might be some wishful thinking involved in their assessment of the state of democracy across the globe. But in Europe, too, the democratic "Zeitgeist" has become more pessimistic (Brunkert et al. 2018), and the current public debate about democracy suggests that liberal democracy is in crisis. In Europe, the rise of populism from the right and the left; the imposition of austerity on southern European countries by the Troika; Brexit; and the illiberal measures taken by governments in Hungary and Poland are interpreted by pundits-academics as well as public intellectuals-as so many signs of a crisis of democracy.
The current situation in Europe reminds me of the early 1970s, when preoccupied observers identified "a breakdown in consensus," "a political and economic decline," and "a crisis of democracy." This crisis talk came in two versions (Held 2006): those arguing from the premises of a pluralist theory of politics and those arguing from Marxist theory. According to the doomsayers from left and right, the liberal democratic state had become increasingly hamstrung or ineffective in the face of growing demands that were either "excessive" (Crozier et al. 1975) or the "inevitable result of the contradictions within which the state is enmeshed" (Offe 1984;Habermas 1973). The crisis talk of these theories was later disconfirmed by a large comparative research program-the "Beliefs in Government" program (Klingemann and Fuchs 1995)-that was triggered by the very preoccupations with the alleged democratic crisis. The results of this program showed that Western representative democracies proved to be perfectly capable of absorbing and assimilating growing pressure from societal problems, and the forms of political expression taken by such pressure could be understood as the normal manifestations of democracy in complex societies. By the time these results were published, however, nobody cared anymore about the crisis of democracy. In the meantime, the Berlin Wall had fallen, democracy had triumphed, pundits had declared the "end of history," and the public had moved on. But the dominant mood soon changed again, and the crisis talk did return in new garb. In 2000, a follow-up study under the title of "Disaffected Democracies" was newly preoccupied by the lack of public confidence in leaders and institutions of democratic governance (Pharr and Putnam 2000). The study argued that the causes for the decline of confidence did not lie in the social fabric, nor were they the result of general economic conditions. The problem, it suggested, was with government and politics themselves. Similarly, the contributors of yet another study on political disaffection in contemporary democracies (Torcal and Montero 2006) highlighted the decisive role of politics and institutions in shaping political disaffection.
In the more recent past, pessimism about the state of Western democracy has again increased. Larry Diamond (2015) writes about "a democratic recession," and Marc Plattner (2017) sees democracies "on the defensive." In 2018, the academic observers' tone became more alarmist still. Sasha Mounk (2018) published a treatise entitled The People vs Democracy, Steven Levitsky and Daniel Ziblatt (2018) discussed "how democracies die," and David Runciman (2018) similarly wrote about "how democracy ends." Is this time different? Are we heading for a truly transformative crisis of western democracy? Although I share some of the concerns of the more alarmist colleagues, I do not think we should dramatize the current situation. I side with The Economist, which, in its June 16, 2018, issue suggested that "reports on the death of democracy are greatly exaggerated." But, it added, "the least bad system of government ever devised is in trouble. It needs defenders." And I think one way to defend it is to get the facts right and to base our expectations about the future of democracy on the best available empirical evidence.
I shall discuss the current situation of democracy from four perspectives. With the exception of the first perspective, I shall focus on the situation of democracy in Europe. First, I briefly adopt the bird's-eye perspective of the long-term trends to reassure us where we ought to situate the argument in the grand scheme of things. Next, I adopt the perspective of the citizens and ask whether they adhere to the principles of democracy and how they evaluate democracy in their own countries. Although I agree with Levitsky and Ziblatt (2018, p. 19) that the citizens' democratic values do not provide a guarantee for the survival of democracy, following Almond and Verba (1965) I believe that such values constitute a necessary condition for democracy's survival, as well as a key constraint of possible relapses into authoritarianism. While people cannot shape at will the government they possess, the governments cannot ignore the will of the people in the long run. I shall compare the citizens' values with their evaluation of democracy and show that there is widespread dissatisfaction with the way democracy works, even if citizens across Europe essentially share democratic principles.
Third, I adopt the perspective of the voters and discuss their reaction to the widespread dissatisfaction-the decline of mainstream parties and the rise of populist challengers. The "Zeitgeist" is not only pessimistic about democracy, it is also heavily populist (Mudde 2004). According to a common view, populists in particular constitute an increasing threat to liberal democracy (Müller 2016). While acknowledging the rise of populism, I shall suggest that we need to be careful about how to interpret it. In particular, we need to distinguish between the two forces on which it builds-on the one hand, the substantive demands of citizens, which are rooted in long-term structural transformations of society, and on the other hand, democratic dissatisfaction. Finally, I shall turn to the perspective of the elites, i.e., to the question of what happens when populists get into power. As Norris (2017) has usefully pointed out, consolidation of democracy not only means that the overwhelming majority of people believe that democracy is the best form of government, it also means that all major actors and organs of the state reflect democratic norms and practices and no significant group actively seeks to overthrow the regime or secede from the state. The question is to what extent this second condition is currently guaranteed in Europe. I shall discuss five factors that serve to constrain the threat posed by populists in power-institutional constraints, partisan constraints, international constraints, and market constraints, as well as the already mentioned citizens' constraints. My overall assessment of the state of democracy in Europe is that there is reason for concern, but no reason to dramatize.

The Bird's-eye Perspective: The Centennial Trends
The bird's-eye perspective is the view of modernization theory and that of Steven Pinker (2018, p. 199-213). The latter presents evidence based on Polity IV data that show a long-term spread of democracy across the world. He also suggests that the spread of democracy across the globe has contributed to the spread of human rights. A recent survey of the state of democracy across the globe (Van Beek et al. 2018) could not find any signs of democratic backsliding for the established democracies of the West, even if there were various problems in other parts of the world. Based on Varieties of Democracy (V-Dem) data, the best data currently available, Brunkert et al. (2018) present the centennial democratic trends for democracy overall as well as for its participatory, liberal, and electoral components, which all have been rising in the long run (see Fig. 1). According to these data, there is evidence that the long-term trend has stalled since the turn of the millennium and shows some signs of a reversal, almost all over the globe. We should note, however, that this setback has been limited so far and that it is occurring at an unprecedented level of democratization across the globe.
Modernization theorists such as Christian Welzel and Ronald Inglehart believe that, despite all trending patterns, democracy always has been and continues to be a strongly culture-bound phenomenon and that democracy, in the long run, will impose itself. Thus, Inglehart (2016) suggests that modernization helps to generate conditions conducive to democracy: Cultural change is a major factor in the emergence and survival of democracy, and cultural change goes in the direction of more K Fig. 1 The centennial democratic trend (global democracy averages). Data are from the Varieties of Democracy (V-Dem) project (www.v-dem.net) and cover between 85 (in 1900) and 175 countries (in 2016). To calculate yearly global averages, we weighted countries for the size of their population. Comprehensive democracy is the product of V-Dem's electoral, participatory, and liberal components. Source: Brunkert et al. (2018, p. 6) democracy. Similarly, while acknowledging that there have been reversals in the trends of democratization, Brunkert et al. (2018) emphasize that they have been only temporary. Like Inglehart, they provide reasons for optimism by pointing out that democracy's cultural seeds-emancipative values-have been rising over the generations and are doing so in most parts of the world, including such seeming strongholds of authoritarianism as China, Russia, Singapore, and Turkey. According to Brunkert et al. (2018), what really matters is the extent to which people's support for democracy is inspired by emancipative values. And this kind of emancipatory support for democracy is neither in temporal nor generational decline in the most mature democracies, among which we find the western European democracies.

The Citizens' Perspective: Support of Democratic Principles and
Dissatisfaction with the Way Democracy Works

Support of Democratic Principles
Ever since Almond and Verba (1965), and in line with the notion that democracies are ultimately rooted in a supportive culture, political scientists have argued that stable democracies require supportive attitudes and norms: If citizens hold democratic values, democracy will be safe. In search of empirical support for this argument, Claassen (2020) has recently built a unique dataset based on 3765 national opinions about democracy, obtained from 1390 nationally representative public opinion surveys in 150 countries. His analysis confirms that the citizens' support for democracy is robustly linked to the stability of democracy, once it has been established.
In adopting the perspective of the European citizens, the pertinent question then becomes whether and to what extent they support liberal democracy today. There are voices that have some doubts about the solidity of the citizens' support for democracy today. Thus, in a widely cited recent paper, Foa and Mounk (2016) claim that citizens in Western democracies in general are increasingly turning away from democracy. Based on World Values Survey data, they maintain (p. 7) that citizens "have become more cynical about the value of democracy as a political system, less hopeful that anything they do might influence public policy, and more willing to express support for authoritarian alternatives." Moreover, they claim that younger cohorts are particularly affected by the decline in support for democracy. Now, the first thing to note about their dramatic message is that even according to their own data, we need to distinguish between the United States and Europe. While there is a clear decline of support of democracy among younger age groups in the United States, the same does not apply to Europe. Thus, in a reaction to their paper, Inglehart (2016, p. 18) interprets their results largely as "a specifically American period effect." He invokes three points that account for this American effect: the virtual paralysis at the top of the American democracy, massive increases in income inequality, and the disproportionate and growing political influence of billionaires as a result of the extraordinary role played by money in U.S. politics. He points out that existential insecurity undermines democracy, and that existential insecurity has been on the rise among most of the population in the United States-especially among the young. The second point to note is that one has to guard against basic errors when analyzing data such as those used by Foa and Mounk. These errors include exaggerating by cherry-picking cases and by using improper visual presentations of the survey data (Norris 2017), failing to distinguish generational from lifecycle effects, neglecting the variable composition of the countries in the aggregate European data from one survey to the other (Voeten 2017), and interpreting the data superficially. Using the same World Values Survey data, Voeten (2017) finds some evidence that U.S. millennials have grown somewhat more accepting of nondemocratic alternatives, but he finds no evidence that people in consolidated democracies in general (Western democracies-EU15 members, as well as Canada, the USA, Norway, Switzerland, Australia, and New Zealand, and also Costa Rica, Cyprus, Japan, Lithuania, Mauritius, Slovenia, and Uruguay) have soured on democracy or have become more likely to accept authoritarian institutions as a way to run their countries.
Using data from the European Social Survey 6 (ESS6), which includes an exceptional set of questions on democracy, I would like to show that support for the democratic ideal is alive and well in Europe, even if there are some regional variations. In the ESS6, which presents the situation in 2012, Europeans were asked how important it was for them "to live in a country that was governed democratically." They could answer on an 11-point scale, where 0 meant not important at all and 10 meant "extremely important." Table 1 presents the average responses in five European regions as well as the percentages of respondents who said that it was extremely important for them to live in a democracy. The five regions include K the Nordic countries, continental Europe (including the UK and Ireland), southern Europe, central and eastern Europe, and a group of three countries (Russia, Ukraine, and Kosovo) that are best characterized as "hybrid democracies." The latter constitute a benchmark for comparative purposes. As the table shows, averages are very high in all the regions. Even in hybrid regimes, they reach appreciable levels. On average, Europeans think that it is very important to live in a country that is governed democratically. In Nordic countries, a two-thirds majority even thinks that it is extremely important to do so. This share is much lower in the other regions, but it is still appreciable in hybrid regimes, where it reaches a bit less than one-third (29%). Based on the ESS6, we can also probe Europeans' detailed understanding of democracy, i.e., the extent to which they support the specific principles of liberal democracy. By way of illustration, Table 2 presents the average importance and the percentage of extremely important responses for two liberal principles of liberal democracy-equality before the law and media freedom. As one can see, the average importance and the shares of citizens who consider the principle of equality before the law extremely important are even higher than the corresponding values for democracy in general, and they are also rather high for media freedom. Strikingly, once we consider specific principles of liberal democracy, there is hardly any variation between the regions anymore. Even in the hybrid democracies, citizens know what it means to live in a democracy. If anything, the exception is now continental Europe, where the corresponding values are lower than in the other regions. In other words, there is a similar understanding of democracy across Europe. This does not only hold for the aspects I have presented here, but it holds more generally across a larger set of specific principles which, together, form a scale for liberal democracy (see . The fact that in some regions specific democratic principles are rated as more (or less) important than democracy overall may refer to region-specific attitudinal differences: Possibly, citizens in the Nordic countries and in continental Europe are very supportive of the ideals of democracy in general but may take some specific principles of democracy (such as media freedom) for granted. By contrast, citizens from the other regions, where the quality of democracy is typically lower, may be somewhat less supportive of democracy in the abstract, but when asked about specific principles of democracy, they may be very sensitive to the importance of the specific rights and obligations implied by them, given their past (or current) experiences with authoritarian regimes. Now, if the overall support of democracy is still guaranteed, there might still be some lack of support among younger generations, as claimed by Foa and Mounk. In an attempt to replicate their results, I rely on the same general question about the importance of living in a country that is governed democratically, and I also follow their choice to consider only the shares of citizens who believe that it is extremely important to live in a democracy. Figure 2 presents the shares of the age cohorts in four out of the five regions (I drop hybrid regimes in this analysis) who believe that it is extremely important to live in a democratic country. As we can In all four regions, younger cohorts are less enthusiastically embracing the democratic ideal than middle-aged cohorts are. However, the relationship is nowhere as strong as in the United States, and it is somewhat curvilinear in all regions, i.e., not only the youngest, but also the oldest cohorts are less enthusiastic about democracy than are the middle-aged. Moreover, the curvilinear relationship between age groups and democratic support is clearly more pronounced in some regions than in others. It is quite strong in the Nordic countries, at a comparatively high level of overall support for democratic ideals, and in southern Europe, at a lower level of overall support. By contrast, the curvilinear relationship of democratic support with age is less pronounced in continental and central and eastern Europe.

Democratic Dissatisfaction
The strong support for democratic principles across Europe contrasts with widespread democratic dissatisfaction among European citizens. The most frequently used indicator for democratic evaluations is based on another general, straightforward question: On the whole, how satisfied are you with the way democracy works in your country? The answers vary from 0 (extremely dissatisfied) to 10 (extremely satisfied). Although this indicator has its problems, it is a rather good measure of citizens' assessment of democracy (Ferrin 2016). We can now compare the average support for democracy and the average satisfaction with the way democracy works in the five regions. The resulting discrepancy between satisfaction and support has been called the "democratic paradox" (Blühdorn 2013, p. 36, 111;Dahl 2000). But there is nothing paradoxical about this discrepancy: In the real world, ideals are difficult to implement, and we are used to getting less than we expect. Citizens tend to be aware of the shortcomings of the democracy they are living in and, accordingly, they are dissatisfied with the quality of democracy they get in their own country (see . Table 3 presents the average expectations (importance) and evaluations (satisfaction) as well as the discrepancy between the two for the same five regions introduced above. On the left-hand side, the table presents the results for the two questions about democracy in general. On the right-hand side, it includes corresponding results for liberal democracy in particular. These results are based on similar questions about expectations and evaluations that have been asked for 12 specific aspects of liberal democracy in the ESS6, two of which I have already used for illustrative purposes in the discussion of support of democratic principles. The answers to the two sets of questions about these 12 aspects have been summarized in an index for support of liberal democracy and a corresponding index for satisfaction with liberal democracy (see Ferrin and Kriesi 2016).
As we can immediately see from Table 3, on average, not only expectations but also satisfactions with democracy in general and liberal democracy in particular are higher in the Nordic countries than in the rest of Europe, and both are particularly low in hybrid democracies. But even in the Nordic countries, reality falls short of expectations. Even in the Nordic countries there is room for improvement in terms of democracy in general and of liberal democracy in particular. But it is striking to observe that democratic satisfaction is on average much lower in southern and central-eastern Europe than in northern or continental Europe.
Returning to the differences between the age groups, Zilinsky (2019) has shown that, across Europe, there are no age differences in terms of democratic satisfaction. Evidence from the ESS cumulative file covering the period from 2000 to 2016 does not reveal any "patterns consistent with the notion that young citizens are less satisfied with democracy than middle-aged or older citizens" (p. 6).

The Origins of Democratic Dissatisfaction
Where does the widespread democratic dissatisfaction come from? Arguably it is the consequence of a crisis of representation, i.e., of an inadequate representation of the citizens' demands in the political system. The crisis of representation, in turn, may result from two sets of factors. On the one hand, it is the result of a lack of responsiveness of the political system, most importantly of the party system, to new demands of the citizens. The citizens' demands are linked to broad societal conflicts. In Western Europe, these conflicts were traditionally based on religion and class, and to some extent also on regional differences. More recently, however, we have seen the rise of new structural conflicts linked to processes of globalization, which can be broadly understood as the opening up of national borders in varying ways. Thus, the increasing international economic competition, the increasing influx of migrants from ever more distant and culturally more different shores, and the increasing political integration in the European Union have created conflicts between "winners" and "losers" of globalization, i.e., between people whose life chances were traditionally protected by national boundaries and who perceive the weakening of these boundaries as a threat to their social status and their social security, and people whose life chances are enhanced by the opening up of national borders. Scholars have used different labels to refer to the new structuring conflict-from "GAL-TAN" (Hooghe et al. 2002), "integration-demarcation" (Kriesi et al. 2008), "universalism-communitarianism" (Bornschier 2010), "cosmopolitanism-communitarianism" (Zürn and de Wilde 2016; de Wilde et al. 2019), and "cosmopolitanism-parochialism" (de Vries 2017) to the "transnational cleavage" (Hooghe and Marks 2018). They agree, however, that the new divide is above all ar-ticulated in terms of two types of issues: immigration and European integration. For multiple reasons-programmatic constraints, internal divisions, cartelization, systemic imperatives, and the requirements of responsible government-mainstream parties have avoided these (and other hot) issues and have failed to respond to the demands of the globalization losers in particular (Blühdorn 2013;Green-Pedersen 2012;Hooghe and Marks 2018;Katz and Mair 2018;Netjes and Binnema 2007;Sitter 2001;Steenbergen and Scott 2004;de Vries and van de Wardt 2011). Those who have been "left behind" by the mainstream parties, i.e., the increasing number of citizens whose demands have not been adequately represented by the established parties, have become dissatisfied with the way democracy works in their own country. Arguably, this version of the crisis of representation is most characteristic of northwestern Europe, where it has increasingly contributed to democratic satisfaction over the last decades.
On the other hand, the crisis of representation also results from more traditional performance failures of party governments. Thus, the perceived poor performance of the economy and of the government in the course of the Great Recession led citizens across Europe to become more critical of the way their national democracy works-especially in the hard-hit regions of southern and central-eastern Europe and in hybrid democracies. In the Great Recession, the southern European countries suffered the consequences of a double crisis that combined economic and political factors (Hutter et al. 2018). All southern European countries experienced a sharp relative decline in economic performance, insufficient government responsiveness, a surge in economic grievances, and a downright collapse of satisfaction with democ- Fig. 3 Satisfaction with the way democracy works in one's own country/at the European level: shares of very/fairly satisfied citizens in four European regions, 2000-2015. a SWD-national. b SWD(satisfaction with democracy)-European racy. The crisis quickly disclosed the inability of the political elites to implement effective countermeasures against growing economic grievances (Beramendi et al. 2015, p. 13), fueling a political crisis that erupted in massive protest waves (Kriesi et al. forthcoming). In central-eastern Europe, the low level of political and administrative performance and the corresponding high level of corruption had contributed to widespread dissatisfaction with the way democracy works in these countries long before the great economic crisis set in (Pop-Eleches 2010, p. 232). In times of economic crisis, however, the tolerance for corruption that previously benefited from the good economic performance diminishes considerably, as is illustrated by the Klasnja and Tucker (2013) experiments in Moldavia. Figure 3, which shows the development of satisfaction with national and European democracy from 2000 to 2015 for four of the five regions, documents the chronic dissatisfaction with national (but not European) democracy in central-eastern Europe and the precipitous fall of satisfaction with both national and European democracy in southern Europe during the Great Recession.

The Voters' Perspective: The Rise of Challenger Parties
It is now the moment to adopt the third perspective, the perspective of the voters, in order to document the consequences of democratic dissatisfaction in electoral terms. As a reaction to both versions of the crisis of representation, voters abandon mainstream parties and turn to new challengers. Thus, the decline of the mainstream parties' capacity to represent the key societal conflicts and their performance failure has made room for the rise of new challengers in the party system. Mair (2002, p. 88) has already pointed to the link between, on the one hand, the weakening of party democracy as we knew it and, on the other hand, the rise of such new challengers. These new challengers not only articulate the new conflicts, but they also turn more explicitly against the established political elites who constitute their direct adversaries in the electoral competition. In other words, they express both the substantive demands of voters who have not been taken into account by the mainstream parties and their political dissatisfaction. While their "host" ideology connects these parties to the fundamental structural conflicts in society, the "thin" populist "ideology" connects them to the more narrowly political sphere and to the political discontent of their constituencies. More specifically, the populist "ideology" refers to the tension between "the elites" and "the people." This "ideology" puts the emphasis on the fundamental role of "the people" in politics, claims that "the people" have been betrayed by "the elites" in charge who are abusing their position of power, and demands that the sovereignty of the people be restored (Mudde 2004;Mény and Surel 2002, p. 11 f.). In their analysis of the vote for radical right parties in 16 western European countries in the 1990s, Lubbers et al. (2002) have shown that this vote is enhanced by political dissatisfaction. More recently, Akkerman et al. (2014Akkerman et al. ( , 2017 have shown for the Netherlands that individuals with pronounced populist attitudes are more likely to vote for populist parties from the right (PVV) and the left (SP), a result that has been confirmed by the cross-national study in nine European countries of van Hauwaert and van Kessel (2018).  Figure 4 presents the development of mainstream parties and of electoral volatility for four selected countries from each of three regions of Europe-central-eastern, continental, and southern Europe-during the Great Recession 1 . The results presented refer to the average differences in vote shares of mainstream parties in precrisis and post-crisis elections (averaged across countries) and to the corresponding change in overall volatility. As the figure shows, the mainstream parties lost vote shares in all three regions during the Great Recession, but their losses were by far the greatest in southern Europe, where (with the exception of Portugal) they lost massively. The increase in the overall electoral volatility tells much the same story: As the mainstream parties lost, volatility shot up in southern Europe, and it also increased in continental Europe. In central-eastern Europe, however, volatility decreased slightly. While the western European party systems increasingly fragmented, the central-eastern European party systems started to consolidate, albeit at a very high level of overall volatility. Figure 5 presents the trends for the rise of radical left and radical right parties in western Europe (given the lack of institutionalization of central-eastern European party systems, corresponding trends are less informative). It presents the respective regional averages for four 5-year periods-two preceding the onset of the Great Recession and two following the fall of Lehman Brothers in September 2008. This deceptively simple graph makes two assumptions, which need to be pointed out. First of all, it excludes all countries where the radical right or the radical left has not participated in any significant way in the national elections (this excludes Iceland, Ireland, and Luxemburg in northwestern Europe and all but two countries (Italy and Greece) in southern Europe for the calculations of the averages for the radical right, and Switzerland, the UK, and Malta for those of the radical left). Including these countries would heavily reduce the trends indicated. The second assumption is that the Italian Movimento 5 Stelle (M5S; Five Star Movement) can be considered a radical left party. Of course, M5S considers itself "neither left nor right," and most observers would not rank it among the radical left. However, even if it is not a party on the radical left, it is nevertheless the functional equivalent of the new radical left in Italy. Keeping these two points in mind, the graph has a very clear message: It is above all in southern Europe that the great economic crisis boosted the rise of radical challengers-mainly from the left, but also from the right in Italy and Greece. In northwestern Europe, the radical left stagnated throughout the more recent past, while the radical right benefited from the economic crisis and also from the more recent refugee crisis.

The Double Logic of the Rising Challengers
As already pointed out, the rising challengers pursue a double logic: They not only express and fuel their voters' democratic dissatisfaction, but they also express and fuel their voters' substantive concerns. Based on an analysis of the cumulative file of the European Social Survey (ESS) that includes the first eight rounds (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016), it is possible to show that political dissatisfaction and substantive concerns each have had an independent effect on the electoral success of radical right and radical left challengers (Kriesi and Schulte-Cloos forthcoming): In substantive terms, the K radical right vote is mostly driven by the voters' critical attitude toward immigration but also to a smaller extent by their Euroscepticism, while, strikingly, both cultural liberalism (opposition to gay rights) and egalitarian economic attitudes have no effect on the radical right vote net of individuals' opposition to immigration and Euroscepticism. By contrast, the radical left vote is mainly driven by egalitarian economic attitudes. Euroscepticism also contributes to the radical left vote, as do culturally liberal and pro-immigration attitudes. In addition to the substantive concerns, political dissatisfaction substantially contributes to the vote for both types of radical parties. While the substantive effect of anti-immigration attitudes is even stronger than the effect of political dissatisfaction for the radical right vote, political dissatisfaction has the largest effect on the radical left vote. In the case of M5S, a "pure" case of a populist party, which at the time of its breakthrough in 2013 hardly campaigned on substantive grounds but almost exclusively focused on its anti-elitist message, political dissatisfaction even became the overpowering factor in determining the vote. On the one hand, this means that the factors related to long-term social change contribute to the determination of the vote for the radical challenger parties from both left and right, independently of the political dynamics, i.e., independently of their populist rhetoric and independently of other types of short-term political considerations. In the final analysis, these parties are driven by the structural conflicts of society, and as long as these conflicts exist, the electoral fortune of these parties will be a function of the substantive concerns linked to these conflicts. On the other hand, this also means that the impact of political dissatisfaction is independent of these long-term structural factors of change. In other words, if the political dynamics change, the populist rhetoric of these parties is likely to change as well, and political discontent may no longer fuel their electoral success. Given the independence of the impact of political dissatisfaction from structural conflicts, it is important to highlight that the role of the populist appeal to political dissatisfaction varies depending on the political context.

Is the Rise of Populist Challengers Dangerous?
Does the rise of populist challengers undermine democracy in Europe? Whatever their faults, and there are many as is well known, the populist parties' "professed aim is to cash in democracy's promise of power to the people" (Canovan 1999, p. 2). As Canovan has pointed out, it is this promise of the "redemptive face" of democracy that creates the tension with its pragmatic face and with liberalism. Populism is both a "politics of resentment" (Betz 1993)-a response of the weak to the wrongdoing of the powerful-and a "politics of hope" (Akkerman et al. 2017, p. 380) that "yearns for a more direct and unmediated relationship between citizens and their political representatives." Mair (2013) has conceptualized the discrepancy between the "pragmatic" and the "redemptive" face of democracy as the tension between "responsibility" and "responsiveness": While the mainstream parties emphasize responsibility to a complex set of stakeholders, the challenger parties from left and right emphasize responsiveness to the voters. Distrust of the political elites and dissatisfaction with the lack of responsiveness, and the lack of unmediated direct democracy, contribute, as we have seen, to the surge of these challengers. Although critical of these parties, Mounk (2018, p. 52) does acknowledge the democratic energy of populism, and he suggests that the refusal to do so will stop us from understanding the nature of their appeal.
We can go even one step further: As Rovira-Kaltwasser (2012) as well as Taggart and Rovira-Kaltwasser (2016, p. 346) have argued, populism can work as a democratic corrective. Populists may develop a corrective force, especially when they represent demands and claims of structurally important groups that have been neglected or "depoliticized" by the established mainstream parties, i.e., if they hold out the promise of "salvation through politics" for unrepresented sectors of society. They expand the political agenda and debate by putting issues and conflicts on the agenda of democratic politics that have been neglected by the mainstream parties. They increase the electoral options and belie the mainstream parties' claim that "there is no alternative" (TINA). They may increase electoral participation by bringing back into democratic politics groups of the population that were left unrepresented by the mainstream parties in the past. Thus, as a result of their mobilization, electoral participation may increase, as has happened in German regional and national elections as a result of the rise of the AfD (Alternative for Germany) party, and as happened in the British Brexit referendum (Evans et al. 2017). If these examples are suggestive, a more systematic analysis of the impact of populist parties on voter turnout by Leininger and Meijers (2017) does, however, not confirm that populist parties systematically spur turnout.

The Elite's Perspective: Populists in Power
The final perspective is the elite's perspective. It deals with populism in power. Populism is, of course, not only a corrective of democracy, but it is also a potential danger for democracy. The danger, however, emanates less from the citizens who vote for the populist parties than from the political elites. On the one hand, it stems from the populist leaders, and on the other hand from the way the other members of the political elite react to successful populists. Let us first look at the populists themselves. As Urbinati (2014) suggests, populism is ultimately more than a "thin" ideology. It requires both an "organic polarizing ideology" and a leader who mobilizes the masses in order to govern in the name of "the people." Combined, the two elements amount to a project of political renewal that seeks "to redress democracy by taking it back to its 'natural' roots" (p. 151). Although it starts as a phenomenon of mass discontent and participation, populism is at the same time also "strategic politics of elite transformation and authority creation" (p. 157). The populist leader uses polarization between "the people" and the elite in order to create the unity of the people that is at the core of populist beliefs, and in order to implement the specific populist, illiberal vision of democracy.
When populists are in opposition, they have to forge an electoral majority that will enable them to rise to power. But as Pappas (2019) observes, once they pass from opposition to power, they face an altogether different task-"they have to serve their people, lest the populist electoral alliance be shattered." Pappas summarizes the inner logic of their rule as "state seizure and institution bending," "political patronage," and "uncompromising polarization." Based on the experience of two Latin American countries (Peron's Argentina and Chavez's Venezuela) as well as of two European countries (Pasok's Greece and Orbán's Hungary), Pappas observes that state seizure consists of state colonization by loyalists; empowerment of the executive at the expense of the other branches of state power; emasculation of checks and balances; control over the media, the judiciary, and the educational system; and, ultimately, regime change. The populists' political economy has been analyzed by Dornbusch and Edwards (1991) based on Latin American cases. Pappas concludes that the main ingredients of populist economics are expansive macroeconomic measures, nationalization of chunks of the economy, and radical redistribution of income in favor of the populists' own constituents, which in turn creates patron-client networks. This explosive mixture is bound to lead to economic difficulties, and even to economic collapse. The uncompromising polarization in words and deeds is illustrated by the example of Trump, which indicates how unrelenting norm-breaking contributes to polarization. As Levitsky and Ziblatt (2018, p. 193) have observed, even if Trump does not directly dismantle democratic institutions, his norm-breaking is almost certain to corrode them. Schedler (2018) suggests that the central lesson of Trump concerns the weight of political language. Trump's most severe democratic norm violations have been discursive, and, Schedler argues, Trump's public discourse is what is dangerous about him. From in-depth case studies, we know that discursive transgressions tend to spill over into behavioral transgressions and that both tend to unfold in vicious spirals (Bermeo 2003).

Constraints On the Populists' Project of Power
How far populists in power will get in implementing anything like the agenda sketched by Pappas depends very much on a set of constraints that I would like to classify into five categories. First, there are the constitutional constraints. In the United States, Donald Trump has (to some extent) been checked by the courts and by Congress. The institutional arrangements in the United States and also in the democracies of western Europe are likely to inhibit the more extreme forms of populist autocracies that we know from the presidential systems of Latin America (Roberts 2017). In the less established democracies of central-eastern Europe, however, populists in power meet with less institutional resistance and use their power to implement illiberal institutional reforms, as is illustrated by the Polish and Hungarian populists in power. The same has applied to Berlusconi's governments in Italy, which have often adopted illiberal measures regarding the checks and balances of Italian democracy (such as media freedom, the judiciary, the constitution, and the president of the republic; Bobba and McDonnell 2015).
The electoral system provides a little noticed but crucial institutional constraint. While majoritarian (or nearly majoritarian) systems allow populists to gain power undivided, proportional systems are likely to force populists to share power with coalition partners who are likely to be mainstream parties. This kind of power sharing goes a long way to moderate both populist parties and their voters. Thus, when the Austrian FPÖ entered the government dominated by Wolfgang Schüssel's ÖVP in 2000, it was seriously weakened by the experience. Its cabinet members adopted more moderate positions, which led to an eventual split between the moderates and the radicals, with the moderates creating a new party, the BZÖ (Luther 2015, p. 143-5). Similarly, the Swiss SVP moderated its populist discourse once its leader was co-opted into the government in 2003, and it also split over the government experience of its leader (Bernhard et al. 2015). Rooduijn et al. (2014) show that, after an electoral success, populist parties generally become less populist in their party programs.
It is when populists gain power undivided and get an opportunity to implement their project of political renewal that they may become a major threat to democracy. In such less favorable institutional circumstances, the strategies of the populists' own parties and of related parties in their own camp become particularly important. Thus, for Levitsky and Ziblatt (2018, p. 7), an essential test for democracies is not whether such populist leaders emerge but "whether political leaders, and especially political parties, work to prevent them from gaining power in the first place." They are afraid that institutions-written constitutions and norms of mutual toleration and forbearance (i.e., patient self-restraint)-are not enough to rein in elected autocrats. For them (p. 19), parties are the essential gatekeepers. Successful gatekeeping requires that mainstream parties isolate and defeat extremist forces. It is "the abdication of political responsibility by existing leaders" that often marks a nation's first step toward authoritarianism.
In this respect, it is worthwhile to remind ourselves of Linz's (1978) analysis of the breakdown of democracy in the interwar period: He distinguished between loyal, semiloyal, and disloyal oppositions in democratic regimes. Fascists and communists were the disloyal oppositions, but most interesting from the contemporary point of view are the semiloyal opponents from less radical parties: Linz characterized them as willing to encourage, tolerate, cover up, treat leniently, excuse, or justify actions of other participants in the political process that go beyond the limits of peaceful, legitimate patterns of politics in democracy. Ultimately, he identified semiloyalty by its greater affinity for radicals on its own side of the political spectrum than for the supporters of the democratic principles. Thus, in the Weimar Republic, the semiloyal opposition contributed to the breakdown of the system by seeking the support of the disloyal opposition and by helping it into power. Building on Linz's analysis, I would like to suggest that the politicians of the populist leaders' own party and of related parties in their own camp become critical for his or her maneuvering space. To the extent that they condone the excesses of the maverick in power, they crucially contribute to the danger he or she poses for democracy. This is what has happened, tragically, in the United States, where most Republican leaders closed ranks behind Trump, and the election became a standard two-party race (Levitsky and Ziblatt 2018, p. 70, 201). Trump's deviance has been tolerated by the Republican party, which has helped make it acceptable to much of the Republican electorate. And by spring 2018, Trump's takeover of the party's institutions has been largely complete. Another instance of this mechanism is the toleration of Victor Orbán within the ranks of the European People's Party. By contrast, after many years of toleration, to be sure, the South African ANC has succeeded in ousting its maverick leader Jacob Zuma and replacing him with a democratic leader, Cyril Ramaphosa.
The third constraint is the international constraint. Following Levitsky and Way (2010), external leverage and linkage may be key factors for keeping populists in power in check. However, the recent survey of a number of key cases of populists in power suggests that "international actors do have a difficult time when it comes to dealing with populists in power" (Taggart and Rovira-Kaltwasser 2016, p. 356). This is related to the fact that international institutions such as the EU or the Organization of American States have a limited tool kit for defending democracy. In the case of the EU, membership is conditional on meeting several democratic conditions. However, once a country has become a member, the EU has limited capacity to impose compliance with democratic rules and procedures. This has been illustrated by the so-called Haider affair, which deeply traumatized the EU (Müller 2013, p. 139). It is equally illustrated by the way the EU has been dealing with Orbán's Hungary and Kaczyński's Poland. Moreover, as Taggart and Rovira-Kaltwasser (2016, p. 356) suggest, outside interventions may backfire, as they give populists in government the opportunity to denounce coalitions between domestic and foreign actors that undermine the will of the people. This is perfectly illustrated by the Greek referendum in 2015, when the Greek voters turned down the austerity package offered by the international actors, although they were under great pressure to accept it.
A fourth constraint is the constraint imposed by the markets. Thus, during the final phase of the protracted Italian government formation in 2018, the markets reacted sharply when the leaders of the two populist parties submitted a list of prospective ministers to the president that included a finance minister who was an explicit opponent of Italy's membership in the Eurozone. When the populists backed down and were prepared to compromise, the markets calmed down to some extent, too. The sanctioning potential of the markets seems to be much bigger than that of international political actors.
Finally, the most important constraint is constituted by the voters. Aspiring despots may be beaten at the polls. Of course, it is by no means certain that voters will vote against populists in power. As Schedler (2018) points out, it has been quite puzzling and disturbing "to see majorities (or at least pluralities) of voters supporting, once and again, illiberal governments who have been, step by step, dismantling their democratic rights and liberties." Schedler suggests that voters might care more about other values (such as social justice, religious piety, security, or the national soul) than democracy; they may fail to see that democracy is being taken apart before their eyes; or they may have competing conceptions of democracy. However, let us not forget that liberal democracy is strong, because, as Galston (2018) suggests, to a greater extent than any other political form, it harbors the power of self-correction. It is possible that voters will unseat aspiring autocrats. Armenia, Malaysia, the Maldives, Turkey, and Ukraine provide recent examples from outside Europe of the voters' power. In April 2018, a protest movement succeeded in ousting Serzh Sargsyan, who had ruled Armenia for the past decade, replacing him with Nikol Pashinyan, a journalist turned lawmaker. Meanwhile, in Malaysia's general elections in early May 2018, a broad coalition of voters ousted Prime Minister Najib in spite of his shameless attempts to rig the elections and hold on to power. Malaysia's citizens have finally been able to reassert their right to change the way the country is run. In the case of Malaysia, the sentiments reflect "less a revolution than a restoration"-a return to an earlier era of settled law, fair judges, and democratic accountability that has survived in the national imagination (The Economist, May 17, 2018). In September 2018, the opposition leader Ibrahim Mohamed Solih was elected to replace the authoritarian incumbent in the Maldives. The Turkish local elections that were equally held in spring 2019 indicate that voters are able to impose limits on the autocratic ambitions of President Erdogan.
In Europe, we need to distinguish between regions. In northwestern Europe, where the radical right has been the main populist force, its illiberal potential has been curbed so far by institutional and partisan constraints-the division of power by proportional electoral systems and the need to form coalitions when getting into power. Moreover, the radical right constituency became less dissatisfied in all the cases where these parties got into power (see Kriesi and Schulte-Cloos forthcoming;Haugsgjerd 2019). In southern Europe, where the radical left has been the main populist force, its illiberal potential has also been limited so far by institutional and partisan constraints-proportional electoral systems and the need to enter into coalition with other parties, as well as by market constraints. The partisan constraints worked even in the case of the coalition of populists who governed Italy in 2018/2019: Because of their incompatible programs, the coalition partners imposed limits on each other. In central-eastern Europe, there are fewer constraints on populists in power because the electoral systems allow them to get undivided power in countries like Hungary and Poland. The market constraints as well as the international constraints have also proved to be weak so far. However, democracy's power of self-correction may also assert itself in this part of Europe. Thus, ZuzanaČaputová's victory in the presidential elections in Slovakia in March 2019 provides a signal that liberal democracy is recovering central-eastern Europe. And the Ukrainian presidential elections in April 2019, which were won by a complete outsider, confirm the power of self-correction that voters wield even in hybrid democracies.

Conclusion
There clearly are dangers for democracy, but let us underscore with Galston (2018) a less fashionable point: This is no time for panic. As we have seen, modernization theorists concur. They believe that democracy, in the long run, will impose itself as a result of the spread of fundamental values that underpin it, even if we find some short-term limited reversal of the long-term progress of democracy. In Europe in particular, there is no reason to dramatize. Among the European citizens, the principles of democracy are widely supported, even if dissatisfaction with democracy is widespread-especially in the European south and the European east. To be sure, there is widespread indifference and alienation in different parts of Europe, in particular among the youngest generations, which is indeed preoccupying. However, there are also large groups of critical citizens who combine support of democratic ideals with dissatisfaction with the existing democracies of their own countries, especially in southern and central-eastern Europe (see Kriesi and Vidal 2020). It is these critical citizens who may reassert themselves and exercise the self-corrective power of democracy.
Democratic dissatisfaction contributes to the rise of radical challenger parties, especially on the radical left in southern Europe but also on the radical right across western Europe. These parties express widespread dissatisfaction with democracy and contribute to it by their populist discourse. The good news is that across western Europe, once these parties are in government, democratic dissatisfaction seems to evaporate among their voters, and they arguably become parties like any other mainstream party, provided they do not gain power undivided. Most dangerous are situations where populist challengers do not have to share power in government. In western Europe, parliamentary democracies and proportional electoral systems tend to guard against populists gaining undivided power. The Italian case shows that populists of different persuasions may join forces to govern, but it also suggests that they may not be able to govern effectively. The institutional context is less favorable for democracy in central-eastern Europe, where semipresidential systems (Poland, Romania) and electoral systems heavily skewed in favor of the largest party (Hungary) provide the possibility for populists to gain undivided power. Other safeguards-partisan, international, and market constraints-may be less reliable than often hoped for. In the final analysis, it is the voters who may resort to sanctioning aspiring autocrats in this part of Europe as well.
Funding Research for this paper has been supported by the EU-grant 810356 -ERC-2018-SyG.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/.