THE SOCIAL DYNAMICS OF POLITICAL AND ECONOMIC CRISES IN THE TWENTY-FIRST CENTURY A Dissertation Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy by Sang Kyung Lee August 2022 © 2022 Sang Kyung Lee THE SOCIAL DYNAMICS OF POLITICAL AND ECONOMIC CRISES IN THE TWENTY-FIRST CENTURY Sang Kyung Lee, Ph.D. Cornell University, 2022 My dissertation research investigates the broad consequences of global crises that have fundamentally transformed the social world over the last few decades. The first two decades of the twenty-first century witnessed profound crises in a wide range of social lives that materialized in economic, political, and cultural dimensions. Particularly in the economically advanced world, the multiplicity of crises was salient: countries in Western Europe and North America were not only hit hardest by the Great Recession in the late 2000s but also confronted with serious challenges rooted in the multicultural turn of the society and declining public confidence in political establishment. I examine social changes that resulted from those crises, focusing on how various actors and groups reacted to the crises and what transformations the crises brought for the political systems and social world. This dissertation comprises three chapters. The first chapter focuses on the economic crisis with particular attention to the Great Recession that broke out in the late 2000s. The second chapter examines more complex dynamics of political crises. Specifically, I explore the rise of populism, which can be best understood as a manifestation of multiple crises that encompassed economic, political, and cultural shifts that have occurred across the world in recent years. The final chapter explores the consequence of the economic and political crises for our contemporary society. I devote particular attention to social integration, which has seriously weakened in years of the prolonged economic recession and populist surge. BIOGRAPHICAL SKETCH Sang Kyung Lee was born and raised in Seoul, South Korea. Before starting the Ph.D. program in the Department of Sociology at Cornell University, he received a M.Sc. in Political Sociology from the London School of Economics in 2011 and a B.A. in Sociology from Yonsei University in 2009. At Cornell, his research focused primarily on politics, political economy, and theory. His latest work investigates the global dynamics of economic and political crises that raised fundamental challenges for democratic governance over the past few decades. In the Fall of 2022, he will begin an appointment as Postdoctoral Associate at the Elliott School of International Affairs, George Washington University. iii For my father iv ACKNOWLEDGEMENT It was a great pleasure to be surrounded by many brilliant people during my time at Cornell. I am very grateful to the members of my committee. My chair, David Strang, has always been supportive of my research and scholarship with great encouragement and challenges. David has been a great supporter for my scholarship as well as a tireless critic of my work, helping me to move further whenever I tended toward complacency. Richard Swedberg enriched my academic life in graduate school more than anyone. Oftentimes conversations I had with Richard not only inspired my new projects but also helped to redirect my research in many fruitful ways. I cannot thank Richard enough. Peter Enns has been the greatest reviewer of my work equipped with constructive critiques and practical suggestions. I could always rely on Peter for both academic feedback and professional advice. Ben Cornwell has also been valuable part of my committee. Ben was always full of helpful scholarly and professional advice. I benefitted from Ben on many occasions. I would like to thank scholars outside my committee who helped me to advance my research at different stages. I am very grateful to Jeff Haydu, Isaac Martin, Ken Roberts, and Sid Tarrow. Jeff and Isaac taught me a great deal about social movements and contentious politics and helped me a lot to develop my own research projects. Ken and Sid provided me with constructive comments on my recent research on populism. I also thank Rick Biernacki, Harvey Goldman, Kwai Ng, and Ian Mullins for their support for my scholarship. I also cannot exaggerate how lucky I was at Cornell to have incredibly intelligent fellow graduate students. I would particularly v like to thank Radu Parvulescu, Yunsub Lee, Tom Davidson, Mario Molina, Lucas Drouhot, Tony Sirianni, Ben Rosche, and Abduallah Shahid. My family members have been passionate supporters for my academic career. I cannot be more proud of my father, Gyoungwoo Lee, and my mother, Heesook Woo. My thanks should also go to my siblings, Sanghyun and Sunhwa. Finally, I owe the most to two humans. My wife, Jeonghwa Yoo, was the one who made my academic journey possible. Without her limitless support and incredible patience over the last seven years, I would not have been able to carry on my research. Her love and presence always stood behind any success I have made. Our son, Yegan, brought wonderful joy to our life. While joy often came along with difficult challenges, he became the greatest source of happiness and the reason for existence for our family. vi TABLE OF CONTENT Bibliographical Sketch………………………………………………………...………iii Dedication……………………………………………………………………………..iv Acknowledgement……………………………………………………………………..v Table of Content……………………………………………………………………...vii List of Tables………………………………………………………………………….ix List of Figures………………………………………………………………………….x Introduction……………………………………...…………………………………....1 Chapter 1. The Politics of Anti-Austerity Protest: South Korea 1997-1998 and Greece in 2009-2010 Intro……………………………………………………………………….…………....5 Theories of divergent anti-austerity movements……………………………………………………………………………..8 South Korea in 1997-1998 and Greece in 2009- 2010…………………….……………………………………………………………..15 The Origins of the Divergent Anti-Austerity Movements……………………………………………………………………...…….30 Concluding Remarks……………………………………………………..………………………..44 References………………………………………………………...…………………..47 Chapter 2. The Origins of the Populist Surge Intro……………………………………………………………………….…………..56 The Social Origins of Populism……………………………………………………………….……………...62 Conceptualizing Populism……………………………………………………………….……………...69 Data and Methods……...……………………………………………………….……………….76 Results………...…..………………………………………………….……………….87 Conclusion……....………………………………………………….……………….103 References…………………………………………………………………………...111 Appendix……………………………………...……………………………………..122 Chapter 3. Toward A New Theory of Social Integration: Durkheim’s Theory of Solidarity Revisited Intro………………………………………………………………….……………....128 Durkheim’s Theory of Social Integration……………………………………………………………...……………128 Problems: Weakening Interdependence And Growing Polarization………………………………………………………………………….134 vii Toward An Alternative Theory Of Solidarity…………..………………………………………...……………………....141 References…………………………………………………………...………………145 Concluding Remarks and Avenues for Future Research…………..…….…………………………………………………..……....148 viii LIST OF TABLES Chapter 1. Major macroeconomic indicators: Korea in 1994-98 and Greece in 2006-10……..…24 Chapter 2. Fractional response probit models of far-right populist support……………………...89 Fractional response probit models of far-left populist support……………….........…95 Fractional response probit models of non-radical populist support…………………101 Descriptive statistics and data sources………………………………………………122 Populist parties (acronyms) from each party family in 27 European countries……..124 Models estimated with the measure of immigration inflows: selected outcomes…………………………………………………………………………….125 Models estimated with the entire sample: selected outcomes………………………126 ix LIST OF FIGURES Chapter 1. Employment rates: Korea in 1994-98 and Greece in 2006-10 …………………….…27 Annual changes (%, real) in average earnings: Korea in 1994-98 and Greece in 2006- 10……………………………………………………………………………………...28 Annual changes (%, real) in gross disposable income: Korea in 1994-98 and Greece in 2006-10…………………………………………………………………………….…29 Annual changes (%, real) in median household disposable income: Korea in 1994-98 and Greece in 2006-10……………………………………………………………..…29 Chapter 2. Average parliamentary vote share (%) of populist parties in EU27 and UK in the 1990s, 2000s, and 2010s…………………………………………………………...…77 Parliamentary vote share (%) of each populist family in Greece, Poland, and Slovakia during 1990-2019…………………………………………………………………..…80 Marginal effect of public concern about immigration on far-right populist support conditioned by unemployment……………………………………………………..…91 Marginal effect of unemployment and poverty gap on far-left populist support conditioned by political distrust………………………………………………………94 Marginal effect of political distrust on far-left populist support conditioned by unemployment and poverty gap………………………………………………………97 Marginal effect of public concern about immigration on non-radical populist support conditioned by poverty gap and Gini index…………………………………………100 x Introduction My dissertation research investigates the social consequences of economic and political crises that have fundamentally transformed the social world over the last few decades. The first two decades of the twenty-first century witnessed profound crises in a wide range of social lives that materialized in diverse social dimensions. Particularly in the economically advanced world, the multiplicity of crises was salient: countries in Western Europe and North America were not only hit hardest by the Great Recession in the late 2000s but also confronted with serious challenges rooted in the multicultural turn of the society and declining public confidence in political establishment. Crises have consistently drawn attention from sociologists (Giddens 1986; Swidler 1986), as they lead to a discontinuation of the status of quo of the current society and in turn pave the way for new social systems. As such, a societal transition triggered by a crisis tends to involve a variety of responses from social actors and groups, as the impact of a crisis manifests differently across individuals and groups (Sorokin 1942). At the heart of the sociological study of the social dynamics of a crisis, therefore, is investigating how various actors and groups reacted to the crisis and what transformations the crisis brought for the political systems and the social world. This dissertation comprises three chapters. The first chapter focuses on the economic crisis with particular attention to the Great Recession that broke out in the late 2000s. The second chapter examines more complex dynamics of political crises that manifested throughout the first two decades of the twenty-first century. 1 Specifically, I explore the rise of populism, which can be best understood as a manifestation of multiple crises that encompassed economic, political, and cultural shifts that have occurred across the world in recent years. The final chapter explores the consequence of the economic and political crises for our contemporary society. I devote particular attention to social integration, which has seriously weakened in years of the prolonged economic recession and populist surge. Chapter 1 studies the social consequences of the economic crisis by examining how social actors reacted to the crisis. The focus is on varying responses to the crisis. I undertake a comparative analysis of Greece in the late 2000s and Korea in the late 1990s, which faced the Eurozone financial crisis and the East Asian financial crisis, respectively. These two cases were chosen, because the financial crisis and austerity mandates prescribed for recovery provoked vastly different reactions from the public. Analyzing anti-austerity protest movements in the two cases, I seek to identify the social processes that gave rise to the different protest movements. Chapter 2 shifts attention to populism, a political manifestation of the multifaceted crisis that broke out in the twenty-first century. This chapter explores the social origins of populism drawing on an up-to-date harmonized dataset of populist parties in 27 European democracies that covers a span of 17 years (2003-2019). Europe over the last two decades offers a great laboratory for populism research, as it was the place where populist parties made unprecedented electoral success simultaneously in many countries. I argue that in order to fully understand changing support for populist parties, we need to consider diverse types of populist parties separately. Specifically, I break populist parties into three mutually exclusive families 2 – far-right, far-left, and non-radicals – and test whether and to what extent the existing theories account for changing support for each populist party family. This decomposition is grounded in the ways in which populism is exercised in mass politics. I will also compare the outcome of the analysis with the recent voting patterns for populists in the U.S., drawing attention to Sanders and Trump in 2016. Chapter 3 discusses the broad consequences of the economic and political crises for our contemporary society. Where the economic and political crises materialized in the global financial crisis and the rise of populism and illiberal politics over the last few decades, we have witnessed ever-growing conflicts and antagonism between diverse social groups. This inspired an explosion of literature on social conflicts and political polarization. Yet, how such conflicts and polarization can be managed has seldom been studied since Durkheim’s original formulation of social integration. In this chapter, I first provide a critical review of Durkheim’s theory of social integration and then propose a set of directions to renew his theory to tackle the novel challenges raised in our contemporary society. REFERENCES Giddens, Anthony. 1986. The Constitution of Society: Outline of the Theory of Structuration. 1. paperback ed. Berkeley: Univ. of California Press. Sorokin, Pitirim Aleksandrovich. 1942. Man and Society in Calamity: The Effects of War, Revolution, Famine, Pestilence upon Human Mind, Behavior, Social Organization and Cultural Life. New York: E.P. Dutton. Swidler, Ann. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(2):273–86. 3 Chapter 1. The Politics of Anti-Austerity Protest: South Korea in 1997-98 and Greece in 2009-10 Abstract Why does austerity confront varying degrees of popular resistance? While prior research primarily addresses the economic threat of austerity as the stimulus for anti- austerity protest, a growing volume of studies highlights the roles of external environments in moderating the mobilizing effect of austerity. This study challenges the recent literature on the moderating role of external environments, pointing out that it tends to overlook the distinction between structural and contingent aspects of external environments. I undertake a paired comparison of carefully chosen cases, South Korea in 1997-98 and Greece in 2009-2010, to examine the moderating roles of the two aspects of external environments – each aspect is characterized through the notion of political opportunity drawing on Rootes (1999). Results reveal that structural and contingent opportunities played distinct roles in promoting (Greece) and hampering (South Korea) the growth of anti-austerity movements, initially triggered by the economic threat of austerity in both countries. 4 When enforced in response to a debt crisis, austerity imposes huge economic burdens on the livelihoods of the public, because many individuals, already suffering substantial material losses, come to not just lose even more (cuts in social spending and more losses in jobs, earnings, and properties) but pay more (tax hikes). It should come as little surprise, therefore, that austerity is often accompanied by mass protests when it is implemented to deal with fiscal challenges. Why then does austerity face divergent protest movements? Much of earlier research highlights the economic threat of austerity, linking it up with the observed protest events. This view consistently appears in studies of earlier debt crises, such as the Latin American crisis in the 1970-80s (e.g., Bernal 1984; Kaufman 1986; Walton 1989) and the East Asian crisis in the late 1990s (e.g., Haggard 2000; Walton and Seddon 1994). Recent studies – notably, those of the Eurozone crisis (e.g., Bermeo and Bartels 2014; Giugni and Grasso 2016) – also stress the severity of economic hardship generated by austerity. Some other empirical analyses offer evidence that harsher austerity provokes stronger protest (Martin and Gabay 2013; Ponticelli and Voth 2011). This threat-centered view, however, faced a critical challenge from a growing volume of studies pointing out that the severity of austerity does not always match the size/duration of protest (Císar and Navrátil 2016; Kriesi 2014). At the heart of this challenge is that the mobilizing effect of austerity is mediated by factors exogenous to austerity. Critiques of the threat-centered view often draw on well-established factors, such as social class (Hylmö and Wennerhag 2016; Peterson, Wahlström, and Wennerhag 2015), relative deprivation (Grasso and Giugni 2016), and political trust 5 (Kriesi 2014), focusing on how these factors come into play alongside the economic threat of austerity to precipitate protest activities. The mediating roles of factors exogenous to austerity are documented well enough to show that anti-austerity movements can be fueled by austerity itself as well as these factors. What is missing, however, is the distinction between structural and contingent aspects of the exogenous factors (Rootes 1999). In effect, prior research tends to treat as a constant or given the factors external to the implementation of austerity that influence anti-austerity mobilization, so that these factors are often considered as fixed attributes of the social system. It is important to note, however, that the external factors are not a mere byproduct of the existing social structure, but they are also in part autonomous of this structure, often contingent on strategic choices made by protesters and the historical conjunctures in which the protest broke out (Rootes 1999). Neglecting those contingencies or reducing them to the structural attributes, therefore, risks false characterization of the observed causes of divergent anti-austerity movements (McAdam, Tarrow, and Tilly 2001). Examining the structural and contingent aspects of factors exogenous to austerity requires careful tracing of anti-austerity activism to reveal how each aspect unfolds and interacts with the economic threat of austerity at different phases of protest mobilization. Particularly suitable for this task is tracing and comparing multiple cases of anti-austerity movements that share a set of key attributes on the input side of anti-austerity mobilization (i.e., austerity enforcement) and differ from each other on the output side (i.e., protest activities). Cases matched in this way help to limit attention to the most critical external factor(s) that caused the divergent anti- 6 austerity movements, whereas controlling for circumstances that led austerity to be implemented the way it was. This research design allows us to rule out trivial findings – namely, those from a comparison of cases where the implemented austerity measures were of different kinds or austerity was enforced in vastly different political and/or economic contexts. The present study selects two cases – South Korea in 1997-98 and Greece in 2009-10 – that exhibit crucial similarities on the input side and diverge from each other on the output side. On the input side, austerity was implemented in very similar economic environments in South Korea and Greece, proposed as a remedy for an unprecedentedly severe debt crisis that broke out after years of steady growth. Next, austerity was implemented in highly analogous political environments, proposed and executed by the governing party as well as supported by the largest opposition party under the two-party system. Last, austerity came out of highly similar (perceived) international environments, as it was mandated by international bodies in both South Korea and Greece and the involvement of those bodies was widely perceived as foreign intervention in both countries. On the output side, austerity faced vastly contrasting protest outcomes in the two countries – see below for a detailed description of case selection. Comparing the processes of protest mobilization in this pair of cases, a method referred to as “dual-process tracing” (Tarrow 2010:238), the analysis presented below reveals how the enduring structures and contingencies moderated the mobilizing effect of austerity and eventually resulted in the contrasting protest outcomes in South Korea and Greece. 7 The rest of this paper is organized as follows. I begin with a theoretical review of existing literature on divergent anti-austerity movements. This review focuses on two pillars of literature: one highlights the mobilizing effect of austerity, and the other the role of external factors in moderating the mobilizing effect of austerity. After reviewing both pillars, I argue that research on divergent anti-austerity movements need to distinguish between structural and contingent aspects of external factors – characterized by Rootes (1999) through the notion of political opportunity – and consider the roles of both aspects in protest mobilization to offer a more comprehensive account of why anti-austerity movements evolve differently. Next, I introduce the two cases selected for this study, South Korea in 1997-98 and Greece 2009-10, and undertake a comparative analysis of anti-austerity movements in the two countries to unveil how the structural and contingent opportunities unfolded and shaped the paths of anti-austerity movements in the two countries. I close with concluding implications. THEORIES OF DIVERGENT ANTI-AUSTERITY MOVEMENTS (Policy) Threat of austerity In earlier studies of anti-austerity activism, especially those examining Latin American debt crisis in the 1980s and East Asian debt crisis in the 1990s (Bernal 1984; Haggard 2000; Kaufman 1986; Walton 1989; Walton and Seddon 1994), the severity of austerity was often associated with the large-scale revolts. While these studies are generally descriptive, the latest empirical analyses (Grasso and Giugni 2016; Martin and Gabay 2013; Ponticelli and Voth 2011) give causal accounts of how harsh 8 austerity results in mass protest movements, testing the association between the measured severity of austerity and the occurrence/size of protest activities. From the theoretical perspective, recent scholarship on policy-oriented protest provides an effective conceptual tool for analyzing how the economic threat of austerity causes protest. This new research was designed to explain how a change in public policy precipitates mass protest. The underlying logic is straightforward: when an enacted, or proposed, policy reform arouses feeling of a loss of security – whether material or immaterial – to individuals, it stirs up collective protest among those individuals, who are encouraged to attempt to avoid the loss by pressing the government into withdrawing the reform. A term policy threat was coined to describe such changes in public policy that generate this negative perception (Campbell 2003; Martin 2013; Reese 2005). Grounded on this basic idea, scholars of policy threat propose a concrete mechanism of policy-oriented protest. The argument contains two interrelated components: the attributability and simultaneity of policy threat. Above all, the literature stresses that policy threats are particularly prone to mass revolts because they give victims “a convenient and legitimate focal target” (Martin 2013:11) to blame, namely incumbent governments. Given that any changes in policy are made by none other than policymakers in office, so that any losses generated by those changes are directly attributable to them, it is evident that once facing unpleasant outcome – whether those involve assets or rights – under an enacted policy reform, individuals are motivated to act against those in office, trying to force them to repeal the reform. Next, it is also pointed out that policy threats are extraordinarily effective in breeding 9 a large number of protestors within a short period of time. The reasoning behind this is that any reform bills, once passed, “affect many people at once” (Martin 2013:11), since each bill identifies its target group – for example, those who become subject to a new tax code or who stop receiving social benefits – and applies to all within the target from its enforcement date. Accordingly, when a loss-generating bill is passed, its target population becomes transformed into a mass of potential protestors, sharing with each other grievances against the same object. Thus policy threats constitute a unique trigger for mass mobilization by enabling the threatened to single out a common enemy and also providing them with a large number of potential allies to fight against that enemy. Since austerity generates immediate losses – mainly, the reduction of earnings/social benefits and the rise of taxes – among individuals subject to them, clearly austerity fits well into the rubric of policy threat: the losses generated under an austerity reform would inspire simultaneous grievances against its executor, the incumbent government. Accordingly, the mobilizing effect of austerity can be tested by examining whether or to what extent the economic losses generated by austerity precipitated protest mobilization against the incumbent government. Looking beyond threat Meanwhile, a growing volume of literature directed attention to factors external to the enforcement of austerity. In particular, recent comparative studies inspired by the Eurozone crisis addressed a broad set of factors that provoke anti-austerity protest. Some (Hylmö and Wennerhag 2016; Peterson et al. 2015) show that social class is still 10 a major driving force of protest participation, offering evidence that traditional working-class individuals are more likely than any other social groups to participate in anti-austerity protest. Others devote attention to the role of environmental factors. Kriesi (2014) reveals that political corruption and public distrust in mainstream parties can further promote anti-austerity movements, drawing on cases of poorly trusted governments facing massive anti-austerity protest movements even before implementing austerity measures. Císar and Navrátil (2016) focus on party system, showing that large-scale anti-austerity protests are more likely to occur where the existing party cleavage is less susceptible to demands from protesters. The importance of institutional environment also receives support from Beissinger and Sasse’s (2014) work that finds that anti-austerity mobilization is affected not only by the existing characteristics of political institutions but also by the configurations of economic institutions, such as the existence of a competent central bank and the relationship with the external control body (notably, the European Union). Gerbaudo’s (2017) analysis of southern European cases highlights the discursive dimension of anti- austerity mobilization, finding that pre-existing anti-regime discourses, when effectively combined with public discontent with austerity reforms, catalyze the spread of mass revolts. The emphasis on such factors does not necessarily go against the threat- centered approach. Instead, recent research suggests that the economic threat of austerity and external factors can jointly influence anti-austerity protest mobilization. Specifically, empirical studies point to the role of external factors in moderating the mobilizing effect of austerity. Bernberg’s (2015) case study of anti-austerity activism 11 in Iceland finds that while the economic threat of austerity played a mobilizing role, inspiring collective grievances among many individuals, it raised protest participation only among those who had a strong feeling of relative deprivation. Grasso and Giugni (2016) add to this finding, offering evidence that relative disadvantages, either perceived or economically measured, increase the propensity to take part in anti- austerity movements. There are also studies that offer macro-level perspectives. Kurer et al. (2019) report that economic hardship generated by austerity is likely to provoke higher levels of participation in protest activities where political costs of mobilization are relatively low. Ancelovici (2015) also focuses on opportunities of anti-austerity mobilization, offering descriptive evidence that political institutions less susceptible to external demands further motivate protest activities among individuals experiencing economic hardship generated by austerity. Dynamics of divergent anti-austerity movements: Structural and contingent opportunities The joint influences of austerity and external factors on anti-austerity mobilization suggest that the growth of anti-austerity protest can be best understood as an interaction between the economic threat of austerity and external factors. This interaction sheds substantial light on the divergence in anti-austerity mobilization, as the conditioning role of external factors helps to account for why severe austerity lead to varying degrees of protest. And yet, there is a fundamental issue overlooked in the debate on divergent anti-austerity activism. That is, prior study tends to treat factors exogenous to austerity as a constant or mere attribute of the existing social system or 12 institutional arrangements. It is important to note, however, that not all external factors are a byproduct of the existing social system. As aptly pointed out by Rootes (1999), external circumstances that constrain or promote protest mobilization need to be distinguished into two parts: one that is structurally drawn from the existing institutional arrangements, and the other that is rather contingent on choices of actors or historical conjunctures (i.e., a set of temporally adjacent events that, combined together, have a critical influence on protest mobilization). Rootes (1999) brings this distinction into the study of protest mobilization by revising the notion of political opportunity developed by earlier literature on social movements (Eisinger 1973; Kitschelt 1986; McAdam 1982; Tarrow 1994). According to Kitschelt’s (1986) influential formulation, political opportunity refers to the degree of openness (or closeness) and strength (or weakness) of state institutions that facilitate (or hinder) protest mobilization: the openness is characterized primarily by the existence of multiple/fragmented parties, factions, and interest groups, the independence of the legislature from the executive branch, and the access of external groups to the executive branch and law-making processes; the strength is represented by the capacity of state authority implementing policies effectively (which often requires a high level of centralization) and autonomously from market forces and the judiciary (cf. Eisinger 1973; Tarrow 1994). In general, openness and strength determine the dominant form and political impact of protest: a relatively open (closed) political system tends to inspire assimilative (disruptive) and potentially more (less) extensive protest activities; a strong (weak) system enables protest activities to have stronger (weaker) and more (less) lasting impact on policy-making. 13 While recognizing the mobilizing role of political opportunity, Rootes (1999) opposes the use of this notion in its original form, contending that Kitschelt conflates two separate dimensions of political opportunity: structural and contingent opportunities. These two dimensions of opportunity differ from each other with regard to their relationship with the existing arrangements of political institutions. That is, the structural opportunity is one immediately drawn from those arrangements and can therefore be characterized as “genuinely structural features” of those arrangements, whereas the contingent opportunity is not strictly determined by those arrangements, but rather “essentially contingent upon the political strategies and tactics” of political actors or historically “conjunctural” (Rootes 1999:4). Accordingly, the structural opportunity often appears to be more or less enduring and durable representing the status quo of political institutions. Its manifestations are clearly suitable for the typological analysis, as exhibited in Kitschelt’s work, because they largely correspond to the established forms of political institutions and cleavages. The contingent opportunity is, however, variable and mutable responsive to short-term changes, because it materializes in choices of protest organizers and/or a sequence of events, both of which tend to be shaped by the particular historical context of protest mobilization rather than by the constraints imposed by the contemporary political systems. As such, the malleable nature of contingent opportunity requires attention to peculiarities of its instances that may not be subsumed under general categories. Considering both dimensions of political opportunity is crucial to understanding the divergence in anti-austerity activism, because it reveals both 14 heterogeneities across protest movements that are rooted in institutional differences and those that are contingent on non-institutional factors. Prior research, however, tends to focus overwhelmingly on the structural factors, drawing on the enduring features of the existing political system (Ancelovici 2015; Beissinger and Sasse 2014; Císar and Navrátil 2016), whereas seldom exploring the contingent opportunity. This in turn risks falsely attributing all the observed difference in protest outcomes to the differences in the structural opportunity. The comparative analysis of anti-austerity movements in South Korea and Greece presented below shows that the contingent opportunity played an important role in the formation of protest movements, hindering and promoting protest mobilization in the two countries. SOUTH KOREA IN 1997-98 AND GREECE IN 2009-10 The present study investigates the origins of divergent anti-austerity movements by examining how the structural and contingent opportunities come into play to interact with the economic threat of austerity and eventually give rise to varying protest movements. Clearly, analyzing the unfolding of each dimension of opportunity in their relation to the threat of austerity requires a careful tracing of a mobilization process with close attention to how both the enduring structural and contingent features of political environments shaped the path of mobilization over time. Particularly suitable for this task is comparing cases of anti-austerity mobilization that share a set of key attributes on the input side of anti-austerity mobilization (i.e., austerity enforcement) and diverge from each other on the output side (i.e., protest activities). Cases of anti- austerity movements that meet these criteria allow the analyst to limit attention to a 15 group of factors that jointly cause the difference in protest outcomes while controlling for the other likely causes.1 Two cases – South Korea (henceforth Korea) in 1997-98 and Greece in 2009- 10 – selected for the present study show fundamental similarities on the input side, whereas differing remarkably from each other on the output side. In the rest of this section, I discuss the similarities and dissimilarities of the two cases in detail and turn then to a method of comparison taken to analyze them. Input side: contextual similarities in austerity enforcement Korea and Greece exhibit three important contextual similarities on the input side of anti-austerity activism. First, austerity was implemented in Korea and Greece under economic contexts that bore structural resemblance to each other, proposed as a remedy for an unprecedentedly severe debt crisis that broke out after years of steady growth. Each in the mid-1990s and the late 2000s, Korea and Greece faced the greatest economic crisis in the history of their democracy. Both Koreans and Greeks had enjoyed prosperous pre-crisis years propelled by the stable growth of the economy 1 This analytical strategy approximates what is often called the “most-similar” comparison (e.g., Collier 1991; Skocpol and Somers 1980), a popular method seeking to differentiate cases matched as closely as possible to the extent that they share many important attributes but differ from each other with regard to only one or two critical aspects. Yet, it is worth noting that the assumption for the similarities between the two cases chosen for the present study is less stringent than what is typically observed in many applications of the most-similar method in that it suffices to ensure contextual similarities in austerity enforcement while leaving room for heterogeneities over the course of anti-austerity mobilization. While less-suited for comparative studies seeking to detect a single critical factor that differentiate the chosen cases, this research design lends itself to tracing processes of mobilization in which multiple factors interact with each other jointly influencing protest mobilization over time. 2 See Kang and Jaung (1999) and Kim (2000) for a review of the political shifts during 16 in the 1980s and early 1990s (Korea) and in the 1990s and the 2000s (Greece) (Emery 2001; Mitsopoulos and Pelagidis 2011). It is also important to note that the welfare state had been in an expansionary phase prior to the crisis in both countries (Kwon and Holliday 2007; Pappas 2014). It was against this background that austerity measures proposed to cope with the debt crisis exposed citizens in each country to economic hardship that was unprecedentedly severe. Next, austerity was implemented in highly analogous political environments in Korea and Greece, brought in during a political transition period and receiving strong support from both major parties under the two-party system. In Korea, austerity formed the major part of the International Monetary Fund (IMF) bailout deal, and negotiations with the IMF on the deal were initiated in late 1997 under the Kim Young-sam government (1993-1997). Yet, after Kim’s conservative party, New Korea, failed to regain power in the 1997 presidential election, facing severe criticism for causing the debt crisis, the deal was taken over by the newly elected Kim Dae-jung government (1998-2003). Apparently, the 1997 victory of Kim Dae-jung and his center-left party, National Congress, emerged out of the fall of their chief competitor, New Korea. And yet, the IMF deal retained its original form under the new government, which, like its predecessor, believed that strong austerity was indispensible for rebalancing the economy.2 The Hellenic austerity reform also came out of a political shift under the two-party system. The rise of the debt crisis in the late 2000s provoked an outburst of grievances among Greek voters, and in turn prompted 2 See Kang and Jaung (1999) and Kim (2000) for a review of the political shifts during the Korean debt crisis and the 1997 presidential election. 17 the fall of the incumbent center-right New Democracy under Kostas Karamanlis (2004-2009). In the early election in October 2009 Karamanlis called to reclaim voter confidence amid the downturn, New Democracy faced a crushing defeat, securing only 91 out of 300 seats, against its longtime socialist rival PASOK led by George Papandreou, which stepped into power with 160 seats (Dinas 2010; Pappas 2014). It was this newly elected Papandreou government that embarked on austerity, which was a key bailout condition proposed by international lenders: the European Commission (EC), the European Central Bank (ECB), and the IMF. Like in Korea, both major parties threw bipartisan support behind the austerity reform, calling for the need for substantial fiscal retrenchment to cope with the mounting debt and fight the recession. Last, austerity came out of highly similar (perceived) international contexts in Korea and Greece, as it was mandated by the international bodies and the involvement of those bodies was widely criticized as foreign intervention. In Korea, the IMF’s economic involvement through the bailout deal was a major source of social unrest. In the first place, the external pressure placed on the Korean government during the bailout negotiations fueled a collective animosity over foreign intervention: evoking the days under Japanese colonial rule (1910-1945) and the U.S. military rule (1945- 1948), Korea’s major newspapers deplored the bailout agreement as ceding sovereignty to foreign powers once again (Shin and Chang 2000). Also, serious doubts were casted on the efficacy of the bailout program. Those doubts evolved out of the existing critique of the IMF’s role in the world economy. Particularly inspired from the Latin American debt crises in the early 1980s, a group of scholars and journalists contended that the IMF’s standardized recovery program promotes further 18 deterioration of debtor countries by exercising excessive control on their government spending and forcibly opening up their feeble financial sector. It was argued in that regard that debtor countries come to not only lose self-control of their economy but also were subjugating themselves to the existing economic order that serves for the benefit of a few developed countries (Bullard, Bello, and Mallhotra 1998; Sachs 1997; Shin and Chang 2000). The pursuit of international aid triggered popular discontent in Greece, too. In particular, the lenders – the EC, ECB and IMF (dubbed “Troika”) – became the major target of criticism, often depicted as invaders. Studies of Greek media under the crisis (e.g., Lialiouti and Bithymitris 2017; Michailidou 2017; Tzogopoulos 2012) brought to light two discursive themes that prevailed in the Greek public sphere. One was a nationalist narrative, which presented the Troika’s economic intervention as “foreign occupation” seriously intruding on Greek sovereignty (Lialiouti and Bithymitris 2017). Proponents of this view contended that by signing the bailout memorandum, the Greek government relinquished their proper role, leaving the lenders to hold sway over the country’s fate. The other was a deep distrust of the lenders. Not only was the existing skepticism of the IMF-led reform, the one that had also been aired amid the Korean crisis, running rampant in the Greek media. But there was also an upsurge of the anti-EU and -German sentiment: the EU and its de facto head, Germany, were portrayed as conspiring to rescue the Eurozone from the crisis at the expense of Greece and a few other peripheries (Michailidou 2017; Tzogopoulos 2012). 19 Output side: contrasting protest outcomes While austerity was enforced under such similar economic and political contexts in Korea and Greece, it faced vastly contrasting protest movements.3 In Korea, three anti- austerity mass rallies were reported in 1998, each on May 5, May 30, and November 8: the rallies in May were attended by around 22,000 and 5,500 citizens, respectively, and the rally in November by 30,000. These figures were obtained from the author’s keyword-based search over the Korea Integrated Newspaper Database System (KINDS), a full-text newspapers archive run by Korea Press Foundation that covers 10 nationwide newspapers and 25 local dallies.4 The scopes of these mobilizations are certainly worth attention. Yet, they may not impress students of Korean social movements, as protest activities with more than 30,000 participants were not uncommon in pre-crisis Korea (Shin, Chang, and Kim 2007). Nor was Korea’s anti- austerity activism particularly long lasting – as will be discussed below, none of the three rallies fueled further waves of local protests during 1998. A much greater number of anti-austerity protests sprang up in Greece: once it embarked on fiscal tightening in February 2010, the Greek government had to face 3 To examine the number and scale of anti-austerity movements in each country, the present study drew on protest events reported by newspapers. This method can be regarded as an application of the Protest Event Analysis (PEA) (Koopmans and Rucht 2002). It is important to note that while the PEA is generally reliable in counting the number and size of events, it may cause potentially serious bias when applied more broadly: most notably, media reports of protest events often misrepresent the original protest agenda (Smith et al. 2001). 4 Unlike its Greek counterpart, the Korean anti-austerity protest has not been systemically studied in its full range. To examine the occurrences of anti-protest events in Korea during 1997-1998, I conducted a keyword search on KINDS (kinds.or.kr), inputting “IMF” and words indicating protest activities, “demonstration”, “rally”, “march”, “riot” and “strike” (January 1, 1997 through December 31, 1998). 20 hundreds of marches every month, which is particularly noteworthy given that even in the early 1990s, the supposedly most contentious era in post-dictatorship Greece, there had been just over 200 demonstrations per year (Rüdig and Karyotis 2014). Among those protest events, three were markedly large, mobilizing 200,000 or more individuals: around 200,000 citizens participated in the protest on March 11, 250,000 on May 5, and 200,000 on December 15. The May 5 rally was reportedly the largest mobilization since the fall of the military junta in 1974 (Rüdig and Karyotis 2014). There were also protests that were smaller than these events, but were still remarkable: 50,000 individuals joined the protest on February 24, 60,000 on November 17, and 30,000 on December 6. These figures were drawn from Diani and Kuosis (2014), who examined 440 newspaper/online sources to construct a dataset for Greek anti-austerity protest during 2010-2012. While it appears evident that the anti-austerity protests in Greece were of much bigger size than those in Korea, the fact that the Greek population in 2010 was less than a quarter of the Korean population in 1998 (OECD 1998, 2011) makes this even more remarkable. It is also noteworthy that Korea’s anti-austerity rallies were all centered on its capital city, Seoul, whereas Greece’s spanned far beyond Athens, reaching not only other major cities but also numerous small towns (Psimitis 2011; Rüdig and Karyotis 2014, 2013). Method of paired comparison Why then did the austerity reforms implemented in the highly similar economic and political contexts lead to the contrasting protest movements? A methodological 21 strategy taken in this study to answer this question is a paired comparison of mobilization processes in Korea and Greece. The focus of comparison is on the divergent trajectories of protest movements from their formation to expansion. Specifically, the analysis begins with an examination of the incipient movements, organized by the principal victims of austerity in both Korea and Greece, and then traces how differently these movements evolved in the two countries. In the social sciences, this method of comparison was devised to draw a key factor(s) that links a particular social event(s) up with similar or contrasting outcomes across the chosen pair of cases. This factor is found through a careful examination of processes in which the event commonly observed in both cases results in the similar or contrasting outcomes: as this method cross-examines a pair of cases, Tarrow (2010) refers to it as “dual-process tracing”5. Here this strategy traces in parallel the processes of protest mobilization across the Korean and Greek cases. The crucial analytical advantage of this method is that it allows the analyst to keep track of the evolution of protest movements at both institutional and individual levels. This in turn helps to uncover the differing roles of existing institutional arrangements and strategic behavior of protesters at critical phases of protest mobilization in Korea and Greece. In terms of political opportunity discussed earlier, the results of analysis will therefore reveal how the structural and contingent opportunities shaped the divergent paths of protest movements in Korea and Greece.6 5 For a general overview of the method of process-tracing, see Mahoney (2012). 6 Since anti-austerity protest broke out in Korea and Greece at the different time points, one might wonder if the cross-time comparison of the two protest movements would omit a critical factor that existed in only one of the two cases and gave rise to the 22 Before embarking on the comparative analysis of Korea and Greece, I shall undertake an analysis of economic threats posed by the austerity reforms in the two countries. Comparing the threats of austerity is an important preliminary step for the present study, because it permits us to test the fundamental intuition drawn from the extant literature: the divergent protest movements stemmed from vastly different austerity measures. Did different economic threats trigger the different protest movements? Austerity in Korea and Greece was implemented to combat crippling debts in the private sector (Korea) and public sector (Greece). Official debt statistics, summarized in Table 1, help examine the financial malaises in each case. contrasting protest outcomes. Most notably, many might cite the extensive use of information technologies as one that helped the Greek protesters to propagate their ideas and organize rallies (whereas the Korean protesters in 1998 were not equipped with such tools). Yet, there are two caveats that lead us to be cautious about taking this claim. Above all, the mobilizing effect of information technologies has yet to be validated, as existing studies report conflicting findings (see Little 2015). Next, it remains uncertain how much effect information technologies actually had on the anti- austerity mobilization in Greece. Even more uncertain is whether the Korean anti- austerity protest in 1998 would have received much broader support had there existed ubiquitous online communication tools: while social media might have boosted the anti-austerity movement in Korea by facilitating the circulation of anti-austerity ideas, one can also make the opposite argument that the fate of Korea’s anti-austerity movement would not have changed much, or would have been confronted with even greater constraints, because social media would have further promoted the pro- austerity sentiment that prevailed in the Korean public sphere in 1998. 23 Korea Greece 1994-96 1997 1998 2006-08 2009 2010 (averaged) (averaged) Government financea (% of GDP) - Debt 9.0 10.2 14.7 105.3 126.7 146.2 - Balance 2.9 2.7 0.6 -7.6 -15.1 -11.2 Private financeb (% of GDP) - Debt 203.3 236.5 271.1 114.0 129.1 139.8 Foreign, short-term 9.6 10.2 9.1 0 0 0 loans and securities Table 1. Major macroeconomic indicators: Korea in 1994-98 and Greece in 2006-10 Sources: a: IMF (2014); OECD (2014b) b: BOK (2015); World Bank (2014) In Greece, the sovereign debt soared up to almost 150 percent of Gross Domestic Product (GDP) in 2010, which was far more than twice the Maastricht treaty’s threshold, 60 percent of GDP (European Communities 1992). In turn, when the global recession hit both Greece and its creditors in the late 2000s, the solvency of the Greek public sector became jeopardized in the face of the collective financial meltdown (Mitsopoulos and Pelagidis 2011). In Korea, the malaise came from the private sector. Korea’s private debt-to-GDP ratio rose to 271 percent in 1998. While the gross private debt certainly undermined the competitiveness of Korea’s private sector, a lethal threat emerged out of a particular subset of it: short-term foreign loans and debt securities. The short-term liabilities Korea’s private sector owed to foreign 24 lenders amounted to almost 10 percent of the nominal GDP during the mid-1990s, and the debtors, mostly enterprises, were not capable of repaying them once the lenders, on the alert for prolonged recession both within the country and across the region, refused to roll them over.7 The different roots of the debt crises determined the forms of austerity reforms in Korea and Greece. On the Korean side, the primary goal of austerity was restructuring a troubled corporate sector, to allow companies to clean up their liabilities and slim down their workforce until they could regain growth momentum. To achieve this goal without delay, the government not only legalized layoffs, temporary employment, and pay reduction for restructuring purposes but also strictly controlled the money supply to pre-empt further expansion of reckless management and fend off inflationary pressures. The latter effort, tight money, took the most intense form during the early phase of restructuring – since it rose from 14.3 percent to 20.4 percent on December 3, 1997, the call loan rate, the Bank of Korea (BOK)’s base interest rate at the time, had not fallen below 20 percent until May 1998 (BOK 2015). This cut off business loans, and in turn drove firms into accelerating their restructuring effort (Lee 2011). On the Greek side, fiscal consolidation was placed at the top of the agenda, a crucial step to convince creditors that Greece would retake control over its debt and put public finances on a sound footing. Under the imminent threat of sovereign default, the Greek government aimed to reduce the deficit, which reached 15.2 percent 7 Lee (2011) and Mitsopoulos and Pelagidis (2011) each provide a comprehensive overview of the Korean and Greek crises. 25 of GDP in 2009, to below 10 percent of GDP by the end of 2010. A set of stringent austerity measures followed, aiming to yield 9.1 percent of GDP in 2010 through cuts in expenditure (3.6 percent of GDP) and increases in tax revenue (5.5 percent of GDP). Such drastic spending cuts and tax hikes each brought out a sharp reduction in Greece’s public sector, i.e., the number of employees and the amounts of compensation and benefits, and dramatic increases in income/property and consumption taxes.8 How severe then were the overall economic threats of austerity in Korea and Greece? The larger protest movement witnessed in Greece leads us to test a hypothesis that the Greek austerity reform posed a more serious threat than did the Korean one. In testing this hypothesis, the theory of policy threat provides a good way to measure the threats of austerity, evaluating the resulting economic losses. To measure the amounts of economic losses imposed under the austerity reforms in Korea and Greece, I examine the trends in two labor market indicators – employment and earnings – and household disposable (that is, post-tax-and-transfer) income: this measurement strategy encompasses those provided by recent studies of anti-austerity activism (Císar and Navrátil 2016; Grasso and Giugni 2016; Ponticelli and Voth 2011). The changes in employment and earnings demonstrate the immediate economic losses working people experienced under the reform. Notice that the major victims of the austerity reform were workers in both Korea and Greece: Korea’s corporate restructuring led to massive job losses and wage cuts in the private sector; 8 For the full review of each country’s austerity program, see IMF (1997) and EC (2010). 26 Greece’s fiscal adjustments caused the steep reduction in compensation among public sector employees. Therefore the two labor market indicators are expected to reveal the direct impact of each austerity program on the livelihoods of the public in the two countries. The changes in household disposable income exhibit the overall effect of each austerity program, extending the scope to all households and also controlling for the two countries’ different social welfare systems. Exactly how much loss was generated under each reform? Were Greeks actually exposed to greater losses than were Koreans? 62# 61# 60# 59# 58# 57# 56# S.#Korea# 55# Greece# 54# 53# 5)#0 06 )# 7)# )#0 08 09 )# 0)#1 4-2 0 4-2 0 -20 -20 -20 -20 Q Q Q4 4 4 4 3#( 4#( 5#( 6#( Q Q Q 9 9 9 9 97 #( 98 #( -194 4-1 9 4-1 9 -194 4-1 9 -19 Q Q Q Q Q Q4 Figure 1. Employment rates Korea in 1994-98 and Greece in 2006-10 Sources: OECD (2014a). 27 10$ 8$ 6$ 4$ 2$ 0$ S.$Korea$ !2$ 1993$ 1994$ 1995$ 1996$ 1997$ 1998$ (2005)$ (2006)$ (2007)$ (2008)$ (2009)$ (2010)$ Greece$ !4$ !6$ !8$ !10$ !12$ Figure 2. Annual changes (%, real) in average earnings Korea in 1994-98 and Greece in 2006-10 Sources: BOG (2011, 2012); OECD (2014c); Statistics Korea (2009) Note. BOG (2011, 2012) and Statistics Korea (2009) each provide the data for earnings in Greece and Korea in current currencies. The author calculated the annual changes in earnings using Consumer Price Index drawn from OECD (2014). Contrary to the hypothesis, the changes in the labor market indicators exhibit that Koreans experienced more severe threats. While 60 percent of the working age population was in the workforce in both Korea in 1997 and Greece in 2009, the figure dropped to 56 percent in Korea in 1998 and to 57.9 percent in Greece in 2010 (see Figure 1). It was also Koreans who faced a sharper decline in earnings: in Korea, the per capita earnings increased by 2.5 percent in 1997 and dropped 10 percent in 1998; in Greece, they increased by 3.3 percent in 2009 and dropped 9.1 percent in 2010 (see Figure 2). 28 10# 8# 6# 4# 2# S.#Korea# 0# Greece# 1993# 1994# 1995# 1996# 1997# 1998# !2# (2005)# (2006)# (2007)# (2008)# (2009)# (2010)# !4# !6# !8# Figure 3. Annual changes (%, real) in gross disposable income Korea in 1994-98 and Greece in 2006-10 Sources: OECD (2014c). 15$ 10$ 5$ 0$ 1993$ 1994$ 1995$ 1996$ 1997$ 1998$ S.$Korea$ !5$ (2005)$ (2006)$ (2007)$ (2008)$ (2009)$ (2010)$ Greece$ !10$ !15$ !20$ !25$ Figure 4. Annual changes (%, real) in median household disposable income Korea in 1994-98 and Greece in 2006-10 Sources: Statistics Korea (2009); OECD (2014c) Note. The median income data for Greece and Korea were each extracted from OECD (2014c) and Statistic Korea (2009). The author calculated the annual changes in the data in the same way with those in Figure 2. Nor do available data for household disposable income, which allow us to examine the changes in gross and median disposable income, bolster the claim for the stronger threat in Greece. Both total and median income data show that Koreans in 1998 faced a steeper decline than did Greeks in 2010: the gross disposable income dropped 6.7 percent in Korea and 5.7 percent in Greece; the median disposable income 29 dropped 19 percent in Korea and 14.7 percent in Greece (see Figure 3 and 4). It should be noted that Greeks in 2010, due to the longer time gap between the outbreak of the crisis and the bailout agreement, were in more cumulative pain than Koreans in 1998, particularly with the gross disposable income declining by 0.9 percent in 2008 and 3.6 percent in 2009 (see Figure 3). So this might provide evidence for greater losses among Greeks. Yet, it still seems insufficient to outweigh the greater instantaneous losses among Koreans observed in the changes in the labor market indicators and median disposable income, which in turn makes it hard to conclude that the Greeks faced greater threats under austerity.9 THE ORIGINS OF THE DIVERGENT ANTI-AUSTERITY MOVEMENTS Now I begin a full comparative analysis of the two anti-austerity movements to uncover why austerity resulted in the divergent protest movements in Korea and 9 A fundamental objection to case selection in the present study might stem from the fact that the Greek crisis lasted longer than the Korean crisis – that is, some may use this as evidence that the former crisis was harsher than the latter one, on one hand, and contend that the prolonged recession in Greece resulted in larger protest movements, on the other. The length can certainly be used to measure the severity of an economic crisis. Critical to anti-austerity mobilization in Korea and Greece, however, were the periods when the austerity reforms began to be enforced in the two countries, as central actors arose against the reforms and sought to broaden the base of protest movements. And the present study shows that Koreans suffered greater economic losses than Greeks during these periods. If harsher economic burdens trigger stronger protest, therefore, Koreans in 1998 must have protested more than Greeks in 2010 – yet, this was not what actually occurred. More importantly, as discussed in this paper, the trajectories of anti-austerity movements in Korea and Greece show that the fates of movements were determined in the first two years of austerity. True, the Greek movements peaked in 2011, but, as prior research (e.g., Diani and Kousis 2014; Kanellopoulos et al. 2017) shows, it was a continuation of the earlier mobilizations in which the major groups of protestors had formed organizational blocs for further coordination of protest movements. 30 Greece. The analysis closely traces how the structural and contingent opportunities unfolded over the course of protest mobilization in Korea and Greece. It turns out that while protest was provoked by the economic threat of austerity in both countries, different configurations of structural and contingent opportunities in the two countries gave rise to the growth (Greece) and fall (Korea) of protest. Korea in 1998 In Korea, all of the three mass rallies briefly sketched out earlier were spearheaded by the Korean Confederation of Trade Unions (KCTU), the radical wing of the country’s two largest union confederations. Initially, along with the moderate wing, the Federation of Korean Trade Unions (FKTU), the KCTU joined and collaborated with the Tripartite Commission of Labor, Management and Government, a presidential advisory council the President-elect Kim Dae-jung set up in January 1998 to reconcile disputes between employers and workers on the IMF bailout conditions. Yet, the KCTU ended up seceding from this body, declaring that it would not accept the legalization of layoffs, which, it argued, left all the burdens with workers (Grey 2008; Shin and Chang 2000). The two protests in May 1998 were organized to press the government to stop permitting layoffs. The first rally was held on May 1 at Jongmyo Park in central Seoul. An estimated 22,000 protestors, consisting of workers, students and the unemployed, joined the rally (Park, Lee, and Kim 1998). The KCTU staged another rally at the end of the month, encouraging like-minded civic groups to participate: around 2,500 protestors took to the streets near Seoul Station on May 30 under the banner of “employment protection, corporate reform, and renegotiation of 31 the IMF memorandum” (W. Lee 1998). The final rally was held after six months. On November 8, the KCTU and its allies gathered at a waterfront of the Han River, accusing government officials and business leaders of having brought the greatest economic disaster on the country. Approximately 30,000 protestors were reported to attend (H. Lee 1998). Meanwhile, the KCTU’s closest ally, the Hyundai Motors Workers Union (HMWU), was carrying on the struggle at the enterprise level. Composed of manual workers at Hyundai Motors, one of the two largest automotive manufacturers in Korea, the HMWU was not only a union with the largest membership and the greatest bargaining power, but it was also the backbone of Korea’s organized labor as a leader in building a unified front during the country’s democratization in 1987 and providing organizational resources to labor movements afterward (Koo 2001). During the implementation of the IMF restructuring program, too, it stepped up to the plate in the struggle for job security. Against Hyundai management’s plan to dismiss 4,830 workers, revealed on June 29, 1998, the leaders of the HMWU called an all-out strike, declaring that the union “would not accept a single lay-off”(Chang 2002:27). This decision received unanimous support from the union members, which led to a halt in production for two months (Neary 2000). Clearly, the revolts of the KCTU and HMWU workers lend strong support to the mobilizing role of austerity, confirming the causal linkage between the economic threat of austerity and protest mobilization among its victims. Despite their strenuous efforts, however, neither of the two militant unions could find an echo in a broader mass movement: the protests organized by the KCTU failed to expand beyond the 32 initial group of adherents; the HMWU also found itself isolated from other unions and in the end conceded to management, accepting thousands of layoffs and forced retirements. Structural opportunity in Korea Why did those protests not spread? In terms of structural opportunity, Korea’s political system can be characterized as closed and strong, much like France’s in Kitschelt’s formulation (1986): closed because the unitary executive branch, the rigid two-party system, and the limited political and social cleavages fundamentally restricted access to agenda-setting and policy-making from political minority factions and external civic groups; strong as the highly centralized government and its control over market forces allow for effective policy implementation without having to deal with critical veto players (Yun 1997). This type of political system left little room, as predicted by Kitschelt (1986), for the pursuit of assimilative strategies among Korea’s militant union workers, who eventually opted for taking to the streets. Once the anti-austerity rallies broke out in Korea, what became a serious constraint on the spread of these rallies was the closedness of the structural opportunity. Particularly consequential for Korea’s short-lived protest movement was the existing political and social cleavages, which made it elusive for the protesters to find influential allies. In the legislature, while both major parties – New Korea and National Congress – were the leading forces of the bailout agreement, the rightwing minority United Liberal Democrats was in coalition with the ruling National Congress (Kang and Jaung 1999). As a result, the protesters in Korea could not find political 33 forces that could “aggregate [their] demands” (Kitschelt 1986:63) and promote their cause. The comparison with the Greek case, discussed below, will make it clear that this seriously hindered the progress of anti-austerity protest in Korea. Further, the Korean protesters also suffered from the lack of allies outside institutional politics. Above all, they found an acute divide rather than solidarity within organized labor. At the upper level, as hinted above, there was a split between the two confederations: while the KCTU resigned from the Tripartite Commission, demanding a full renegotiation on the bailout conditionality, the FKTU stayed and devoted all its efforts to expanding workers’ compensation schemes under the proposed bailout plan (Grey 2008).10 The division was also prevalent among individual unions. On one side, the HMWU was a vanguard for militant unionism, staking out its hardline stance against the bailout reform. On the other side, there were unions, which, alongside the FKTU, chose to accommodate themselves to reform. The latter type of unionism was manifested in a series of campaigns organized to help insolvent firms. Among those firms was KIA Motors, a longtime competitor of Hyundai Motors. Whereas Hyundai Motors workers came into fierce antagonism with management, those at its rival company did the exact opposite: as KIA Motors was driven into bankruptcy during the financial crisis, its workers made common cause with management to save the company, collecting voluntary contributions, returning bonuses and allowances, and accepting temporary layoffs (Chang 2002). 10 On May 1, when the KCTU organized the first anti-austerity rally, the FKTU also held a rally, which was attended by around 4,000 workers, in Seoul. Yet, unlike the KCTU, the FKTU focused on expansion of social welfare, especially benefits for discharged workers, rather than prohibition of dismissal or renegotiation with the IMF (Park, Lee, and Kim 1998). 34 The divide within organized labor was by no means of a temporary nature, but deep-seated in the history of Korea’s union activism. Established in 1960 as a sole representative body approved and funded by the government, the FKTU had long maintained amicable relations with the authoritarian leaders in pre-democratic Korea, helping them to keep control of union activism (Lee 2002, 2007). Consequently, the rise of independent unions after the 1987 democratization accelerated in direct opposition to the FKTU. The KCTU arose as a new collective body representing those independent unions. After the foundation of the KCTU in 1990, the conflictual relationship between the FKTU and the KCTU became the order of the day, as the KCTU spoke out against a series of bills for labor market deregulation in the early 1990s backed by the FKTU (Lee 2002, 2007). The confrontation over the bailout conditions, which embraced the chief elements of the previous deregulation bills, was a continuation of the existing conflict between the two organizations. Korea’s anti-austerity unionism also had to face strong opposition from without. Indeed, what predominated over Korea’s public sphere during the debt crisis was a pro-austerity movement. While the government was negotiating with the IMF on the bailout conditionality in November 1997, various campaigns promoting austere lifestyles were being carried out by civic groups: “Such civic group members [could] be seen in parks and on streets shouting, appealing to the people to be thrifty and save hard currency” (Kim and Finch 2002:125). These pro-austerity campaigns urged citizens to exert strict control over consumption and lead a frugal life to meet the economic hardship. Shortly afterward, as the bailout program was officially announced on December 3, the campaigns gained a lot greater momentum, 35 culminating in the so-called “Gold Collection Drive”: to save the economy suffering from acute foreign currency shortages, Korean citizens started to donate or sell gold, one of the surest means to add to foreign reserves, to government agencies. While this initially took the form of local campaigns, organized by voluntary organizations, the Gold Collection Drive rapidly spread out over the nation: by March 14, 1998, approximately 3.5 million citizens, representing 23 percent of the total households, participated, bringing their beloved treasures, which in total amounted to 226 tons of gold, valued at 2.2 billion dollars (Kim and Finch 2002; Lee 2011). Contingent opportunity in Korea While the division between the social groups surrounding the protesters impeded anti- austerity mobilization in Korea, it is important to note that the protesters were also confronted with a serious internal constraint. This obstacle had nothing to do with the existing social or political cleavages. Rather, it was a consequence of strategic choices made by the principal protest organizers: the KCTU. As discussed earlier, the KCTU initially joined with the FKTU in the collective effort by the government, unions, and business leaders to draw up an agreement on the scope and duration of austerity. This decision resulted in the KCTU’s joining the Tripartite Commission in January. The KCTU, however, withdrew from the Tripartite Commission in May, declaring its opposition to massive layoffs and organizing a mass rally. And yet, this stance did not last a month, as the KCTU rejoined the Tripartite Commission in June and resumed negotiations with the government and business leaders; it then quit again and staged another anti-austerity rally in November. 36 This vacillation stemmed from an internal conflict the economic crisis triggered among the members of the KCTU: the organization had been divided over how to respond to the bailout deal, and its leaders bumbled along from one position to the other (Neary 2000). Not only did this irresoluteness render the KCTU’s own resistance less sustainable, but it also had a devastating effect on its ally’s struggle. Especially, while the HMWU was carrying on its battle against the company’s layoff plan, the KCTU’s decision in June to rejoin the Tripartite Commission “certainly isolated the HMWU and left the impression that the nation-wide resistance to the lay- offs, declared by the KCTU, was more rhetorical than real” (Neary 2000:4). Indeed, with the retreat of its sole ally, the HMWU became an easy target for oppression: when the HMWU’s struggle reached its peak with the shutdown of the workplace on August 14, the government concentrated police forces in the factory district, deploying over 10,000 riot police, to disperse the strikers. Just a few days after this intervention, the HMWU called off the strike on August 24, withdrawing its no-layoffs slogan (Neary 2000; Chang 2002). Greece in 2010 The Greek anti-austerity protest movement was initiated by the country’s two largest union confederations, the Civil Servants’ Confederation (ADEDY) and the General Confederation of Greek Workers (GSEE). Each representing public and private sector employees, the two organizations had been acting in concert from the earlier phase of the bailout negotiation. Starting with the first joint general strike on February 24, 2010, the collaboration between the two organizations was seen to be strengthened further 37 after the announcement of the bailout plan on March 3, as the ADEDY and GSEE went out on two more general strikes on March 11 and May 5 (Diani and Kousis 2014; Psimitis 2011; Simiti 2015). The latter two strikes occurred in conjunction with other anti-austerity marches and rallies that swept over the country during March and April, which will be discussed below (Psimitis 2011). This accounts in part for a rapid increase in the number of participants in the joint general strikes: while 30,000-50,000 participants – union members and other demonstrators combined – were reported in the first joint general strike in February, 200,000 in the second in March, and 250,000 in the third in May (Diani and Kousis 2014). The mobilization of workers turned out to be quite successful during the whole year of 2010. Alongside the ADEDY and GSEE’s second general strike on March 11, the All Workers Militant Front (PAME), the communist affiliate of the GSEE with nationwide membership, organized independent rallies in multiple cities (Psimitis 2011). Strikes also sprang up at the industry and firm levels. In March, workers at the National Power Company (DEH), one of the largest public enterprises in Greece, joined the wave of anti-austerity protests, declaring a 48-hour strike on the 16th and 17th (Psimitis 2011). In April, both industry-level and enterprise-level unions as well as workers’ voluntary associations, encompassing schoolteachers, transportation workers, doctors, and vendors, joined the wave of protests, organizing strikes, work stoppages or rallies (Psimitis 2011). And, finally in May, the workers’ uprising culminated in the general strike on the 5th, which took place alongside a large mass demonstration in Syntagma Square in central Athens: in most branches of the public and private sectors, over 90 percent of the employees were reported to join the strike 38 and demonstration (Psimitis 2011). Strikes continued after May, as the Papandreou government adhered to the original bailout conditions. In the end, a total of 7 general strikes were called in 2010, which surpassed the total number of general strikes, 3, organized during 2007-2009 (Rüdig and Karyotis 2014). The Greek uprising reached beyond organized labor. According to Rüdig and Karyotis’ (2013) survey, which was conducted in December 2010 using stratified quota sampling and received 1,014 valid responses, trade union members made up 27 percent of all respondents who had participated in anti-austerity rallies at least once. Especially after the announcement of the bailout plan in March, Greece’s anti- austerity protest spread across a wide variety of social groups. Among those groups were political minorities that were staying the course with the protestors from the early phase of the anti-austerity movement. Three leftwing minority parties – the Coalition of Radical Left (SYRIZA), and the Front of the Greek Anticapitalist Left (Antarsya), and the Greek Communist Party (KKE) – rebelled11, claiming that the bailout agreement should be annulled (Psimitis 2011; Simiti 2015). These three parties formed the largest bloc along with the two union confederations at the May protest in Syntagma Square (Simiti 2015). There was also an outburst of grievances from anarchists, the largest grassroots activist groups in Greece, which took the form of spontaneous revolts in urban areas (Kritidis 2014; Psimitis 2011). Besides all these 11 Unlike those on the left, political minorities on the right did not take part in the protest movements. The parliamentary right, the Popular Orthodox Rally (LAOS), opted for giving support to the bailout reform. The far-right extra-parliamentary party Golden Dawn opposed the bailout deal, carrying out electoral campaigns against it, and voluntary participations among its members in anti-austerity rallies were reported. Yet, Golden Dawn did not officially join those rallies (Kanellopoulos et al. 2017). 39 groups, a large number of citizens with no previous experience in collective action joined the anti-austerity protests: Rüdig and Karyotis’ (2013) survey reports that these “novice” protestors made up 20 percent of the respondents who engaged in anti- austerity protest in 2010. Taken in summary, while Greece’s anti-austerity movement was initiated, like Korea’s, by the principal victims of austerity, the workers, it turned into a broad mass movement, winning strong support from both organized labor and various political allies, which was in stark contrast with the Korean movement. What then account for the stronger protest movement in Greece? Structural opportunity in Greece The structural political opportunity surrounding the Greek uprising was distinct from its Korean counterpart in both aspects spelled out above. First, Greece’s political system was more prone to extensive and durable mobilization. Clearly, the Greek system can be characterized, much like the Korean system, as strong, because the highly centralized administrative power ensures effective policy enforcement on behalf of the ruling majority. It was more open, however, than the Korean system: although Greece’s two-party system that had long persisted in the pre-crisis years, like in Korea, allowed the ruling majority to exercise exclusive control over the legislature and executive branch, there existed, unlike in Korea, minority parties with past experiences in activism in the Greek parliament that were willing to join the protest movement and turn its demands into policy agenda (Kriesi 2004). As such, the extensive political cleavage in the Greek institutional politics facilitated the growth of 40 anti-austerity protest. Particularly remarkable was the uprising among the radical left parties: SYRIZA, ANTARSYA, and KKE. The mobilization from the radical left parties broke into two groups: on one side, the parliamentary socialists, SYRIZA with 13 seats in 2010, and the young extra-parliamentary communists, ANTARSYA founded in 2009, formed a joint protest group; on the other side, the parliamentary communists, KKE with 21 seats in 2010, organized separate rallies. (Simiti 2015; Tsakatika 2016) The former two, both established as a coalition of diverse factions, not only mobilized their party members – the party membership of SYRIZA was estimated at over 16,000 – but also worked as an umbrella organization for a broader group of individuals opposing the bailout agreement (Simiti 2015; Tsakatika 2016). KKE’s split was reported to occur due to its long-maintained sectarian activism equipped with abundant organizational resources: over 40,000 party members and local units under strict control of the central committee. (Tsakatika 2016; Tsakatika and Eleftheriou 2013) The simultaneous mobilization from the radical lefts formed a massive wave of protests, exerting collective pressure against the incumbent PASOK (Pappas 2014). Next, where unions in Korea were divided over how to cope with the bailout conditions, being split into the pro- and anti-austerity groups, those in Greece showed a strong solidarity from the onset. As described above, unlike the KCTU and FKTU, Greece’s two largest labor confederations – the ADEDY and GSEE – made a joint effort to mobilize against the austerity reform, winding up drawing participation from the majority of the union members. This strong unity within organized labor was not driven by a contingent convergence of interests that would have otherwise conflicted 41 with each other. Rather, the various fractions of organized labor in Greece had built a united front in pushing collective demands for compensation and social benefits during pre-austerity years (Ioannou 2000), which facilitated coordination and expansion of anti-austerity protest movements among a growing number of unions during early months.12 Contingent opportunity in Greece In addition to the structural opportunity characterized by the extensive political cleavages and strong solidarity within organized labor, the Greek anti-austerity movement benefited from a contingent dimension of opportunity that unfolded at a particular historical juncture conducive to protest mobilization. In order to unveil how this dimension of opportunity came into play, it is crucial to understand that the Greek anti-austerity protest occurred in times of growing distrust of political leaders and institutions and therefore need to be situated in the tradition of anti-establishment movements (Andronikidou and Kovras 2012; Dalakoglou and Giovanopoulos 2011; Porta 2015). An especially important catalyst for the anti-austerity protest in 2010 was brought about by the 2008 protest. In December 2008, a large number of Greek citizens assembled in Syntagma Square and staged rallies against the government. These protests were ignited by a death of a 15 year old student, Alexandros 12 In Greece, while the GSEE represents the private sector, its large share consists of public sector employees, so that the GSEE, much like the ADEDY, is controlled by public sector workers (Kretsos 2011). This particular configuration of union confederations in Greece certainly contrasts with the longstanding split within Korea’s organized labor. 42 Grigoropoulos, by two policemen in Athens. According to reporters and witnesses, on 6 December 2008, one of the two policemen shot his gun – two warning shots and a third to the ground – to threaten a group of boys after exchanging verbal insults with them; one of the boys was shot by the third bullet by accident. Although the exact sequence of events was not clearly addressed during the criminal investigation (while the police authority announced that there was a provocation from the boys, eyewitnesses argued that the boys did not provoke the policemen at all), a number of irate citizens, composed of thousands of workers and university students, began to express their accumulated distrust of the state authority by taking to the streets. The protests lasted for three weeks. Initially spurred on by the tragic accident involving the young boy, the protest activities evolved into mass revolts, which expressed outrages and grievances over perennial issues, such as corruption and the lack of government accountability, that had motivated much of the Greek public discourses in the post- junta era – see Karamichas (2009) for an overview of the 2008 protest. With the presence of the long-standing discourses of anti-establishment and civic disobedience, the 2008 protest movement played an important role in facilitating the protest mobilization in 2010. Recent research (e.g., Georgiadou et al. 2019; Karakatsanis 2016; Roos and Oikonomakis 2014) reports that the rapid expansion of the 2010 anti-austerity movement was driven in part by the 2008 protest, stressing that the protesters in 2010 could capitalize on anti-regime grievances that had inspired the 2008 protest. Clearly, the temporal adjacency of the austerity reform and the 2008 43 protest, given the predominance of anti-regime sentiment, constituted a historical conjuncture conducive to protest mobilization in 2010.13 CONCLUDING REMARKS This study examined the causes of divergent anti-austerity movements through a comparative analysis of Korea in 1997-98 and Greece in 2009-10. The focus was primarily on the roles of external factors in moderating the mobilizing effect of austerity. The results advance our understanding of these roles, breaking external factors into the structural and contingent dimensions of political opportunity and accounting for how each dimension of opportunity shaped the progress of protest provoked by the economic threat of austerity. In terms of structural opportunity, the differences in the existing political and social cleavages facilitated the varying degrees of anti-austerity mobilization in the two countries: the limited cleavage among political elites and the high level of fragmentation within organized labor seriously restricted the scope of protest mobilization in Korea, whereas the existence of political factions that capitalized on public discontent with austerity and the unified labor movement fueled the growth of protest movements in Greece. In terms of contingent opportunity, the strategic inconsistencies among the KCTU leaders wound up dismantling the organizational base of Korea’s anti-austerity activism. On the contrary, the rapid spread of the Greek protest movement in 2010 was fostered in part by the 13 The role of historical conjuncture also implies that protest movements are seldom spontaneous, but they are often manifestations of pre-exiting movement cultures (Accornero and Ramos Pinto 2015; Flesher Fominaya 2015; Morell 2012). 44 temporal contingencies that crystalized in grievances and anger that had been predominant during the 2008 protest. While this result demonstrated the processes in which the structural and contingent opportunities jointly promoted (Greece) and restricted (Korea) anti- austerity protest mobilization, future research may find various ways in which the two dimensions of opportunity interact over the course of mobilization. Clearly, the two dimensions of opportunity can be decoupled from each other, as they are largely autonomous of each other. That said, comparative studies of cases where structural and contingent opportunities have conflicting effects on anti-austerity mobilization would help to flesh out plural mechanisms that give rise to divergent anti-austerity movements, accounting for how the two dimensions of opportunity could not only reinforce but also counteract each other. The concrete understanding of the structural and contingent opportunities also helps to refine the notion of political opportunity and bring it back into the study of social movements. Since its formulation in earlier literature (Eisinger 1973; Kitschelt 1986; McAdam 1982; Tarrow 1994), political opportunity has inspired strong criticism pointing to the tendency among its proponents to overstretch its conceptual boundary. At the heart of this criticism was the concern that political opportunity is often addressed as an umbrella concept that embraces not just the configuration of political systems but also changes in broad socio-economic circumstances: as a result, the concept of opportunity, argue critics, was “in danger of becoming a sponge that soaks up every aspect of the social movement environment” (Gamson and Meyer 1996:275). The all-encompassing nature of opportunity reduced its analytic use value, 45 as it risks being tautological (Goodwin, Jasper, and Khattra 1999; Meyer and Minkoff 2004) in that it ends up describing, rather than explaining, protest mobilization by listing all social changes that had occurred prior to its emergence. Decomposing opportunity into its structural and contingent dimensions and examining instances of each dimension allows us to make better use of it by accounting for how the variation in each dimension facilitates or hinders the evolution of movements. 46 REFERECNES Accornero, Guya, and Pedro Ramos Pinto. 2015. “‘Mild Mannered’? Protest and Mobilisation in Portugal under Austerity, 2010–2013.” West European Politics 38(3):491–515. Ancelovici, Marcos. 2015. “Crisis and Contention in Europe: A Political Process Account of Anti-Austerity Protests.” Pp. 189–209 in Europe’s Prolonged Crisis, edited by H.-J. Trenz, C. Ruzza, and V. Guiraudon. London: Palgrave Macmillan. Andronikidou, Aikaterini, and Iosif Kovras. 2012. “Cultures of Rioting and Anti- Systemic Politics in Southern Europe.” West European Politics 35(4):707–25. Beissinger, Mark R., and Gwendolyn Sasse. 2014. “An End to ‘Patience’? The Great Recession and Economic Protest in Eastern Europe.” Pp. 334–70 in Mass politics in tough times: opinions, votes and protest in the Great Recession, edited by N. G. Bermeo and L. M. Bartels. Oxford: Oxford University Press. Bermeo, Nancy Gina, and Larry M. Bartels, eds. 2014. Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. Oxford: Oxford University Press. Bernal, Richard L. 1984. “The IMF and Class Struggle in Jamaica, 1977-1980.” Latin American Perspectives 11(3):53–82. Bernburg, Jón Gunnar. 2015. “Economic Crisis and Popular Protest in Iceland, January 2009: The Role of Perceived Economic Loss and Political Attitudes in Protest Participation and Support*.” Mobilization: An International Quarterly 20(2):231–52. BOG. 2011. Summary of The Annual Report 2010. Athens, Greece: Bank of Greece. BOG. 2012. Summary of The Annual Report 2011. Athens, Greece: Bank of Greece. BOK. 2015. “Economic Statistics System.” Bullard, Nicola, Walden Bello, and Kamal Mallhotra. 1998. “Taming the Tigers: The IMF and the Asian Crisis.” Third World Quarterly 19(3):505–55. Campbell, Andrea Louise. 2003. “Participatory Reactions to Policy Threats: Senior Citizens and the Defense of Social Security and Medicare.” Political Behavior 25(1):29–49. Chang, Dae-oup. 2002. “Korean Labour Relations in Transition: Authoritarian Flexibility?” Labour, Capital and Society / Travail, Capital et Société 35(1):10–40. 47 Císar, Ondřej, and Jiří Navrátil. 2016. “At the Ballot Boxes or in the Streets and Factories: Economic Contention in the Visegrad Group.” Pp. 35–53 in Austerity and Protest: Popular Contention in Times of Economic Crisis, edited by M. Giugni and M. T. Grasso. Burlington, VT: Routledge. Collier, David. 1991. “The Comparative Method.” Pp. 105–19 in Political Science: The State of the Discipline II, edited by A. W. Finifter. Washington, D.C.: The American Political Science Association. Dalakoglou, Dimitris, and Christos Giovanopoulos. 2011. “From Ruptures to Eruption: A Genealogy of Post-Dictatorial Revolts in Greece.” Pp. 91–114 in Revolt and Crisis in Greece: Between a present yet to pass and a future still to come, edited by A. Vradis and D. Dalakoglou. Oakland, Baltimore, Edinburgh, London & Athens: AK Press & Occupied London. Diani, Mario, and Maria Kousis. 2014. “The Duality of Claims and Events: The Greek Campaign Against the Troika’s Memoranda and Austerity, 2010-2012.” Mobilization: An International Quarterly 19(4):387–404. Dinas, Elias. 2010. “The Greek General Election of 2009: PASOK – The Third Generation.” West European Politics 33(2):389–98. EC. 2010. The Economic Adjustment Programme for Greece. Brussels, Belgium: European Commission Directorate-General for Economic and Financial Affairs. Eisinger, Peter K. 1973. “The Conditions of Protest Behavior in American Cities.” American Political Science Review 67(01):11–28. Emery, Robert F. 2001. Korean Economic Reform: Before and since the 1997 Crisis. Aldershot: Ashgate. European Communities. 1992. “Treaty on European Union.” Official Journal of the European Communities 35:1–112. Flesher Fominaya, Cristina. 2015. “Debunking Spontaneity: Spain’s 15-M/ Indignados as Autonomous Movement.” Social Movement Studies 14(2):142– 63. Gamson, William A., and David S. Meyer. 1996. “Framing Political Opportunity.” Pp. 275–90 in Comparative Perspectives on Social Movements, edited by D. McAdam, J. D. McCarthy, and M. N. Zald. Cambridge, UK; New York: Cambridge University Press. Georgiadou, Vasiliki, Anastasia Kafe, Spyridoula Nezi, and Costis Pieridis. 2019. “Plebiscitarian Spirit in the Square. Key Characteristics of the Greek 48 Indignants.” International Journal of Politics, Culture, and Society 32(1):43– 59. Gerbaudo, Paolo. 2017. “The Indignant Citizen: Anti-Austerity Movements in Southern Europe and the Anti-Oligarchic Reclaiming of Citizenship.” Social Movement Studies 16(1):36–50. Giugni, Marco, and Maria T. Grasso. 2016. Austerity and Protest: Popular Contention in Times of Economic Crisis. 1 edition. Burlington, VT: Routledge. Goodwin, Jeff, James M. Jasper, and Jaswinder Khattra. 1999. “Caught in a Winding, Snarling Vine: The Structural Bias of Political Process Theory.” Sociological Forum 14(1):27–54. Grasso, Maria T., and Marco Giugni. 2016. “Protest Participation and Economic Crisis: The Conditioning Role of Political Opportunities.” European Journal of Political Research 55(4):663–80. Grey, Kevin. 2008. “The IMF Crisis and the Crisis of the Korean Labour Movement.” Pp. 241–64 in South Korea: Challenging Globalisation and the Post-Crisis Reforms, edited by Y.-C. Kim, D.-J. Kim, and Y. Kim. Oxford: Chandos. Haggard, Stephan. 2000. The Political Economy of the Asian Financial Crisis. Washington, D.C.: Institute for International Economics. Hylmö, Anders, and Magnus Wennerhag. 2016. “Does Class Matter in Anti-Austerity Protests? Social Class, Attitudes toward Inequality, and Political Trust in European Demonstrations in a Time of Economic Crisis.” Pp. 83–107 in Austerity and Protest: Popular Contention in Times of Economic Crisis, edited by M. Giugni and M. T. Grasso. Burlington, VT: Routledge. IMF. 1997. “Korea Letter of Intent.” https://www.imf.org/external/np/loi/120397.htm IMF. 2014. World Economic Outlook: Legacies, Clouds, Uncertainties. Washington, D.C.: IMF. Ioannou, Christos A. 2000. Trade Unions in Greece : Development, Structures & Prospects. Electronic Edition. Bonn: Friedrich Ebert Stiftung Library. Kanellopoulos, Kostas, Konstantinos Kostopoulos, Dimitris Papanikolopoulos, and Vasileios Rongas. 2017. “Competing Modes of Coordination in the Greek Anti-Austerity Campaign, 2010–2012.” Social Movement Studies 16(1):101– 18. Kang, Won-Taek, and Hoon Jaung. 1999. “The 1997 Presidential Election in South Korea.” Electoral Studies 18(4):599–608. 49 Karakatsanis, Leonidas. 2016. “Radicalised Citizens vs. Radicalised Governments? Greece and Turkey in a Comparative Perspective from the December 2008 Uprising to the 2013 Gezi Park Protests.” Journal of Contemporary European Studies 24(2):255–79. Karamichas, John. 2009. “The December 2008 Riots in Greece: Profile.” Social Movement Studies 8(3):289–93. Kaufman, Robert R. 1986. “Democratic and Authoritarian Responses to the Debt Issue: Argentina, Brazil, Mexico.” Pp. 187–217 in The Politics of International Debt, edited by M. Kahler. Ithaca: Cornell University Press. Kim, Hyuk-Rae. 2000. “The State and Civil Society in Transition: The Role of Non- Governmental Organizations in South Korea.” The Pacific Review 13(4):595– 613. Kim, Seung-Kyung, and John Finch. 2002. “Living with Rhetoric, Living against Rhetoric: Korean Families and the IMF Economic Crisis.” Korean Studies 26(1):120–39. Kitschelt, Herbert P. 1986. “Political Opportunity Structures and Political Protest: Anti-Nuclear Movements in Four Democracies.” British Journal of Political Science 16(1):57–85. Koo, Hagen. 2001. Korean Workers: The Culture and Politics of Class Formation. Ithaca: Cornell University Press. Koopmans, Ruud, and Dieter Rucht. 2002. “Protest Event Analysis.” Pp. 231–59 in Methods of Social Movement Research, edited by B. Klandermans and S. Staggenborg. Minneapolis: University of Minnesota Press. Kretsos, Lefteris. 2011. “Union Responses to the Rise of Precarious Youth Employment in Greece: Precarious Youth Employment and Trade Unions in Greece.” Industrial Relations Journal 42(5):453–72. Kriesi, Hanspeter. 2004. “The Political Opportunity Structure of New Social Movements: Its Impact on Their Mobilization.” Pp. 167–98 in The Politics of Social Protest: Comparative Perspectives on States and Social Movements, edited by C. Jenkins. London: Routledge. Kriesi, Hanspeter. 2014. “The Political Consequences of the Economic Crisis in Europe: Electoral Punishment and Popular Protest.” Pp. 297–333 in Mass politics in tough times: opinions, votes and protest in the Great Recession, edited by N. G. Bermeo and L. M. Bartels. Oxford: Oxford University Press. Kritidis, Gregor. 2014. “The Rise and Crisis of Anarchist and Libertarian Movement in Greece, 1973- 2012.” Pp. 63–94 in The City Is Ours: Squatting and 50 Autonomous Movements in Europe from the 1970s to the Present, edited by B. van der Steen, A. Katzeff, and L. van Hoogenhuijze. Oakland, CA: PM Press. Kurer, Thomas, Silja Häusermann, Bruno Wüest, and Matthias Enggist. 2019. “Economic Grievances and Political Protest.” European Journal of Political Research 58(3):866–92. Kwon, Soonman, and Ian Holliday. 2007. “The Korean Welfare State: A Paradox of Expansion in an Era of Globalisation and Economic Crisis: The Korean Welfare State.” International Journal of Social Welfare 16(3):242–48. Lee, Byoung-Hoon. 2002. “Labor Unions in the Evolution of Social Dialogue: Korea’s Experience.” International Employment Relations Review 8(1):47–55. Lee, Byoung-Hoon. 2007. “Militant Unionism in Korea.” Pp. 155–72 in Strikes around the world: case studies of 15 countries, edited by S. van der Velden, H. Dribbusch, D. Lyddon, and K. Vandaele. Amsterdam: Aksant. Lee, Hogap. 1998. “60gae si-min-dan-che min-jung-dae-hoe: 3man-yeo-myeong chung-dol-eobs-i kkeut-na (60 civic groups demonstrated: 30,000 rallied without clashes).” Donga Ilbo, November 9, 22. Lee, Kyu-Sung. 2011. The Korean Financial Crisis of 1997: Onset, Turnaround, and Thereafter. Washington, D.C.: World Bank; Korea Development Institute. Lee, Wonhong. 1998. “Min-no-chong deung 52gae dan-che eo-je 2,500myeong jib- hoe (52 groups with 2,500 participants, headed by KCTU, rallied yesterday).” Donga Ilbo, May 31, 14. Lialiouti, Zinovia, and Giorgos Bithymitris. 2017. “A Nation under Attack: Perceptions of Enmity and Victimhood in the Context of the Greek Crisis.” National Identities 19(1):53–71. Little, Andrew T. 2015. “Communication Technology and Protest.” The Journal of Politics 78(1):152–66. Mahoney, James. 2012. “The Logic of Process Tracing Tests in the Social Sciences.” Sociological Methods & Research 41(4):570–97. Martin, I. W., and N. Gabay. 2013. “Fiscal Protest in Thirteen Welfare States.” Socio- Economic Review 11(1):107–30. Martin, Isaac William. 2013. Rich People’s Movements: Grassroots Campaigns to Untax the One Percent. Oxford ; New York: Oxford University Press. McAdam, Doug. 1982. Political Process and the Development of Black Insurgency, 1930-1970. 2nd ed. Chicago: University of Chicago Press. 51 McAdam, Doug, Sidney G. Tarrow, and Charles Tilly. 2001. Dynamics of Contention. Cambridge; New York: Cambridge University Press. Meyer, David S., and Debra C. Minkoff. 2004. “Conceptualizing Political Opportunity*.” Social Forces 82(4):1457–92. Michailidou, Asimina. 2017. “‘The Germans Are Back’: Euroscepticism and Anti- Germanism in Crisis-Stricken Greece.” National Identities 19(1):91–108. Mitsopoulos, Michael, and Theodore Pelagidis. 2011. Understanding the Crisis in Greece. London: Palgrave Macmillan UK. Morell, Mayo Fuster. 2012. “The Free Culture and 15M Movements in Spain: Composition, Social Networks and Synergies.” Social Movement Studies 11(3– 4):386–92. Neary, Michael. 2000. “Hyundai Motors 1998-1999: The Anatomy of a Strike.” Capital & Class 24(1):1–7. OECD. 1998. OECD Economic Surveys: Korea 1998. Vol. 1998. OECD Publishing. OECD. 2011. OECD Economic Surveys: Greece 2011. Vol. 2011. OECD Publishing. OECD. 2014a. Employment Rate. OECD Publishing. OECD. 2014b. General Government Deficit. OECD Publishing. OECD. 2014c. Household Disposable Income. OECD Publishing. OECD. 2014d. Inflation (CPI). OECD Publishing. Pappas, Takis S. 2014. Populism and Crisis Politics in Greece. London: Palgrave Macmillan UK. Park, Cheonho, Joohoon Lee, and Jeeyoung Kim. 1998. “No-dong-jeol do-sim gyeog- lyeol-si-wi: min-no-chong deung 2man-myeong jib-hoe te-go-yong-an-jeong- te chog-gu (Violent demonstration on Labor Day: KCTU and twenty thousand protestors call for employment security).” Hankook Ilbo, May 2, 23. Peterson, Abby, Mattias Wahlström, and Magnus Wennerhag. 2015. “European Anti- Austerity Protests – Beyond ‘Old’ and ‘New’ Social Movements?” Acta Sociologica 58(4):293–310. Ponticelli, Jacopo, and Hans-Joachim Voth. 2011. “Austerity and Anarchy: Budget Cuts and Social Unrest in Europe, 1919-2008.” SSRN Electronic Journal. 52 Porta, Donatella della. 2015. Social Movements in Times of Austerity: Bringing Capitalism Back Into Protest Analysis. 1st edition. Cambaridge, UK; Malden, MA: Polity. Psimitis, Michalis. 2011. “The Protest Cycle of Spring 2010 in Greece.” Social Movement Studies 10(2):191–97. Reese, Ellen. 2005. “Policy Threats and Social Movement Coalitions: California’s Campaign to Restore Legal Immigrants’ Rights to Welfare.” Pp. 259–87 in Routing the Opposition: Social Movements, Public Policy, and Democracy, edited by H. Ingram, V. Jenness, and D. Meyer. Minneapolis: University of Minnesota Press. Roos, Jérôme E., and Leonidas Oikonomakis. 2014. “They Don’t Represent Us! The Global Resonance of the Real Democracy Movement from the Indignados to Occupy.” Pp. 117–36 in Spreading Protest: Social Movements in Times of Crisis, edited by D. della Porta and A. Mattoni. Colchester, UK: ECPR Press. Rootes, Chris. 1999. “Political Opportunity Structures: Promise, Problems and Prospects.” La Lettre de La Maison Française d’Oxford 10:75–97. Rüdig, Wolfgang, and Georgios Karyotis. 2013. “Beyond the Usual Suspects? New Participants in Anti-Austerity Protests in Greece.” Mobilization: An International Quarterly 18(3):313–30. Rüdig, Wolfgang, and Georgios Karyotis. 2014. “Who Protests in Greece? Mass Opposition to Austerity.” British Journal of Political Science 44(3):487–513. Sachs, Jeffrey. 1997. “The IMF and the Asian Flu.” The American Prospect, November 30, 16–21. Shin, Gi-Wook, and Kyung-Sup Chang. 2000. “Social Crisis in Korea.” Pp. 35–74 in Korea Briefing 1997-1999: Challenges and Change at the Turn of the Century, edited by K. Oh. Armonk, NY: M. E. Sharpe. Shin, Gi-Wook, Paul Y. Chang, and Sookyung Kim. 2007. South Korea’s Democracy Movement (1970-1993): Stanford Korea Democracy Project Report. Shorenstein Asia Pacific Research Center, Stanford University. Simiti, Marilena. 2015. “Rage and Protest.” Contention 3(2). Skocpol, Theda, and Margaret Somers. 1980. “The Uses of Comparative History in Macrosocial Inquiry.” Comparative Studies in Society and History 22(2):174– 97. 53 Smith, Jackie, John D. McCarthy, Clark McPhail, and Boguslaw Augustyn. 2001. “From Protest to Agenda Building: Description Bias in Media Coverage of Protest Events in Washington, D.C.” Social Forces 79(4):1397–1423. Statistics Korea. 2009. “Household Income and Expenditure Survey.” Tarrow, Sidney. 2010. “The Strategy of Paired Comparison: Toward a Theory of Practice.” Comparative Political Studies 43(2):230–59. Tarrow, Sidney G. 1994. Power in Movement: Social Movements, Collective Action, and Politics. Cambridge: Cambridge University Press. Tsakatika, Myrto. 2016. “SYRIZA’s Electoral Rise in Greece: Protest, Trust and the Art of Political Manipulation.” South European Society and Politics 21(4):519–40. Tsakatika, Myrto, and Costas Eleftheriou. 2013. “The Radical Left’s Turn towards Civil Society in Greece: One Strategy, Two Paths.” South European Society and Politics 18(1):81–99. Tzogopoulos, George N. 2012. “It’s Germany Stupid! The Greek-German Misunderstanding.” Pp. 6–9 in Germany as Viewed by Other EU Member States, edited by A. Möller and R. Parkes. Brussels: European Policy Institutes Network. Walton, John. 1989. “Debt, Protest and the State in Latin America.” Pp. 299–328 in Power and Popular Protest: Latin American Social Movements, edited by S. Eckstein. Berkeley, Los Angeles, CA: University of California Press. Walton, John, and David Seddon. 1994. Free Markets & Food Riots: The Politics of Global Adjustment. Oxford; Cambridge: Blackwell. World Bank. 2014. “Quarterly External Debt Statistics (SDDS).” Yun, Seongyi. 1997. “Democratization in South Korea: Social Movements and Their Political Opportunity Structures.” Asian Perspective 21(3):145–71. 54 Chapter 2. The Origins of the Populist Surge Abstract The unprecedented growth of populist parties across the world over the past two decades inspired an ever-growing interest in social changes that facilitated this electoral success. Leading theories offer three types of explanations. One features the economy, addressing worsening deprivation as a major stimulus for populist support. Another proposes non-economic accounts, focusing primarily on mass immigration and political distrust. There is also a hybrid explanation that stresses the joint effect of the economic and non-economic factors. This paper explores the social origins of populism drawing on an up-to-date harmonized dataset of populist parties in 27 European democracies that covers a span of 17 years (2003-2019). In order to uncover the roots of growing support for populism, the analysis decomposes populist parties into three families (far-rights, far-lefts, and non-radicals). Results show that far-right populism is fuelled by growing concern about immigration under worsening economic deprivation, that far-left populists find the most favorable environment where both economic deprivation and political distrust seriously worsen, and that non-radical populists, like their far-right counterpart, benefit from immigration worries, but win growing support only under low economic deprivation. These findings help to refine the theory of populism by offering extensive explanatory leverage over the heterogeneous origins of populism. 55 Over the past two decades, populism has been a major defining characteristic of contemporary politics across the world. Some might attribute this to a few recent political events – most notably, the Brexit referendum and the election of Donald Trump in 2016 – that provoked an explosion of news coverage in both traditional and new media. The rise of populism, however, is not simply reducible to a handful of high-profile events. Rather, it materialized in rapidly growing support for populist parties and leaders in many regions of the world spanning from Latin America, where populism flourished more than elsewhere during the post-war era, to North America and Europe and to some extent other parts of the world, such as sub-Saharan Africa, the Middle East, and Southeast Asia (Mudde and Rovira Kaltwasser 2017). This in turn encouraged many to view populism as a mainstay of mass politics since the 2000s (Mudde and Rovira Kaltwasser 2018; Müller 2016; Norris and Inglehart 2019) and some to go on a little further to declare that “[w]e are seemingly living in populist times” (Moffitt 2016:1). What then fuelled the populist surge? Clearly, greater electoral support for populism cannot be dissociated from populist parties’ issue responsiveness and strategies/tactics to mobilize voters. Consistent with the established literature on party politics (e.g., Huber and Powell 1994; Powell 2000), students of populism have documented instances in which populist campaigns and leadership could facilitate the widening of voter base (e.g., Bornschier 2012; van der Brug and Mughan 2007; Hutter and Vliegenthart 2018; Kriesi 2014). Yet, a rapidly growing volume of literature moves beyond the realm of party competition, calling attention to broader areas of 56 social life in which the growth of populism is rooted.14 This strand of literature, inspired by the unprecedented rise of populism in recent years, seeks to uncover the social origins of populism by linking the populist surge with changes in the social world that had fundamentally transformed preferences and desires among many individuals over the past few decades. This approach offers important insight into populism research, as it helps to view populism as political mobilization grounded in social changes rather than a mere byproduct of electoral competition. Relatedly, it sheds light on why and how populist parties, many of which had long remained marginal in the political landscape, could make electoral success simultaneously across many countries without notable changes in ideological stances and programs. Existing theories concerned with the social origins of populism fall into one of the following three types: economic, non-economic, and hybrid. These accounts differ from – if not necessarily exclusive of – each other with regard to the proposed primary sources of growing support for populism. Firstly, many recent studies center around the economy, addressing worsening economic circumstances as the chief driver of populism. While various economic conditions, such as downturn/recession (Blekesaune 2007; Dustmann et al. 2017; Eichengreen 2018), material hardship (Swank and Betz 2003), and globalization shocks (Rodrik 2018), were taken into consideration, much of recent research focused primarily on the role of deprivation in 14 Studies of populism concerned with the role of endogenous institutions, such as parties and state institutions, are often called “supply-side” research, whereas those geared toward exogenous environments “demand-side” research. While the present study primarily contributes to the demand-side literature, it is important to note that explaining changing support for populism requires advancing both sides of research. I will discuss this issue in the concluding section. 57 fostering electoral support for populist parties, pointing out that populists tend to receive strong support from the share of population that is economically marginalized (e.g., Algan et al. 2017; Gidron and Mijs 2019; Hernández and Kriesi 2016). This view has received ever-growing support from scholars contending that the latest electoral success of populists in the US and Europe was driven by the Great Recession and the Eurozone crisis in the late 2000s. An increasing volume of literature, however, calls into question the adequacy of the economic explanation, pointing to a limited causal linkage between economic deprivation and populist support (e.g., Doyle 2011; Mudde and Rovira Kaltwasser 2018). Some scholars direct attention to non-economic factors that might better predict electoral performances of populist parties. Among those factors are mass immigration and political distrust. Mass immigration was addressed by many studies as a major stimulus for populist voting, especially for radical rights espousing nativism. Recent scholarship on cultural backlash (Inglehart and Norris 2017; Norris and Inglehart 2019) and identity politics (e.g., Fukuyama 2018; Jardina 2019) provide theoretical elaborations, characterizing growing support for populists as collective responses to the multicultural transformation that has occurred over the last several decades. Research on political distrust was also well-established in the study of populism. Previous studies give consistent evidence that populists tend to gain extraordinary support from individuals discrediting governing institutions and politics, feeling their voice unheard and ignored (e.g., Aslanidis and Rovira Kaltwasser 2016; Doyle 2011; Schumacher and Rooduijn 2013). According to this account, populists arise as an 58 alternative political force when a serious crisis of representation emerges from contemporary democracies. There is also a third type of account, i.e., one contending that populism has both economic and non-economic roots. Proponents of this account offer a hybrid view combining the economic and non-economic explanations. This line of research usually pairs economic deprivation with either mass immigration or political distrust, pointing to their joint effect on populist voting (e.g., Kriesi and Pappas 2016; Quillian 1995; Ramiro 2016). Some research characterizes populism as an outcome that results from a concurrence of multiple crises or a bundle of factors that contain both economic and non-economic shifts (Berezin 2018; Brubaker 2017; Lynch 2019). Where does growing support for populism come from? Do we need to adjudicate between the three leading theories – economic, non-economic, and hybrid – to choose the most satisfactory account? Or, do the three theories need to be considered in non-conflicting terms, so that they can collectively help to advance our understanding of populist support? This study tackles these issues, seeking to unravel the social origins of growing electoral support for populist parties. I argue that in order to fully understand changing support for populist parties, we need to consider diverse types of populist parties separately. Specifically, I break populist parties into three mutually exclusive families – far-right, far-left, and non-radicals – and test whether and to what extent the existing theses on the social origins of populism account for changing electoral support for each populist party family. This decomposition is grounded in the ways in which populism is exercised in mass politics. As will be discussed below in detail, populism 59 is typically combined with one of the two radical ideologies when materializing in political campaigns, whereas it is in relatively fewer cases practiced in an ideologically scrupulous manner or across the partisan divide. Considering the three types of populist parties thus makes it possible to capture all existing manifestations of populism. By investigating the social determinants of electoral support for the three populist groups, this study makes two important contributions to populism research. Above all, it reveals the distinct social dynamics leading to the growth of the three populist groups. Most of recent influential studies tends to treat populists and populist parties as a single analytic unit (e.g., Brubaker 2017; Inglehart and Norris 2017) or focus exclusively on one or two of the radical types (e.g., Algan et al. 2017; Burgoon et al. 2019; Gidron and Mijs 2019). Accordingly, our understanding of changing support for the three populist groups remains fragmentary at best. Nor have the existing theories been fully assessed over all the three populist groups. This study is the first to test the leading theories vis-à-vis each other drawing on all existing types of populist parties in an attempt to unveil the social mechanisms that result in the growth of each party family. The analysis presented below will demonstrate discrepant sources of support for the three populist party families. Combined together, the heterogeneous social origins of populist support would allow us to refine the theory of populism by fleshing out the distinct routes to the growth of the three populist groups, which remains incompletely understood in prior research. Methodologically, this study offers a rigorous test of the existing explanations for the social origins of growing support for populism, dealing with two major 60 challenges that confronted prior research. One challenge for populism research, as pointed out by many scholars (e.g., Gidron and Bonikowski 2013; Mudde and Rovira Kaltwasser 2018), arises from the conceptual ambiguity of populism, which often makes it elusive to draw generalizable lessons from existing studies that adopt inconsistent, and often conflicting, definitions of populism. Another challenge stems from the cross-sectional design of many studies that examine correlates of populist support. As these studies focus primarily on how variations in individual-level traits affect populist voting, they do not tell much on the longitudinal dynamics of electoral support for populist parties. The social determinants of electoral support for populist parties over time, therefore, have yet to be systemically investigated. To resolve the conceptual ambiguity of populism, I draw on an up-to-date list of European populist parties collected and validated by over 80 scholars and journalists according to Cas Mudde’s (2004) ideational definition of populism – I discuss below why this definition is preferable over the other alternatives. This harmonized list ensures the definitional consistency across the populist parties from both cross-sectional and longitudinal perspectives, allowing us to analyze overtime variation in populist support in multiple countries. I examine electoral support for populist parties in parliamentary elections in 27 European countries from 2013 to 2019. The case of Europe over the last two decades offers a great laboratory for populism research, because it exhibits an unprecedented wave of electoral successes of populist parties in many regions. This paper proceeds as follows. The next section offers an overview of the literature on of the social origins of growing populist support. I turn then to a 61 discussion of how populism should be defined, introducing the ideational definition adopted in this paper to collect the harmonized data of populist parties and explaining why this definition is preferred over the other alternatives. The empirical analysis tests the validity of each theory of the social origins of populism over the three populist party families (far-right, far-left, and non-radicals). A discussion of theoretical implications and avenues for future research follows. THE SOCIAL ORIGINS OF POPULISM Economic explanation The worsening state of economy occupies a central place in the existing literature on the social origin of populisms (e.g., Dornbusch and Edwards 1991; Eichengreen 2018; Hernández and Kriesi 2016; Judis 2016; Kriesi and Pappas 2016). Among its other manifestations, scholars devote particular attention to increasing deprivation – one’s sense of “losing out” (Burgoon et al. 2019) – and its systemic linkage to growing voter support for populist parties. Specifically, recent cross-country analyses address unemployment as a chief driver of populist voting, evidencing that soaring unemployment electorally favors populist radical rights, which in turn undermines the voter base of established parties (Algan et al. 2017; Arzheimer 2009; Funke, Schularick, and Trebesch 2016; Jackman and Volpert 1996; Vlandas and Halikiopoulou 2019; Werts, Scheepers, and Lubbers 2013). Research also shows that populists tend to win substantially more votes in socioeconomically deprived regions or regions that have undergone a deep recession and exposed many individuals to job loss or insecurity in recent years (Algan et al. 2017; Broz, Frieden, and Weymouth 62 2019; McQuarrie 2017; Muro and Liu 2016). Increasing deprivation at the aggregate level is sometimes associated with the so-called “loser of globalization” thesis: while the cross-border flow of workforces at unprecedented scope and volume facilitated the remarkable growth in many business sectors and benefited a large share of workforces on the globe, it also created those “left behind” (Mudde and Rovira Kaltwasser 2018). At the heart of growing electoral support for populist parties over the last few decades, argue proponents of this thesis, were outrages among those who believed that they were deprived of economic opportunities and rights due to the openness of the global economy (Colantone and Stanig 2018; Ford and Goodwin 2014; Goodhart 2017). Individual-level studies delve further into the political mechanisms linking economic deprivation to populist voting, drawing on divergent aspects of deprivation. Gidron and Mijs (2019) take a straightforward approach, testing whether income losses increase the likelihood of voting for radical populist parties. Their case study of Dutch elections offers longitudinal evidence that voters experiencing income loss become more supportive of radical left parties. Morgan (2018) gives a similar account with an analysis of the American case. Much of the latest literature examines relative aspects of deprivation (Gurr 1970) and their effect on populist voting. Instead of focusing solely on within- individual income changes, Burgoon et al. (2019) devised a measure for what they call “positional deprivation”, which allows to view the extent to which each individual earns less than the other members of society, so as to estimate the level of deprivation each individual feels relative to the others. Their analysis reports that positional deprivation has strong predictive power for populist voting. In a similar vein, 63 Rooduijn and Burgoon (2018) decompose individual-level economic hardship into one under favorable macroeconomic circumstances and the other under unfavourable circumstances, and give evidence that voters experiencing hardship in good times are more likely than those experiencing in bad times to vote for radical right populists. There is also a growing volume of literature taking a more comprehensive approach to relative deprivation. Gidron and Hall (2017) highlight the importance of perceived deprivation, giving cross-sectional evidence that electoral support for populist radical right parties is particularly strong among those suffering high levels of status anxiety that originated from economic insecurity as well as cultural shifts – also see Mols and Jetten (2016) and Gidron and Hall (2019) for similar studies addressing perceived marginalization as the primary stimulus for populist voting. Non-economic explanation The causal linkage between economic deprivation and populist voting, however, remains contested. Although this linkage receives some empirical support, it should also be noted that the electoral success of populist parties over the past decades did not occur in the economically deprived countries: especially, populist radical right parties in Europe had great electoral success in the affluent economies with extensive social welfare, whereas they achieved very little in the poorest regions (Mudde 2016; Mudde and Rovira Kaltwasser 2018). Further, some recent studies of European and Latin American elections give evidence against the effect of economic deprivation on populist voting (Doyle 2011; Lubbers, Gijsberts, and Scheepers 2002; Swank and Betz 2003). 64 The mixed support for the association between economic deprivation and electoral populism might call attention to alternative accounts moving beyond the economic explanation. Among those accounts is ever-growing scholarship that addresses mass immigration as a key determinant of the recent populist surge – see Rydgren (2007:250-251) and Noury and Roland (2020:13-14) for an up-to-date review of literature on immigration and electoral populism in North America and Europe. An elaborate formulation of this account is found in the “cultural backlash” theory, a well- known thesis recently proposed by Inglehart and Norris (2017; Norris and Inglehart 2019). According to this theory, increasing electoral support for populist parties is best interpreted as reactionary opposition to a structural turn to a post-materialist, pluralist society that had occurred – particularly in North America and Western Europe – during the second half of the twentieth century. An important component of this turn was the influx of immigrants and greater emphasis on the rights and cultures of those people, which was perceived as an existential threat to individuals espousing the nativist ideals. It was in response to this threat that those “social conservatives” turn to populist parties, especially parties propagating nativism and xenophobia (Norris and Inglehart 2019). This explanation certainly echoes scholarship on “identity politics” (Fukuyama 2018; Jardina 2019; Kaufmann 2019; Knigge 1998; Koopmans et al. 2005; Müller 2016) contending that the rise of radical right parties was primarily triggered by white working-class individuals who felt a severe erosion of social identity with the advent of multicultural society and the increasing number of immigrants. Another leading candidate for alternative explanation is political distrust. Many previous studies address declining trust in political institutions as an important 65 cause of populist voting (Berman 2019; Dustmann et al. 2017; Lubbers et al. 2002; Oliver and Rahn 2016; Roberts 2006). Further, some recent empirical studies testing the effect of both economic deprivation and political distrust on electoral populism report that declining political trust is robustly associated with the growth of populist parties, whereas finding very little evidence for the effect of economic deprivation (Doyle 2011; Ramiro and Gomez 2017). An important consequence of growing distrust of political institutions is often addressed as strong discontent with institutional politics. Recent studies suggest that voters who have lost confidence in governing institutions become disenchanted with politics and political leaders, believing that political elites do not listen to there voice and feeling excluded and unheard by established parties and those in office (Lubbers et al. 2002; Norris 2005; Rooduijn and Akkerman 2017). Empirical research gives evidence that political discontent motivates voting for radical populist parties that appeal to voter frustration with mainstream parties and political elites (Schumacher and Rooduijn 2013). In the broader field of party politics, the importance of political distrust has long attracted attention from political sociologists. The best-known account is one that addresses the crisis of political representation, which occurs when many voters discredit political institutions because they believe those institutions do not serve their interests, but govern on behalf of a small group of elites or particular social classes (Lipset 1959, 1960). Not only does this crisis erode the fundamental ideal of representative democracy (since people feel unrepresented), but it can also pave the way for the rise of populism and extreme politics where parties with anti- 66 establishment ideologies capitalize on this crisis for their electoral success (Laclau 1977, 2005; see Roberts 2015 for a more recent work). Hybrid explanation Finally, instead of disregarding the explanatory power of economic deprivation, some scholars explore the possibility that it yields a partial effect on voting for populist parties. This approach holds that economic deprivation, or an overall decline in living standards, cultivates electoral support for populists. But, it also stresses that it does so only when a certain set of other conditions are met. This approach gives a hybrid explanation for populism, combining the economic and non-economic explanations discussed above. Many studies in this group of research pair economic deprivation with either mass immigration or political distrust, exploring their joint spurring effect on populist support or the mediating role of the non-economic factors. An elaborate study of the joint effect is offered by recent comparative research on European populism, published in an edited volume entitled European Populism in the Shadow of the Great Recession (Kriesi and Pappas 2016). Collecting 17 case studies of populist parties across Europe, this research finds that populist parties made great electoral success in most cases where severe economic adversity was accompanied by strong public distrust of politicians and representative bodies, whereas having faced conflicting electoral outcomes where only one of the two conditions was witnessed. The rise of populism is therefore attributed to the interplay of economic and political crises that generated collective grievances against 67 both economic hardship and institutional politics. This approach bridges the economic deprivation literature and the political distrust literature discussed above. Similarly, prior research highlights factors that mediate the effect of economic deprivation. Magni (2017) addresses political efficacy – the perception of one’s ability to influence institutional politics – as a moderator, reporting that voters experiencing economic hardship are more likely to support radical populist parties when they have limited confidence in their ability to influence politics, whereas those with high confidence tend to vote for mainstream opposition parties (also see Algan et al. 2017; Werts et al. 2013). Existing studies also point to the moderating role of nativist and anti-immigration attitudes, giving cross-sectional evidence that these attitudes tend to inspire voting for radical right parties among individuals with low socio-economic status attributing economic distress to the increasing number of foreign-born residents (e.g., Ivarsflaten 2005; Quillian 1995; Ramiro 2016; Semyonov, Raijman, and Gorodzeisky 2006; Van Der Brug, Fennema, and Tillie 2000). Some scholars take more holistic approaches. Brubaker’s (2017:377) “perfect storm” argument fits into this type of explanation, suggesting that the trans-Atlantic growth of populism in recent years stemmed from a concurrence of three major crises, “the economic, refugee, and security crises”, each triggered by the global financial crash and subsequent recession, the influx of refugees and asylum seekers from Africa, and the wave of terror attacks over the last decade – see Berezin (2018) for a similar account. The contingent convergence of the three crises in the US and some regions of Europe, argues Brubaker (2017:373), created “a ‘perfect storm’ supremely conducive to populism”. Accordingly, changing economic circumstances alone may 68 not sufficiently account for the populist surge. Instead, they need to be treated as an element of the particular set of causes that ultimately led to the dramatic rise of populism. This argument is in line with Lynch’s (2019:675) proposal for “a fully developed political economic theory of populist mobilization”, which encourages analysts to examine both conventional threats, such as economic adversity and increasing immigration, and institutional constraints, especially, the neoliberal turn among mainstream parties and growing intervention of supranational bodies in the making of economic policies. While the threats inspired anger and depression among many voters in advanced economies, the constraints motivated those voters to explore political forces that resist both mainstream parties and foreign intervention. CONCEPTUALIZING POPULISM The economic, non-economic, and hybrid explanations constitute an important part of the setting for research on the populist surge, offering the primary explanatory variables to be tested. The other key part lies in operationalizing the notion of populism and populist parties, which would provide the main response variable. The study of populism, however, requires special attention and care to defining the response variable, as the idea of populism remains exceedingly contested, and therefore the way it is operationalized largely determines both the scope and adequacy of any given populism research. The term populism has inspired a seemingly never-ending series of debates over the past several decades, yielding thousands of books, academic articles, columns, and news reports. The huge popularity of this term often led to an explosion 69 of messages and arguments equipped with heterogeneous ideas of populism, which resulted in an overwhelming definitional variety and ambiguity.15 Although the term populism is still being used among many to denote various political entities and phenomena, however, there has been a growing movement among scholars that employs a common conceptual approach. Referred to as the “ideational” approach by its protagonist, Cas Mudde, it defines populism as “an ideology16 that considers society to be ultimately separated into two homogenous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’, and which argues that politics should be an expression of the volonté générale (general will) of the people” (Mudde 2004:543; emphasis in original). Central to this approach is the Manichean distinction of the social world into “the elite” and “the people”, viewing the former as corrupt, immoral, and illegitimate and the latter as innocent, morally superior, and sovereign.17 This worldview addresses the following three properties of populism. First, populism is an anti-elitist ideology that stands for the people and against the elite, and populists are those who fight 15 For up-to-date overviews of theories of populism, see Berezin (2019), Gidron and Bonikowski (2013), and Rovira Kaltwasser et al. (2017) among many others. 16 While the ideational approach characterizes populism as an ideology, other analogous approaches, which address the moralistic representation of the people against the elite as the chief characteristic of populism, categorize populism in different ways: for instance, as a discursive rhetoric (e.g., Brubaker 2017; Hawkins 2009), a style (e.g., Kazin 1995; Moffitt 2016), a political claim (e.g., Bonikowski and Gidron 2016), or a political mobilization (e.g., Jansen 2011; Levitsky and Roberts 2011). Although such different categorizations, as Gidron and Bonikowski (2013) suggest, may facilitate research on divergent aspects of populism, I agree with Mudde and Rovira Kaltwasser (2018) that the conceptual differences between those views are too minor to alter findings of empirical research. 17 To this “vertical” opposition between ordinary people and those on the top, Brubaker (2017) adds a “horizontal” opposition that excludes some non-elites groups from “the people”. 70 against the elite in the name of the people. As the elite are typically in power to oppress the people while seeking their own interests, populists often arise as outsiders challenging the existing establishment to take over political leadership. Second, populism is an anti-pluralist ideology that denies the diversity of values and interests among social groups. As characterized by the Rousseauean “general will”, the people are believed to share the homogenous views and desires, so as to be uniformly represented as a single group, and populists solely appeal to these views and desires. This accordingly supports the core claim of populists that “they, only they, represent the people” (Müller 2016:20; emphasis in original). Third, populism is a thin-centered ideology. Originally coined by political theorist Michael Freeden (1996), the thin- centered ideology refers to “ideologies whose morphology, whose conceptual patterns and arrangements, [are] insufficient to contain the comprehensive solutions for the full spectrum of socio-political problems”, and which needs to be combined with other ideologies to “thicken out” to give such solutions (Freeden 2017:2). Mudde (2004) characterizes populism as a variant of the thin-centered ideology, in that populism gives nothing but the moralistic distinction between the people and the elite, leaving no concrete program for political mobilization. This in turn accounts for why populism “almost always appears attached to other ideological elements” (Mudde and Rovira Kaltwasser 2017:6), such as nativism, socialism, and nationalism, when it crystalizes into party positions, slogans, and campaigns. 71 I employ the ideational approach in the present study, because it has two important analytical advantages.18 First, the ideational approach offers broader applicability than many alternative approaches, permitting to view an extensive range of political actors and groups through the lens of populism (Mudde and Rovira Kaltwasser 2018). Obviously, its simplistic account that solely addresses the moralistic representation of the people against the elite19 lacks specificity and precision, especially compared to some alternative approaches delving into particular modalities of populism. The well-established literature on Latin American cases, for example, characterizes populism as an economic policy regime that prioritizes expansionary social welfare and redistribution at the expense of fiscal sustainability (Acemoglu, Egorov, and Sonin 2013; Dornbusch and Edwards 1991). Similar policy- centered approaches highlighting the irresponsible economic governance were also widely shared among studies of other cases, notably, post-authoritarian Greece in the 1980s and post-communist Eastern Europe in the 1990s and 2000s (Innes 2002; Lyrintzis 1987; Mavrogordatos 1997; Pappas 2014). Prior study also calls attention to political leadership and communicative strategy, documenting parties led by charismatic leaders who speak directly to the masses bypassing, and often neglecting, institutional channels and veto points (Taggart 1995, 2000; Weyland 2001). Some 18 In addition to these advantages, Mudde and Rovira Kaltwasser (2017, 2018) address two more advantages, arguing that the ideational approach is better-suited than its alternatives for taking into account the followings: (a) the relationship between populism and democracy and (b) both supply and demand sides of populism. Yet, no sufficient explanation was provided for why the other approaches are less useful or inadequate for investigating these two issues. Nor does it seem that any of these issues is intrinsically related to the key properties of the ideational approach. 19 This is why the ideational approach has often been called a “minimal” definition of populism (Mudde and Rovira Kaltwasser 2013). 72 research focuses on more symbolic aspects of populist leadership that exhibit unconventional behavioral patterns, deemed crude among many, to mobilize individuals with anti-elitist attitudes (Ostiguy 2017). Despite the greater specificity, however, those alternative approaches offer limited analytical leverage for empirical research. A fundamental weakness of the policy-centered approach is that even parties characterized as populist display substantial heterogeneity in enforced economic policies (Roberts 2006; Weyland 2001), which causes a serious classification problem. Similarly, the approaches centering on leadership strategies and styles are often criticized for the existence of many radical forces whose messages, slogans, and programs are considered populist, but which are not led by charismatic leaders or do not exhibit idiosyncratic styles of politics (Levitsky and Roberts 2011; Mudde and Rovira Kaltwasser 2018). The limited scope of these alternative approaches, therefore, risks underutilizing research on populism by dropping many potentially valid observations. Contrary to the approaches with the restricted boundaries, the ideational approach encompasses a broad range of political actors and groups. This offers a crucial analytical advantage, because once applied for empirical research, the ideational approach reveals that populists (i.e., those who claim to exclusively represent ordinary people while demonizing political elites) have been on the rise in recent decades in many regions of the world (Mudde and Rovira Kaltwasser 2017). This allows to view populism as a political mobilization (Jansen 2011) that sprang up at local, national, and global scales, and accordingly to analyze its spread and growth across time and space. 73 Second, the ideational approach is well-suited for taking into account the practical malleability of populism, allowing for its manifestations on both ends of the political spectrum (Mudde and Rovira Kaltwasser 2018). This turns out to be particularly relevant given the deeply ambiguous nature of “the people”, an expression that can virtually refer to any social group(s) without any validation or justification. As pointed out by many populism scholars, the persuasive power of this semantically ambiguous expression is of the very essence of populist ideology: by speaking in the name of the (“ordinary”, “powerless”, “innocent”) people, populists create solidarity among those who recognize themselves, or are recognized, as the people and hostility toward those excluded from the people (Brubaker 2017; Mudde 2004). Since “the people” is an empty category with no fixed referent (Laclau 2005), it always needs to be defined, or imagined, for political mobilization. This is why populism is properly conceptualized as a thin-centered ideology, because defining “the people” requires merging other “host” ideologies with populism (Mudde and Rovira Kaltwasser 2018). Prior research reports that populism is interwoven with divergent sets of ideologies across the left-right spectrum (Gidron and Mijs 2019; Judis 2016; Moffitt 2016; Mudde and Rovira Kaltwasser 2018; Roberts 2019). Typically, right-wing populism fits well with nativism and nationalism, which often leads to the construction of “the people” sharing the communal values (imagined to be) deeply rooted in the nation’s history and tradition. This in turn leaves the group of outsiders who do not belong to “the people” or are believed to be their enemy. Among these outsiders are immigrants and foreign-born residents as well as non- native religious and political opponents (in the European context, for instance, 74 Muslims and pro-EU citizens) (Berezin 2009; Brubaker 2017; Fukuyama 2018; Mudde 2007). The demarcation of “us” from those “aliens” was intensely propagated amid the unprecedented surge of European far-right populists in recent years that brought Marine Le Pen of Front National (FN) and Nobert Hofer of the Freedom Party of Austria (FPÖ) to the final round of the latest presidential elections, made the Euroskeptic UK Independence Party (UKIP) the third most popular party in Britain in 2015, and promoted hard-line xenophobic far-rights such as the Party for Freedom (PVV) and the Danish People’s Party (DF) to the largest opposition party in the Netherlands and Denmark. Left-wing populism usually marries with socialism and, in many cases, anti- globalization. Left-wing populists accordingly advocate the rights of the unprivileged, working-class “people” against their enemies comprising domestic economic elites as well as external threats such as transnational capital and foreign/international bodies that intervene in the domestic economy (Judis 2016; Mudde and Rovira Kaltwasser 2017; Stavrakakis and Katsambekis 2014). The mixture of socialist and anti-globalist ideologies was at the core of recent left-wing populist leadership in many regions of Latin America – notably, Evo Morales (Bolivia), Hugo Chávez (Venezuela), Rafael Correa (Ecuador), and Daniel Ortega (Nicaragua) who came into power with radical socialist and anti-imperialist reform plans. On the other side of the globe, the Eurozone crisis invigorated left-wing populism in the hardest-hit region of Europe. Among others, the highly successful electoral campaign of the Coalition of Radical Left (Syriza) in Greece exhibited a variant of populist discourses driven by socialism and anti-interventionism, seeking to represent the workers while calling for resistance 75 against the external agents (the European Central Bank, European Commission, and International Monetary Fund – dubbed “Troika”) that requested an extensive austerity reform to the Greek government as a key condition of the bailout package. DATA AND METHODS The present study explores the social origins of growing electoral support for populist parties. Since the focus is on the overall growth of electoral support for populist parties, the response variable of primary interest is aggregate vote share of populist parties in parliamentary elections in each country. Accordingly, the explanatory factors addressed by the three leading theories need to be measured/aggregated at the country level. This requires a collection of comparative cross-country data merging vote shares of populist parties in each country’s election years with the country-level predictors. The populist parties examined in the analysis below were drawn from an up-to- date expert survey – the “PopuList” (Rooduijn et al. 2019) – conducted through close cooperation between academics and The Guardian.20 This survey reviewed parties in 31 European countries that have been represented at least once in parliament, and selected those that fit the ideational definition. The latest list of populist parties offered by this survey, updated in January 2020, was peer-reviewed by over 80 scholars of 20 Many recent populism studies draw on this dataset. For comparative applications, see Foster and Frieden (2019), Ruth-Lovell, Lührmann, and Grahn (2019), and Zulianello (2020). 76 populism.21 I measured electoral support for populist parties by the total vote share (%) won by all populist parties in each parliamentary election, drawing on past election data from the ParlGov database (Döring and Magnow 2019). 1990−1999 2000−2009 2010−2019 Vote Share (%) 0 20 40 60 Figure 1. Average parliamentary vote share (%) of populist parties in EU27 and UK in the 1990s, 2000s, and 2010s. Sources. Calculated from the PopuList (Rooduijn et al. 2019) and ParlGov (Döring and Magnow 2019) data Figure 1 presents changes in electoral support for populist parties in 27 EU member states and the United Kingdom over the last three decades, mapping average total vote shares won by populist parties in parliamentary elections held during each decade: the 1990s, 2000s, and 2010s. As clearly shown on the three maps, populists have been on the rise. Not only did nearly all countries – except Malta where populist parties have not won parliamentary seats – witness increasing support for populist parties: across 28 countries, the average vote share of populist parties increased from 7 percent in the 1990s to 14 percent in the 2000s to 22 percent in the 2010s. Populists 21 The “PopuList” website (https://popu-list.org) offers a detailed overview of data and coding rules. Also see the Guardian report on the survey method (https://www.theguardian.com/world/2018/nov/20/measuring-populism-how- guardian-charted-rise-methodology). 77 made particularly remarkable success in some countries. For example, the average populist vote shares in Netherlands, Lithuania, Poland, and Bulgaria fell short of 4 percent in the 1990s, but skyrocketed to 23 percent (Netherlands), 26 percent (Lithuania), 40 percent (Poland), and 47 percent (Bulgaria) in the 2010s. Emerging populists also made great success in some countries. In Latvia and Spain, populist parties were all founded in the 2010s, and won 9 percent (Latvia) and 15 percent (Spain) vote shares on average in elections held in the 2010s. Lastly, there are countries where populists had already put themselves on a strong footing in the 1990s, but achieved even more in the 2010s. Populist parties in Hungary and Italy each won 17 percent and 36 percent vote shares on average in the 1990s, and 67 percent and 61 percent in the 2010s. Yet, while this overall upward trend in aggregate populist voting is certainly worth investigating, it is important to note that this movement has been driven by divergent types of populist parties. As discussed above, the thin-centeredness of populist ideology facilitates merging with other “host” ideologies that offer concrete identities of “the people” to be advocated. Most notable among those host ideologies are far-right and far-left ideologies. As discussed earlier, these extreme ideologies help to “thicken out” the thin-centered populist ideology, filling the empty category of the people with particular social groups. Prior research on extreme politics gives good reason to consider populists with far-right and far-left orientations separately, detecting distinct voting patterns for parties on each extreme end (e.g., Gidron and Mijs 2019; Rooduijn and Burgoon 2018). Discerning far-rights and far-lefts from the sample of populist parties is possible without referring to external sources, as the 78 “PopuList” dataset provides peer-reviewed lists of well-defined far-right and far-left parties: far-rights are defined as parties that advocate nativism and authoritarianism (Mudde 2007), and far-lefts as those that deny the capitalist arrangement of economic and political structures and pursue radical redistribution of resources (March 2011). Extracting parties that are both populist and far-right and those that are both populist and far-left, therefore, allows to obtain subsamples of populist far-right and populist far-left parties. This subsampling leaves a third group, i.e., populist parties that do not belong to either far-right or far-left party families. I refer to these parties as non-radical populists, because their populist appeal do not strictly rely on a particular type of extreme ideology but is driven by rather weak partisan orientations. While the populist ideal – the moral distinction between “the people” and “the elite” – lies at the heart of these parties’ political mobilization, these parties, unlike the other two populist party families, lack the common specific ideology that provides a concrete identity of the people. These parties in turn tend to appeal to broad masses, inspiring general grievances against political elites, or to move across the partisan divide in the pursuit of electoral benefits. See Appendix Table A2 for the entire sample of populist parties from the three party families. 79 Greece 40 20 0 Poland 40 Far−left Far−right 20 Non−radical 0 Slovakia 40 20 0 1990 2000 2010 2020 Year Figure 2. Parliamentary vote share (%) of each populist family in Greece, Poland, and Slovakia during 1990-2019 Sources. Calculated from the PopuList (Rooduijn et al. 2019) and ParlGov (Döring and Magnow 2019) data The analysis below examines the social origins of electoral support for each of the three families of populist parties (far-rights, far-lefts, and non-radicals), drawing on total parliamentary vote shares won by each party family. This strategy has a clear advantage over analyzing the entire sample of populist parties all at once, because the three party families not only evolved at different paces but also flourished in different places. As noted above, the average vote share of all populist parties in 28 European countries increased from 7 percent in the 1990s to 14 percent in the 2000s to 22 percent in the 2010s. Once decomposing these figures into the average vote shares of 80 Parliamentary Vote Share (%) the three subgroups, we see that the three party families grew at different paces: far- right populists won 4 percent, 7 percent, and 10 percent, far-left populists 1 percent, 2 percent, and 3 percent, and non-radical populists 2 percent, 5 percent, and 8 percent in the 1990s, 2000s, and 2010s. The three party families also thrived in different regions. Figure 2 plots their vote shares between 1990 to 2019 in three selected countries where the electoral success of populists stemmed from each of them: as displayed on each graph, the populist surge was led by far-rights in Poland, by far-lefts in Greece, and by non-radicals in Italy. What then drives growing electoral support for each of the three populist party families? I now turn to measures of predictor variables chosen to test the three major explanations of the social origins of populism. First, economic deprivation was measured by three indicators: unemployment rate, relative poverty gap, and Gini index. As discussed earlier, unemployment was proposed by many proponents of the economic explanation as a basic measure of deprivation (e.g., Algan et al. 2017; Arzheimer 2009; Funke et al. 2016), since it results in economic losses that often have the greatest impact on living standards for many individuals and households. The other two well-known indicators were chosen to take into account the insights from the literature that highlights the importance of relative deprivation (e.g., Burgoon et al. 2019; Rooduijn and Burgoon 2018). Both the Gini index and poverty gap gauge relative deprivation through income inequality. It is important to note, however, that these two indicators differ with regard to what they focus primarily on. The Gini index assesses the overall relative deprivation by examining income disparity across the entire population. It accordingly tells the extent to which wealth is unequally 81 distributed among all residents of a country. The poverty gap gives a more specific view of relative deprivation, directing attention to low-income individuals and evaluating the level of deprivation this social group undergoes relative to the better- offs. Existing poverty gap indices measure relative deprivation against a fixed poverty line. The Eurostat poverty gap index adopted in this paper sets the poverty line at 60 percent of the median disposable income, and computes the distance between the poverty line and the median income of those below the poverty line: the magnitude of this distance is expressed as a percentage of the poverty line. Testing the effects of the Gini index and poverty gap, therefore, give a comprehensive account of the linkage between relative deprivation and the growth of populist parties, analyzing this linkage among both the whole population and the most deprived share. Alongside the measures of economic deprivation, I also considered a broader set of economic factors treated in prior study as facilitators or catalysts for populist voting. Among these factors are poor economic performances (Blekesaune 2007; Dustmann et al. 2017; Eichengreen 2018), weak welfare states (Swank and Betz 2003), and globalization shocks (Rodrik 2018). I estimated the effects of each factor on electoral support for populism using real GDP growth and interest rates, social spending, and foreign direct investment. All measures of the economic variables were drawn from the Eurostat and AMECO database, which offer the largest harmonized data for EU member and affiliated states. Next, I took account of mass immigration, one of the key non-economic drivers of populist support addressed by the state-of-the-art literature (e.g., Dustmann et al. 2017; Kriesi and Pappas 2016; Norris and Inglehart 2019; Semyonov et al. 82 2006). I examined the influence of mass immigration by assessing how changing public concern about immigration affects populist vote share. The measure of public concern about immigration was drawn from the standard Eurobarometer (EB)22, a series of cross-national surveys conducted every year in all EU members and some affiliated states. I attended to an item that asked respondents to choose “two most important issues facing their country”, and analyzed the share of respondents that mentioned immigration as one of their two choices.23 To test the electoral impact of mass immigration on populism, one might also want to draw on a measure of immigration inflows along with the attitudinal measure. While it reduces the sample size due to the limited data availability, the inclusion of immigration inflows does not appear to change the substance of findings from models estimated with the attitudinal measure of immigration – see Appendix Table A3.24 Along with public concern about immigration, I considered public attitudes toward European integration, following the established literature that views Euroscepticism as a proxy for anti-immigrant attitudes 22 In case EB surveys conducted in the same year contain items that ask an identical question, I aggregated the shares of respondents across the surveys to obtain an annual average. 23 There exist cross-national data, such as European Social Survey and International Social Survey Programme, that offer more precise measures of anti-immigrant attitudes: notably, items assessing respondents’ views on increasing immigrants and the economic impact of immigration. These measures, however, are offered for only one or two survey waves and therefore of very limited longitudinal use. 24 The only difference is found in the model outcome for non-radical populism: as reported in Table 3, the interaction terms between the attitudinal measure of immigration and relative deprivation (poverty gap and Gini index) are negative and statistically significant; however, the interactions between immigration inflows and relative deprivation are inconsistent (the interaction with Gini is positive) and do not gain statistical support. The models of non-radical populism estimated with immigration inflows, however, need to be interpreted with great caution, because the sample size is reduced more than half due to the limited data availability of immigration flows. 83 and xenophobia (Taggart 1998; Taggart and Szczerbiak 2004). The measure of Euroscepticism was drawn from an EB item that asked the respondents if the EU membership is “bad” for their country, and the share of negative responses was taken. I then took into account the role of the other established non-economic factor: political distrust. The measure of political distrust was drawn from three EB items, which asked respondents whether they “tend to” or “tend not to” trust their national government, parliament, and political parties, respectively. I aggregated negative responses from each item to obtain the share of respondents expressing distrust, and then merged these aggregated responses (Cronbach's 𝛼 = .94) to compute a factor score estimate.25 The compiled dataset of all available measures yields a sample of 27 European countries – Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, and United Kingdom – observed between 2003 and 2019. Vote shares of 25 While many studies of populism consider economic deprivation and political distrust separately, some (e.g., Algan et al. 2017) might see them as causally dependent (i.e., increasing economic deprivation raising political distrust), which in effect may find evidence from prior research (e.g., Solt 2008; Zmerli and Castillo 2015). And yet, there is good reason to treat the two factors independently when modeling populist support. That is, economic deprivation does not always accompany greater political distrust. This is primarily because there are other factors – most notably, partisanship (e.g., Evans and Andersen 2006; Ladner and Wlezien 2007) – that not only exert greater impact on political distrust but also mediate the effect of economic deprivation on political distrust. This is why, as reported by recent comparative research (Kriesi and Pappas 2016), the levels of economic deprivation and political distrust seldom matched each other amid the surge of populism in Europe. Nor do the two factors raise the issue of collinearity, measured by the Variance Inflation Factors, in the models estimated in this study. 84 populist parties were collected from election years, which obviously vary across the countries, whereas the predictor variables are annual estimates. The dataset thus takes the (unbalanced) time-series cross-section format in which the unit of analysis is country-year. See Appendix Table A1 for descriptive statistics of all the variables. Model specifications The analysis below examines the social determinants of aggregate electoral support for each of the three populist party families by running a series of multivariate regression models predicting changes in the vote share won by each party family. An important issue that needs to be dealt with in specifying the models here is the bounded nature of the response variable. That is, like many other fractional or proportional variables (such as the share of income spent on political donations, turnout rates, and protest participation rates), vote share must lie between 0 and 1. This raises a well-known estimation issue for linear models, because ordinary least-squares (OLS) regression assumes that the response variable is unbounded and therefore could take any value on the real number line. Practically, OLS models estimated with the proportional response variable may give predicted values falling outside the unit interval. To resolve this issue, I estimate a fractional response model (FRM), a class of generalized linear model (GLM) developed by Papke and Wooldridge (1996) as a robust estimation method with satisfying econometric properties for fractional or proportional response variables. This modelling strategy offers two important advantages. For one, its functional form makes no assumption about a structure underlying the data generating process, while ensuring fully robust estimation under 85 the GLM assumptions. This feature is particularly relevant to the analysis of fractional data collected from diverse groups like vote shares from many different elections. For another, the functional form of FRM remains well defined even when the probability of the response variable being 0 or 1 is greater than 0. This property makes FRM preferable for the study of populist parties over the popular logistic transformation alternatives in electoral studies (e.g., Katz and King 1999; Tomz, Tucker, and Wittenberg 2002) that are inapplicable to observations with 0 or the unit value. In effect, quite a few populist parties often end up winning no parliamentary votes, so that the inability of handling 0s, which may end up dropping or systemically transforming all 0s, risks yielding biased estimates of treatment effect. The models estimated below take the following form: 𝐸 𝑉!,!,!|𝕩 = 𝐺 𝕩!,!!!Β+ 𝑉¬!,!,!𝛾 + 𝛿! ¬! where 𝐺(∙) is a probit function fulfilling 0 < 𝐺 𝑧 < 1 for all 𝑧 ∈ ℝ. 𝐸 𝑉!,!,!|𝕩 ∈ (0,1) is the expected share of the vote won by populist party family 𝑖 in country 𝑗 in election year 𝑡. 𝕩!,!!! is a row vector of the country-level predictors for country 𝑗 measured in year 𝑡 − 1. The models consider lagged values of the predictors to ensure that the predictors reflect the state of each country prior to elections. Β is a vector of corresponding regression coefficients. The models also control for the vote shares of the other two populist party families, 𝑉¬!,!,!, that might be associated with the vote 86 share of populist party family 𝑖. Coefficient 𝛾 estimates this association. 𝛿! is the country-fixed effects term. RESULTS Below I report findings from three sets of multivariate analyses. Each set predicts overtime changes in electoral support for each of the three populist party families: far- right, far-left, and non-radicals. It is important to note that not all populist parties competed for all elections held between 2003 and 2019. Each model set, therefore, restricts attention to each country’s election years when at least one party from the populist party family of interest competed. This treatment makes the sample sizes differ across the three sets of models.2627 All the analyses begin by testing the individual effect of the established factors addressed by the economic and non- 26 The whole analysis was replicated using the entire sample (see Appendix Table A4). This replication gives highly similar estimates of the main effects for all the three populist party families. While the estimates overall remain statistically significant for far-right populism and far-left populism, however, they do not for non-radical populism. This difference is likely to stem from the fact that many of non-radicals have relatively shorter histories – 13 out of 35 non-radical populist parties were founded in the 2010s – than parties from the two radical populist groups. This results in limited variation in electoral performances of non-radicals during much of the time period (2003-2019) chosen for the present study, and in turn gives good reason to restrict the scope of analysis to election years when each chosen populist party family competed. 27 To check whether or not the findings from this paper are largely drawn from a few influential cluster of observations, the whole analysis was also replicated with the jackknife treatment, running the entire set of models dropping countries in the sample one at a time. Overall, the results show that the findings from this paper are robust: the main findings for far-right populism (the spurring effect of immigration conditioned by unemployment) appear to be consistent and retain statistical significance in 22 out of 23 trials; those for far-left populism (the interactions between unemployment/poverty gap and political distrust) in 11 out of 12 trials; those for non- radical populism (the interactions between immigration and poverty gap/Gini) in 12 out of 16 trials. 87 economic explanations, and then consider interactions between the economic and non- economic factors to capture the insights from the hybrid explanations. Far-right populists Table 1 reports the probit coefficients and heterogeneity-consistent standard errors from the multivariate fractional response models of overtime variation in electoral support for far-right populist parties. As noted earlier, all predictor variables were lagged to capture the state of economy and public opinion prior to each parliamentary election. I estimated three subsets of models (1-3, 4-6, 7-9) to test the three measures of economic deprivation – unemployment, poverty gap, and Gini index – separately while controlling for all else. In each subset, the first models (1, 4, 7) test the effect of the main predictors. As reported by these models, economic deprivation is not predicted to fuel electoral support for far-right populists, as the coefficients on the three indicators are all negative. The two measures of relative deprivation (poverty gap and Gini) fail to receive statistical support, whereas the negative impact of unemployment appears to be significant at the .05 alpha-level: the role of unemployment, however, can be better understood in its interaction with immigration, as will be discussed shortly. It is also noteworthy that none of the conventional macroeconomic variables gains statistical support. This in turn casts serious doubt on the validity of the economic explanation for the rise of far-right populists in Europe. 88 Measure of Economic Deprivation Unemployment Poverty gap Gini M1 M2 M3 M4 M5 M6 M7 M8 M9 Unemployment!!! -.02* -.01 -.03* (%) (.01) (.01) (.01) Poverty Gap!!! -.02 -.02 -.02 (%) (.01) (.01) (.02) Gini!!! -.02 -.01 -.02 (%) (.01) (.01) (.02) GDP growth!!! -.004 -.004 -.01 -.001 .003 -.001 -.005 - -.004 (real, %) (.01) (.01) (.01) (.01) (.01) (.01) (.01) .0009 (.01) (.01) Interest rate!!! -.02 -.02 -.02 -.03 -.03 -.03 -.03 -.03 -.03 (%) (.01) (.02) (.02) (.02) (.02) (.02) (.02) (.02) (.02) FDI!!! .001 .0008 .0007 -.001 - -.001 .001 001 .0007 (% of GDP) (.004) (.004) (.004) (.005) .0003 (.005) (.005) (.005) (.005) (.005) Social Spending!!! .07 .06 .03 .06 .08 .06 .05 .06 .04 (% of GDP) (.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04) (.04) Political distrust!!! -.02 .02 -.01 -.09 .29 -.03 -.06 .56 -.05 (.12) (.15) (.12) (.11) (.27) (.12) (.11) (.37) (.12) Immigration concern!!! .01** .01** -.001 .01** .01** .006 .01** .01** -.004 (% of r) (.003) (.003) (.005) (.003) (.003) (.01) (.003) (.003) (.02) Euroscepticism!!! -.002 -.002 -.001 - .001 -.0007 -.001 - -.002 (% of r) (.008) (.008) (.008) .0006 (.008) (.008) (.008) .0004 (.008) (.008) (.008) Interactions Unemp!!!×Distrust!!! -.006 (.01) Unemp!!! .002** ×Immigration!!! (.0007) Pov Gap!!!×Distrust!!! -.01 (.01) Pov Gap!!! .0002 ×Immigration!!! (.0007) Gini!!!×Distrust!!! -.02 (.01) Gini!!!×Immigration!!! .0005 (.0009) 𝑁† 85 85 85 81 81 81 81 81 81 BIC 200.9 205.4 205.3 197.0 201.3 201.3 197.0 201.3 201.3 Table 1. Fractional response probit models of far-right populist support. Note: Heterogeneity-consistent standard errors in parentheses. All models control for vote shares of far-left and non-radicals. *p<.05; **p<.01 (two-tailed significance levels). †To examine only elections in which at least one far-right populist party competed, the scope of analysis was restricted to the following cases: Austria, Belgium, Bulgaria, Czech Republic, Denmark, Finland, France, Greece, Hungary, Italy, Poland, Romania, Slovakia, Slovenia, Sweden, UK (all, 2003-2019), Croatia (2007-2019), Estonia (2015-2019), Germany (2013-2019), Lithuania (2016-2019), Netherlands (2006-2019), Portugal (2019), and Spain (2015-2019). 89 Among the two established non-economic factors, political distrust turns out to be hardly relevant. Not only are its the predicted effects reported to be negative in Model 1, 4, and 7, but the overwhelmingly large standard errors give little evidence for the causal linkage between declining trust of political institutions and the growth of far-right populists. On the contrary, immigration, the other established non-economic factor, receives consistent statistical support, as the coefficient on public concern about immigration is positive and significant in Model 1, 4, and 7. The magnitudes of the predicted effects are also stable across the three models. This finding bolsters the recent literature that addresses far-right populists as the major beneficiaries of growing anti-immigrant sentiment (Mudde 2007; Mudde and Rovira Kaltwasser 2018). Theoretically, it also lends support to the influential theses on cultural backlash (Norris and Inglehart 2019) and identity politics (Fukuyama 2018; Jardina 2019) attributing the recent surge of populism to ethic and social groups that took mass immigration and multicultural shifts as an existential threat to their lives. Interestingly, however, the effect of Eurosceptic sentiment, which has been thought to in part reflect xenophobia and nativism, appears to be marginal and insignificant across all the model specifications. One possible reason for this might be that Euroscepticism is by no means intrinsic to far-right populism. In effect, many far-lefts and non-radicals are also opposing the EU or the ideal of European integration, contending that European integration winds up marginalizing peripheral members, or criticizing the lack of transparency within the EU’s decision-making processes (Caruso 2017; Stavrakakis and Katsambekis 2014). The existence of Euroscepticism across the ideological 90 spectrum is therefore likely to break up Eurosceptic votes into the three populist party families rather than attract them into far-right populists. 0.5 Unemployment High Low 0.4 0.3 0.2 0.1 0 10 20 30 Immigration concern (% of r) Figure 3. Marginal effect of public concern about immigration on far-right populist support conditioned by unemployment (high=13.8%; low=4.4%) Note: Vote fractions convertible to vote percents for interpretive convenience, e.g., .01 to 1 percent of the vote. The subsequent two models (2-3, 5-6, 8-9) in each model subset consider interaction effects. Following the literature on the hybrid explanation of populism, I interacted economic deprivation with political distrust and immigration. A key finding here lies in a positive and significant interaction between unemployment and immigration (see Model 3), implying that the spurring effect of public concern about immigration28 on far-right populist support becomes even stronger under high 28 The positive association between the attitude toward immigration and far-right populist support certainly offers strong evidence for the spurring effect of the former on the latter. There is, however, one important caveat to this interpretation. That is, 91 Vote share (fraction) unemployment. It adds to this implication that the interaction terms between public concern about immigration and the other two measures of economic deprivation are also positive – although the large standard errors give little statistical support for these estimates. Figure 3 presents the marginal effect of public concern about immigration on far-right populist vote share predicted by Model 3. To allow for the conditioning role of unemployment, the graph displays two sets of predicted values: one under high unemployment (thick line) and the other under low unemployment (dashed line). High and low levels of unemployment are each characterized by one standard deviation above and below the sample mean of unemployment. The difference is remarkable. Under low unemployment, the predicted vote share of far-right populists increases by 2 percentage points when the share of EB respondents concerned about immigration goes up from 10 percent to 20 percent, whereas it increases by more than 11 percentage points under high unemployment. This certainly sheds light on the recent surge of far-right populists in countries that underwent either or both of a prolonged recession or a rise of nativist sentiment. For instance, Hungary went through a long downturn from the late 2000s that accompanied double-digit unemployment rates for 5 years (2009-2013). Once the economy bounced back, the Hungarian public began to express growing worries about immigration: the share of EB respondents concerned growing far-right support could also have a spurring effect on the attitude toward immigration over time. If so, one would see a feedback loop between public attitudes toward immigration and public support for far-right populism (Enns and Jardina 2021). Although this possibility does not raise the issue of reverse causation (as the present analysis considers a one-year lag of the attitude toward immigration), it does offer an important avenue for further research on the self-reinforcing cycle of far-right populism. The same issue may also be addressed for the relationship between the other attitudinal variables and populist support: notably, the relationship between political distrust and far-left populist support. 92 about immigration had never exceeded 3 percent in the 2000s, but it rose up to 29 percent in 2016. Meanwhile, the average vote share of far-right populists in Hungary increased by 23 percentage points between the 2000s and the 2010s. In Sweden, Finland, and the Netherlands, the average far-right populist vote share increased 10 to 15 percentage points between the 2000s and the 2010s. While unemployment remained more or less stable in these countries despite the outbreak of the Eurozone crisis, more individuals began to see immigration as a problem during the 2010s: between 2005 and 2015, the share of EB respondents concerned about immigration soared in all of these countries (8 to 38 percent in Sweden, 8 to 16 percent in Finland, and 11 to 28 percent in the Netherlands). These results give strong evidence for the variant of the hybrid account that highlights the joint effect of mass immigration and economic deprivation on electoral support for populism, implying that far-right populists are most likely to find their constituency among economic “have-nots” attributing blame to non-native individuals (e.g., Ivarsflaten 2005; Quillian 1995). They also add to the insight on the role of economic deprivation in channelling the perceived cultural threat of immigrants into electoral support for far-right populists (Fukuyama 2018; Jardina 2019). Far-left populists Table 2 reports model outcomes with far-left populist vote share. A main result here is an interaction effect between economic deprivation and political distrust. As reported in Model 2, 5, and 8, the interaction terms between each measure of economic deprivation and political distrust are all positive, and two out of the those interaction 93 terms are significant (unemployment with political distrust) and marginally significant (poverty gap with political distrust, 𝑧 = 1.94) at the .05 alpha level. With regard to the role of economic deprivation, this result indicates that worsening economic deprivation would help far-left populists to win greater electoral support when the public highly discredits political institutions, whereas it may not otherwise. It also sheds light on which aspect of economic deprivation comes into play in the interaction with political distrust: the unemployment rate and poverty gap measure the level of deprivation among the economically disadvantaged, each focusing on those out of work and the low-income individuals, whereas the Gini index assesses the gross income disparity across the whole population. The statistical significance of the interactions between the former two measures of deprivation and political distrust accordingly implies that economically motivated support for far-left populists is most likely to come from the least well-offs. 0.4 Political distrust High 0.4 Low 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0.0 8 12 16 15 20 25 30 Unemployment (%) Poverty Gap (%) Figure 4. Marginal effect of unemployment and poverty gap on far-left populist support conditioned by political distrust (high= 1.1; low=-.9) 94 Vote share (fraction) Measure of Economic Deprivation Unemployment Poverty gap Gini M1 M2 M3 M4 M5 M6 M7 M8 M9 Unemployment!!! .06** .01 .05** (%) (.01) (.02) (.01) Poverty Gap!!! .04 .04 .03 (%) (.03) (.03) (.04) Gini!!! .06 .07 .06 (%) (.08) (.08) (.08) GDP growth!!! .01 .01 .007 .04* .02 .05 .05** .04* .06** (real, %) (.01) (.01) (.01) (.02) (.02) (.03) (.01) (.01) (.01) Interest rate!!! -.005 .009 .004 - .01 -.004 -.008 -.001 -.01 (%) (.02) (.02) (.02) .0006 (.02) (.02) (.03) (.03) (.04) (.02) FDI!!! -.01* -.01 -.01 -.01* .-01 -.02 -.02** -.01* -.02 (% of GDP) (.005) (.006) (.005) (.008) (.009) (.01) (.006) (.006) (.01) Social Spending!!! -.30 - -.28** -.13 -.14 -.19 -.17* - -.20 (% of GDP) (.06) .28** (.05) (.09) (.07) (.17) (.08) .23** (.12) (.06) (.08) Political distrust!!! 1.09** .94** .91** .96** -.53 1.21* 1.18** -.31 1.28** (.15) (.16) (.21) (.22) (.29) (.61) (.21) (1.13) (.35) Immigration concern!!! .001 .0002 -.005 -.001 -.001 .02 -.0001 .0006 .05 (% of r) (.002) (.002) (.004) (.003) (.003) (.05) (.003) (.003) (.09) Euroscepticism!!! -.01 -.01 -.007 .0004 -.01 -.001 .002 -.005 .006 (% of r) (.02) (.01) (.01) (.01) (.01) (.01) (.01) (.01) (.01) Interactions Unemp!!!×Distrust!!! .02* (.009) Unemp!!! .001 ×Immigration!!! (.0009) Pov Gap!!!×Distrust!!! .02 (.01) Pov Gap!!! -.001 ×Immigration!!! (.002) Gini!!!×Distrust!!! .05 (.03) Gini!!!×Immigration!!! -.002 (.003) 𝑁† 39 39 39 36 36 36 36 36 36 BIC 99.0 102.6 102.6 96.5 100.0 100.0 96.5 100.0 100.0 Table 2. Fractional response probit models of far-left populist support. Note: Heterogeneity-consistent standard errors in parentheses. All models control for vote shares of far-right and non-radicals. *p<.05; **p<.01 (two-tailed significance levels). †To examine only elections in which at least one far-left populist party competed, the scope of analysis was restricted to the following cases: Germany, Greece, Ireland, Netherlands, Slovakia, UK (all, 2003- 2019), Croatia (2011-2019), Cyprus (2016-2019), France (2017-2019), Lithuania (2004-2008), Slovenia (2014-2019), and Spain (2015-2019). 95 Figure 4 plots the marginal effects of unemployment and poverty gap on far- left populist vote share predicted by Model 2 and 5. As with the conditioning variable in the previous figure, high and low levels of political distrust are each set at one standard deviation above and below the sample mean. Under low political distrust, as clearly shown on the two graphs, both growing unemployment and poverty gap demonstrate very weak associations with far-left populist support, each leading to a minor decrease and increase in the vote share of far-left parties. This indicates that many voters are not likely to opt for far-left populists so long as they give credit to existing governing institutions. This result raises a serious challenge to the state-of-the-art literature on populism that validates, albeit to varying degrees, the direct spurring impact of economic deprivation on far-left populist voting (Algan et al. 2017; Burgoon et al. 2019; Gidron and Mijs 2019; Rooduijn and Burgoon 2018). Why then would worsening economic deprivation alone not be sufficient to raise far-left populist support? This insufficiency would probably stem from the fact that many of European democracies have established left-wing parties – notably, social democrats – that generally claim to represent the economically disadvantaged voters, basing much of their program on redistributive politics. As pointed out by research on class voting and welfare states (Evans 2000; Korpi 2006), the mainstream left-wing parties tend to electorally benefit from grievances and outrages generated by worsening economic deprivation. As far as many left-wing partisans view these parties as a viable alternative, therefore, far-left populists with little or no experience in parliamentary 96 politics would hardly be seen as an attractive choice. In relation to the above analysis of far-right populism, one might suggest that far-right populists have stronger competitive edge on traditional conservatives than do far-left populists on social democrats, as the hard-line stance on immigration and nativist propaganda allow them to facilitate a type of voter mobilization distinct from those of conservatives (Fukuyama 2018). Under high political distrust, both unemployment and poverty gap are found to have an evident positive impact on far-left populist support. As exhibited by the thick lines on Figure 4, 5 percentage point increases in the unemployment rate and poverty gap are each predicted to raise far-left populist vote share by approximately 4 percentage points and 6 percentage points. This result clearly demonstrates that far-left populists take advantage of worsening economic circumstances to raise voter support only when political institutions confront a serious decline in trust. Unemployment Poverty gap 0.3 High High Low 0.3 Low 0.2 0.2 0.1 0.1 0.0 0.0 −1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0 Political distrust Political distrust Figure 5. Marginal effect of political distrust on far-left populist support conditioned by unemployment (high= 13.8%; low=4.4%) and poverty gap (high=27.4%; low=17.0%) 97 Vote share (fraction) The interaction between economic deprivation and political distrust also reveals the effect of political distrust conditioned by economic deprivation. Figure 5 plots the marginal effect of political distrust on far-left populist vote share under relatively high (thick line) and low (dashed line) levels of unemployment and poverty gap set to be one standard deviation above and below their sample means. As exhibited by both plots, declining political trust is expected to spur electoral support for far-left populists even under moderate economic deprivation: when the factor score of political distrust increases from -.5 to .5, the vote share of far-left populists is predicted to rise 8 percentage points and 10 percentage points under the low levels of unemployment and poverty gap, respectively. Increasing poverty gap and Gini index further strengthen this spurring impact, as the same increase in the political distrust score is expected to raise the vote share of far-left populists by 13 percentage points under high unemployment and 14 percentage points under high poverty gap. Taken in summary, the interaction effect of economic deprivation and political distrust yields two important findings. For one, while economic deprivation spurs electoral support for far-left populists only under high political distrust, political distrust does so even when the level of economic deprivation is low. As far as far-left populist support is concerned, therefore, this lends support to the variant of the non- economic explanation that highlights the importance of declining political trust, whereas limiting the empirical validity of the economic explanation. For another, far- left populists are most likely to maximize their representation in parliament when both economic deprivation and political distrust seriously worsen. This result obviously offers strong support for the hybrid explanation that addresses the joint effect of 98 economic deprivation and political distrust (Kriesi and Pappas 2016; Magni 2017). Empirically, the result is particularly relevant to the two most successful cases, Greece and Spain, where emerging far-left populists made great electoral success in the 2010s. In Greece, Syriza, the far-left populist party founded in 2004, won less than 4 percent vote share on average in elections in the 2000s, but gained more than 30 percent in all elections since 2015. Between 2005 and 2015, the five-year moving averages of the unemployment rate, poverty gap, and political distrust score in Greece rose 15.2 percentage points, 5.4 percentage points, and 2.0 points, respectively. In Spain, the nascent far-left populist Podemos, founded in 2014, won 17 percent and 18 percent vote shares in elections held in 2015 and 2016. Like in Greece, the five-year moving averages of the unemployment rate, poverty gap, and political distrust score dramatically increased in Spain, each by 12.2 percentage points, 6.3 percentage points, and 1.9 points between 2005 and 2015. Non-radical populists The final set of analyses examines the determinants of non-radical populist support. Model outcomes are reported in Table 3. Like in the case of far-right populists, a key finding for non-radical populists is drawn from the interaction between immigration and economic deprivation. There are crucial differences, however, between the findings for the two party families. First, while it was unemployment that received statistical support for the interaction with immigration in the analysis of far-right populism, it is the other two measures of economic deprivation – the poverty gap and Gini index – that turn out to be significant for non-radical populist support. This 99 indicates that electoral support for non-radical populism is more prone to changes in the general level of deprivation one perceives relative to others – recall that poverty gap and Gini are two different measures of relative deprivation – than to immediate experiences of deprivation such as job losses. Second, contrary to far-left populists, non-radical populists are not expected to benefit from worsening economic deprivation. As reported in Model 6 and 9, the interaction terms that couple immigration with the poverty gap and Gini index are negative and significant, whereas the coefficients on immigration are positive. This suggests that the positive electoral impact of immigration on non-radical populist parties becomes weaker under worsening deprivation. Poverty gap 0.5 Gini 0.5 High High Low Low 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0 5 10 15 20 25 0 5 10 15 20 25 Immigration concern (% or r) Immigration concern (% or r) Figure 6. Marginal effect of public concern about immigration on non-radical populist support conditioned by poverty gap (high=27.4%; low=17.0%) and Gini index (high=33.9%; low=25.9%) 100 Vote share (fraction) Measure of Economic Deprivation Unemployment Poverty gap Gini M1 M2 M3 M4 M5 M6 M7 M8 M9 Unemployment!!! -.005 -.03 .003 (%) (.02) (.05) (.02) Poverty Gap!!! .0001 -.009 -.006 (%) (.04) (.04) (.04) Gini!!! -.09* -.09* -.03 (%) (.03) (.04) (.03) GDP growth!!! .01 .008 .01 .008 .005 .01 -.006 .0006 -.003 (real, %) (.02) (.02) (.02) (.02) (.02) (.02) (.02) (.02) (.02) Interest rate!!! .05 -.05 -.04 -.05 -.06 -.05 -.06 -.05 -.06 (%) (.05) (.05) (.05) (.04) (.04) (.04) (.03) (.04) (.03) FDI!!! .0007 .001 .002 -.01 -.002 -.002 -.001 -.003 -.001 (% of GDP) (.005) (.005) (.005) (.007) (.007) (.008) (.006) (.005) (.005) Social Spending!!! .01 .02 .07 -.002 -.03 .06 -.05 -.05 -.01 (% of GDP) (.11) (.11) (.13) (.09) (.09) (.10) (.06) (.06) (.07) Political distrust!!! .31 .13 .23 .23 -.45 .20 .44 -.52 .36 (.23) (.41) (.24) (.26) (.57) (.25) (.25) (.87) (.24) Immigration concern!!! .01 .01 .02* .01 .008 .07** .02* .02 .09** (% of r) (.01) (.01) (.01) (.01) (.01) (.02) (.01) (.01) (.02) Euroscepticism!!! -.01 -.01 -.01 -.01 -.01 -.007 -.01 -.01 -.01 (% of r) (.01) (.01) (.01) (.02) (.02) (.01) (.01) (.01) (.01) Interactions Unemp!!!×Distrust!!! .01 (.03) Unemp!!!×Immigration!!! -.002 (.001) Pov Gap!!!×Distrust!!! .02 (.02) Pov Gap!!!×Immigration!!! -.002* (.001) Gini!!!×Distrust!!! .03 (.02) Gini!!!×Immigration!!! -.002** (.0007) 𝑁† 42 42 42 41 41 41 41 41 41 BIC 117.4 121.2 124.9 116.0 119.7 123.3 119.6 119.6 123.3 Table 3. Fractional response probit models of non-radical populist support. Note: Heterogeneity-consistent standard errors in parentheses. All models control for vote shares of far-right and far-lefts. *p<.05; **p<.01 (two-tailed significance levels). †To examine only elections in which at least one non-radical populist party competed, the scope of analysis was restricted to the following cases: Bulgaria, Italy, Luxembourg, Lithuania, Slovakia (all, 2003-2019), Austria (2006, 2013), Belgium (2007-2019), Croatia (2011-2019), Czech Republic (2010- 2017), Finland (2019), Greece (2012-2015), Latvia (2011, 2018), Netherlands (2002-2006), Poland (2005-2015), Slovenia (2018), and Romania (2012). 101 Figure 6 plots the marginal effect of public concern about immigration on the vote share of non-radical populists under high and low levels of economic deprivation set to be above and below the sample means of the poverty gap and Gini index. As clearly displayed on the two graphs, growing concern about immigration has very limited positive impact on electoral performances of non-radical populists when the poverty gap or Gini index are high. The magnitudes of the impact, however, becomes much larger under the low poverty gap or the low Gini index, predicting a 7 to 8 percentage point increase in the vote share of non-radical populists when the share of EB respondents concerned about immigration goes up from 10 percent to 20 percent. This result offers an interesting insight into non-radical populism. Obviously, the positive electoral impact of immigration, like in the case of far-right populism, points to a causal linkage between nativist sentiment and non-radical populism. The negative moderating role of economic deprivation, however, suggests that non-radical populism is fuelled primarily by the perceived cultural threat of immigrants rather than its interplay with economic blame for them. This not only contradicts the hybrid account of populism that stresses the joint effect of immigration and economic deprivation (e.g., Ivarsflaten 2005; Quillian 1995), but is also at odds with the well- established literature on blame attribution to immigrants and ethnic minorities (e.g., Dancygier 2010; Gest 2016). How then should we understand this finding for non- radical populism? One plausible answer might lie in the fact that non-radical populists in Europe have often competed with far-right populists in elections: non-radicals faced far-rights in 14 (and far-lefts in 6) out of 16 countries where they had fielded candidates at least once during 2003-2019. As far as far-right populists are more likely 102 to win support from voters perceiving immigrants as cultural and economic threats, as evidenced by the above analysis of far-right populists, non-radical populists may have less opportunity to attract those voters, especially under worsening economic circumstances. Apart from the competition with far-right populists, it is also important to note that analyzing electoral support for non-radical populism requires particular caution, because, unlike the other two families, non-radical populists do not share any fixed set of antagonistic ideologies other than the general populist ideal, i.e., the moral representation of “the people” against “the elite”. It may come as little surprise, therefore, that non-radicals often appear to be ideologically ambiguous and scrupulous, as they lack a common frame that allows to identify “the people” and to develop a practical solution to mobilize them. The ideological heterogeneity within the non-radical populist family remains unexplored in the extant populism literature while it may add another dimension to our understanding of non-radical populism. I will discuss this issue in detail below. DISCUSSION AND CONCLUDING IMPLICATIONS The remarkable growth of populism over the last few decades has fuelled ever- growing attention and heated debates in the media and scholarly community. Concerned with the unprecedented scale of the growth as well as the illiberal and anti- pluralist tendencies of populist parties that may be consequential for contemporary democracies (Berezin 2009; Canovan 1999; Urbinati 1998), political sociologists and political scientists have made great endeavor to reveal where the populist surge came 103 from, winding up proposing a set of possible explanatory factors. This paper tested the predictive power of these factors, confirming/disconfirming their varying electoral effects on the three populist party families: the rise of far-right populism was found to be fuelled by growing concern about mass immigration and further driven under high unemployment; far-left populists find the most favorable electoral environment where both economic deprivation and public distrust of political institutions seriously worsen; non-radical populists, like their far-right counterpart, appeared to benefit from anti-immigrant sentiment, but could win greater support only under low economic deprivation. This paper made two important contributions. First, its key findings helped to refine the theory of populism by revealing the distinct social mechanisms leading to the growth of the three different types of electoral populism. To the best of the author’s knowledge, this paper is the first to test the three predominant accounts of the social origins of electoral populism vis-à-vis each other over all existing types of populist parties and to unravel a varieties of social changes each type of electoral populism is rooted in. From the existing accounts, this paper drew a twofold lesson. On one hand, the results of the analysis confirmed that the three key factors – economic deprivation, immigration, and political distrust – play the central role in shaping populist support. On the other hand, the results favored the hybrid explanation over the economic and non-economic explanations, validating the significance of the two pairwise interactions, each bundling economic deprivation with immigration and with political distrust, identified by the extant literature. It turned out, however, that these two pairwise interactions have differential effects on populist support: the 104 interaction between economic deprivation and immigration is positively associated with support for far-right populists, whereas one between economic deprivation and political distrust is predicted to foster support for far-lefts. Further, non-radical populists were found to benefit from a type of interaction opposed to one that drives far-right populism, appearing to win a greater share of the vote under growing public concern about immigration and declining economic deprivation. Combined together, therefore, the stimulating impacts of each distinct interaction on the three populist party families give a comprehensive view of changing support for populism, unveiling the heterogeneous social origins of the populist surge. Next, this paper offered a rigorous empirical test of social dynamics of electoral populism, building statistical models with the harmonized data of populist parties and country-level attributes that cover 27 European countries during a span of 17 years (2003-2019). This research design yielded two important analytical advantages. First, it provided comparative rigor, allowing to analyze populist parties across Europe classified according to the ideational approach. Not only did this make the cross-country analysis legitimate by resolving the definitional inconsistencies that had long plagued populism research, but it also permitted to take conceptual advantage the ideational approach claims over the alternative definitions of populism. Second, the present research design facilitated the longitudinal analysis of electoral populism, as the data covered the two decades during which most part of the populist surge had been witnessed and ensured the overtime comparability of populist parties. This was crucial to analyzing changing support for populist parties. 105 While this study analyzed populist support in Eurpoe, the results shed light on the recent US case as well. Studies of the 2016 presidential election show that a large share of citizens who voted for Donald Trump was the white working-lass group with strong hostility toward immigrants (Hooghe and Dassonneville 2018; Morgan and Lee 2018; Smith and Hanley 2018). There is no systemic longitudinal evidence for the joint effect of economic deprivation and immigration worries on Trump voting; but, cross-sectional research is in line with the positive interaction between economic deprivation and immigration worries (Goetz et al. 2019). The results for far-left populism from this study also echo electoral support for Bernie Sanders in 2016. Like in the case of Trump voters, we do not have evidence for the interaction effect: for far- left populism, political distrust and economic deprivation. And yet, prior study offers consistent cross-sectional evidence for the positive effects of both distrust and economic distress on Sanders voting (Altamura and Oliver 2022; Dyck, Pearson- Merkowitz, and Coates 2018). It is seems hard to test the results for non-radical populism with the US case, as we do not see any candidates or parties employing the broad or opportunistic populist appeal. In addition to advancing our understanding of social changes that drive the growth of populism, this paper raises two important questions for future research. One question is concerned with how the role of populist parties and leaders comes into play alongside economic deprivation, immigration, and political distrust. The focus of this paper was on the conditions under which more citizens turn to populist parties, and the varied interaction effects between the three key factors were found to constitute these conditions. This side of research on populism is often called the “demand-side” study, 106 because it is primarily interested in the conditions or circumstances that generate growing demands for populist parties (Golder 2016). The advantage of this research design is to detect and corroborate the effect of changes in social environments on electoral support for populist parties. Since those changes are often measurable and comparable across different countries, demand-side explanations typically seek to find general social conditions under which populist parties could make electoral success. Given that many populist parties, especially those in Europe, made unprecedented electoral success in recent years without any significant changes in party ideologies and programs, this analytical property gave rise to an explosion of demand-side studies that attributed the remarkable growth of populist parties to a set of key radical changes in the social world. Obviously, however, the salience of demand-side research by no means restricts the relevance of “supply-side” research, which focuses on strategic choices made by party organizations and political systems and contexts that shape the consequences of these choices (e.g., Art 2011; Garrido 2017; Lubbers and Scheepers 2002; Tavits 2012). As for research on any other types of political forces, research on populism reaps the full benefit from the interaction between the two sides of research. Especially, supply-side research on country-specific arrangements of political institutions and historical contingencies in which election campaigns of populist parties came into play would help to account for cases of populist parties that showed varied electoral performances under highly similar demand-side conditions. While much of prior supply-side research devoted overwhelming attention to far-right populists (Golder 2016; Mudde 2010), future research need to extend the focus to the 107 other two populist party families, as the demand-side sources of populist support turned out to vary across the three party families. A second question relates to the ideological manifestations and persuasive appeal of non-radical populism. Recall that populism is a variant of the thin-centered ideology, i.e., an ideology that merely offers the moralistic representation of “the people” against “the elite” and tends to be thickened out with other “host” ideologies that help to build the identity of the people and to come up with a practical solution to mobilize them (Mudde and Rovira Kaltwasser 2018). Prior study documented many cases of populism exercised in combination with the far-right or far-left ideologies, wherein the native group sharing a particular set of racial/ethnic and communal characteristics (far-right) or the economically underprivileged groups (far-left) arise as the people and populist parties seek to exclusively represent each of these groups (e.g., Bustikova 2014; Golder 2016; March 2011; Mudde 2010; Ramiro and Gomez 2017; Stavrakakis and Katsambekis 2014). The results from this paper provided evidence for the persuasive appeal of far-right populism and far-left populism, bolstering the central role of immigration and economic deprivation on growing electoral support for populist parties on each extreme end. Little is known, however, about the ideological manifestations of non-radical populism and its appeal to voters. Unlike far-right populists and far-left populists, non-radical populists do not stick to particular types of antagonistic ideologies, often appearing to woo broad classes of voters without partisan cues or to straddle both ideological ends. Consider four cases of non-radical populists that have been in office or received the largest parliamentary vote share at least once: Italy’s Five Star Movement (M5S), Croatia’s 108 Bridge of Independent Lists (Most), Bulgaria’s Citizens for European Development of Bulgaria (GERB), and Slovakia’s Direction–Social Democracy (Smer). M5S claims that it advocates the rights of ordinary citizens by allowing them to vote online for each proposed law. In this pursuit of direct democracy, governing institutions are expected to be a technocratic, non-partisan apparatus that merely facilitates public decision-making (Caruso 2017). Most’s populist appeal is aimed at voters disenchanted with the long reign of the country’s two major parties, and breaking the rigid two-party system is the primary objective of its proposed “reform” (Grbeša and Šalaj 2018). Like M5S, Most does not address any particular social group as its constituency. GERB and Smer, unlike M5S and Most, have been engaged in partisan politics and more attentive to policy agenda, but in ways to move across the partisan divide: GERB, generally considered a center-right party, has consistently supported European integration and the Keynesian-style economic management driven by large public spending; Smer, the hardline social democrat, overturned its longstanding pro- immigrant stance, embracing the ethnocentric ideology, prior to the recent 2016 election while retaining its extensive social welfare program (Deegan-Krause and Haughton 2009; Ganev 2018; Kazharski 2018). The stimulating impact of immigration on the vote share of non-radial populist parties certainly gives evidence that anti-immigrant sentiment fuels non-radical populism. While this points to an affinity between far-right populism and non-radical populism, the negative moderating role of economic deprivation seems to reveal the peculiarity of non-radical populism. As discussed earlier, competition with far-right populists might offer a plausible account of this peculiarity. And yet, much of support 109 for non-radical populism seems to remain unexplored. That is, we are still poorly informed of how non-radical populists appeal to voters: the absence of “host” ideologies and the ideological ambiguities among non-radical populists make it particularly difficult to navigate demand-side factors that foster support for them. While it turned out that immigration has positive electoral impact on non-radical populists, it is unclear why it does so given the fact that most non-radical populist parties do not adhere to nativism. Further research on non-radical populist mobilization need to delve into how the populist appeal can be made in conjunction with non-radical ideologies, investigating the ways in which populism becomes thickened out with non-radical ideologies. This will not only raise our understanding of the ideological manifestations and persuasive appeal of non-radical populism, but also facilitate broader research on the varieties of populist mobilization rooted in diverse types of ideological causes. 110 REFERENCES Acemoglu, Daron, Georgy Egorov, and Konstantin Sonin. 2013. “A Political Theory of Populism.” The Quarterly Journal of Economics 128(2):771–805. Algan, Yann, Sergei Guriev, Elias Papaioannou, and Evgenia Passari. 2017. “The European Trust Crisis and the Rise of Populism.” Brookings Papers on Economic Activity 2017(2):309–400. Altamura, Christopher, and Brenden Oliver. 2022. “Who Feels the Bern? An Analysis of Support for Bernie Sanders in the 2020 Democratic Primary.” American Politics Research 1532673X2211123. Art, David. 2011. Inside the Radical Right: The Development of Anti-Immigrant Parties in Western Europe. New York: Cambridge University Press. Arzheimer, Kai. 2009. “Contextual Factors and the Extreme Right Vote in Western Europe, 1980-2002.” American Journal of Political Science 53(2):259–75. Aslanidis, Paris, and Cristóbal Rovira Kaltwasser. 2016. “Dealing with Populists in Government: The SYRIZA-ANEL Coalition in Greece.” Democratization 23(6):1077–91. Berezin, Mabel. 2009. Illiberal Politics in Neoliberal Times: Culture, Security and Populism in the New Europe. Cambridge, UK; New York: Cambridge University Press. Berezin, Mabel. 2018. “Past Is Prologue.” Pp. 109–29 in Populism and the Crisis of Democracy. Volume 1: Concept and Theory, edited by G. Fitzi, J. Mackert, and B. S. Turner. London: Routledge. Berezin, Mabel. 2019. “Fascism and Populism: Are They Useful Categories for Comparative Sociological Analysis?” Annual Review of Sociology 45(1):345– 61. Berman, Sheri. 2019. “Populism Is a Symptom Rather than a Cause: Democratic Disconnect, the Decline of the Center-Left, and the Rise of Populism in Western Europe.” Polity 51(4):654–67. Blekesaune, Morten. 2007. “Economic Conditions and Public Attitudes to Welfare Policies.” European Sociological Review 23(3):393–403. Bonikowski, Bart, and Noam Gidron. 2016. “The Populist Style in American Politics: Presidential Campaign Discourse, 1952–1996.” Social Forces 94(4):1593– 1621. 111 Bornschier, Simon. 2012. “Why a Right-Wing Populist Party Emerged in France but Not in Germany: Cleavages and Actors in the Formation of a New Cultural Divide.” European Political Science Review 4(1):121–45. Broz, J. Lawrence, Jeffry Frieden, and Stephen Weymouth. 2019. “Populism in Place: The Economic Geography of the Globalization Backlash.” SSRN Electronic Journal. Brubaker, Rogers. 2017. “Why Populism?” Theory and Society 46(5):357–85. van der Brug, Wouter, and Anthony Mughan. 2007. “Charisma, Leader Effects and Support for Right-Wing Populist Parties.” Party Politics 13(1):29–51. Burgoon, Brian, Sam van Noort, Matthijs Rooduijn, and Geoffrey Underhill. 2019. “Positional Deprivation and Support for Radical Right and Radical Left Parties” edited by T. Beck. Economic Policy 34(97):49–93. Bustikova, Lenka. 2014. “Revenge of the Radical Right.” Comparative Political Studies 47(12):1738–65. Canovan, Margaret. 1999. “Trust the People! Populism and the Two Faces of Democracy.” Political Studies 47(1):2–16. Caruso, Loris. 2017. “Digital Capitalism and the End of Politics: The Case of the Italian Five Star Movement.” Politics & Society 45(4):585–609. Colantone, Italo, and Piero Stanig. 2018. “The Trade Origins of Economic Nationalism: Import Competition and Voting Behavior in Western Europe: Trade and Economic Nationalism.” American Journal of Political Science 62(4):936–53. Dancygier, Rafaela M. 2010. Immigration and Conflict in Europe. Cambridge: Cambridge University Press. Deegan-Krause, Kevin, and Tim Haughton. 2009. “Toward a More Useful Conceptualization of Populism: Types and Degrees of Populist Appeals in the Case of Slovakia.” Politics & Policy 37(4):821–41. Döring, Holger, and Philip Magnow. 2019. Parliaments and Governments Database (ParlGov): Information on Parties, Elections and Cabinets in Modern Democracies. Development Version. Dornbusch, Rudiger, and Sebastian Edwards, eds. 1991. The Macroeconomics of Populism in Latin America. Chicago: University of Chicago Press. 112 Doyle, David. 2011. “The Legitimacy of Political Institutions: Explaining Contemporary Populism in Latin America.” Comparative Political Studies 44(11):1447–73. Dustmann, Christian, Barry Eichengreen, Sebastian Otten, André Sapir, Guido Tabellini, and Gylfi Zoega. 2017. Europe’s Trust Deficit: Causes and Remedies. London: Centre for Economic Policy Research. Dyck, Joshua J., Shanna Pearson-Merkowitz, and Michael Coates. 2018. “Primary Distrust: Political Distrust and Support for the Insurgent Candidacies of Donald Trump and Bernie Sanders in the 2016 Primary.” PS: Political Science & Politics 51(02):351–57. Eichengreen, Barry. 2018. The Populist Temptation: Economic Grievance and Political Reaction in the Modern Era. New York: Oxford University Press. Enns, Peter K., and Ashley Jardina. 2021. “Complicating the Role of White Racial Attitudes and Anti-Immigrant Sentiment in the 2016 US Presidential Election.” Public Opinion Quarterly 85(2):539–70. Evans, Geoffrey. 2000. “The Continued Significance of Class Voting.” Annual Review of Political Science 3(1):401–17. Evans, Geoffrey, and Robert Andersen. 2006. “The Political Conditioning of Economic Perceptions.” The Journal of Politics 68(1):194–207. Ford, Robert Anthony, and Matthew J. Goodwin. 2014. Revolt on the Right: Explaining Support for the Radical Right in Britain. Abingdon, Oxon; New York, NY: Routledge. Foster, Chase, and Jeffry Frieden. 2019. “Compensation, Austerity, and Populism: Social Spending and Voting in 17 Western European Countries (Working Paper).” Freeden, Michael. 1996. Ideologies and Political Theory: A Conceptual Approach. Oxford: Clarendon Press. Freeden, Michael. 2017. “After the Brexit Referendum: Revisiting Populism as an Ideology.” Journal of Political Ideologies 22(1):1–11. Fukuyama, Francis. 2018. Identity: The Demand for Dignity and the Politics of Resentment. First edition. New York: Farrar, Straus and Giroux. Funke, Manuel, Moritz Schularick, and Christoph Trebesch. 2016. “Going to Extremes: Politics after Financial Crises, 1870–2014.” European Economic Review 88:227–60. 113 Ganev, Venelin I. 2018. “‘Soft Decisionism’ in Bulgaria.” Journal of Democracy 29(3):91–103. Garrido, Marco. 2017. “Why the Poor Support Populism: The Politics of Sincerity in Metro Manila.” American Journal of Sociology 123(3):647–85. Gest, Justin. 2016. The New Minority: White Working Class Politics in an Age of Immigration and Inequality. New York: Oxford University Press. Gidron, Noam, and Bart Bonikowski. 2013. “Varieties of Populism: Literature Review and Research Agenda.” Weatherhead Working Paper Series (13-0004). Gidron, Noam, and Peter A. Hall. 2017. “The Politics of Social Status: Economic and Cultural Roots of the Populist Right.” The British Journal of Sociology 68(S1):S57–84. Gidron, Noam, and Peter A. Hall. 2020. “Populism as a Problem of Social Integration.” Comparative Political Studies 53(7):1027–59. Gidron, Noam, and Jonathan J. B. Mijs. 2019. “Do Changes in Material Circumstances Drive Support for Populist Radical Parties? Panel Data Evidence from the Netherlands during the Great Recession, 2007–2015.” European Sociological Review 35(5):637–50. Goetz, Stephan J., Meri Davlasheridze, Yicheol Han, and David A. Fleming-Muñoz. 2019. “Explaining the 2016 Vote for President Trump across U.S. Counties.” Applied Economic Perspectives and Policy 41(4):703–22. Golder, Matt. 2016. “Far Right Parties in Europe.” Annual Review of Political Science 19(1):477–97. Goodhart, David. 2017. The Road to Somewhere: The Populist Revolt and the Future of Politics. London: Hurst & Company. Grbeša, Marijana, and Berto Šalaj. 2018. “Populism in Croatia: The Curious Case of The Bridge (Most).” Anali Hrvatskog Politološkog Društva: Časopis Za Politologiju 14(1):7–30. Gurr, Ted Robert. 1970. Why Men Rebel. Princeton, NJ: Princeton University Press. Hawkins, Kirk A. 2009. “Is Chávez Populist?: Measuring Populist Discourse in Comparative Perspective.” Comparative Political Studies 42(8):1040–67. Hernández, Enrique, and Hanspeter Kriesi. 2016. “The Electoral Consequences of the Financial and Economic Crisis in Europe.” European Journal of Political Research 55(2):203–24. 114 Hooghe, Marc, and Ruth Dassonneville. 2018. “Explaining the Trump Vote: The Effect of Racist Resentment and Anti-Immigrant Sentiments.” PS: Political Science & Politics 51(03):528–34. Huber, John D., and G. Bingham Powell. 1994. “Congruence between Citizens and Policymakers in Two Visions of Liberal Democracy.” World Politics 46(3):291–326. Hutter, Swen, and Rens Vliegenthart. 2018. “Who Responds to Protest? Protest Politics and Party Responsiveness in Western Europe.” Party Politics 24(4):358–69. Inglehart, Ronald, and Pippa Norris. 2017. “Trump and the Populist Authoritarian Parties: The Silent Revolution in Reverse.” Perspectives on Politics 15(2):443–54. Innes, Abby. 2002. “Party Competition in Postcommunist Europe: The Great Electoral Lottery.” Comparative Politics 35(1):85. Ivarsflaten, Elisabeth. 2005. “The Vulnerable Populist Right Parties: No Economic Realignment Fuelling Their Electoral Success.” European Journal of Political Research 44(3):465–92. Jackman, Robert W., and Karin Volpert. 1996. “Conditions Favouring Parties of the Extreme Right in Western Europe.” British Journal of Political Science 26(4):501–21. Jansen, Robert S. 2011. “Populist Mobilization: A New Theoretical Approach to Populism.” Sociological Theory 29(2):75–96. Jardina, Ashley. 2019. White Identity Politics. Cambridge, United Kingdom: New York, NY, USA : Cambridge University Press. Judis, John B. 2016. The Populist Explosion: How the Great Recession Transformed American and European Politics. New York: Columbia Global Reports. Katz, Jonathan N., and Gary King. 1999. “A Statistical Model for Multiparty Electoral Data.” American Political Science Review 93(1):15–32. Kaufmann, Eric P. 2019. Whiteshift: Populism, Immigration, and the Future of White Majorities. New York: Abrams Press. Kazharski, Aliaksei. 2018. “The End of ‘Central Europe’? The Rise of the Radical Right and the Contestation of Identities in Slovakia and the Visegrad Four.” Geopolitics 23(4):754–80. 115 Kazin, Michael. 1995. The Populist Persuasion: An American History. Ithaca, NY: Cornell University Press. Knigge, Pia. 1998. “The Ecological Correlates of Right-Wing Extremism in Western Europe.” European Journal of Political Research 34(2):249–79. Koopmans, Ruud, Paul Statham, Marco Giugni, and Florence Passy. 2005. Contested Citizenship: Immigration and Cultural Diversity in Europe. Minneapolis: University of Minnesota Press. Korpi, Walter. 2006. “Power Resources and Employer-Centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists.” World Politics 58(2):167–206. Kriesi, Hanspeter. 2014. “The Populist Challenge.” West European Politics 37(2):361–78. Kriesi, Hanspeter, and Takis S. Pappas, eds. 2016. European Populism in the Shadow of the Great Recession. Colchester, UK: ECPR Press. Laclau, Ernesto. 1977. Politics and Ideology in Marxist Theory: Capitalism— Fascism—Populism. Atlantic Highlands, NJ: Humanities Press. Laclau, Ernesto. 2005. On Populist Reason. London; New York: Verso. Ladner, Matthew, and Christopher Wlezien. 2007. “Partisan Preferences, Electoral Prospects, and Economic Expectations.” Comparative Political Studies 40(5):571–96. Levitsky, Steven, and Kenneth M. Roberts, eds. 2011. The Resurgence of the Latin American Left. Baltimore: Johns Hopkins University Press. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy: Economic Development and Political Legitimacy.” American Political Science Review 53(1):69–105. Lipset, Seymour Martin. 1960. Political man: the social bases of politics [computer program]. Version [1st ed.]. Lubbers, Marcel, Merove Gijsberts, and Peer Scheepers. 2002. “Extreme Right-Wing Voting in Western Europe.” European Journal of Political Research 41(3):345–78. Lubbers, Marcel, and Peer Scheepers. 2002. “French Front National Voting: A Micro and Macro Perspective.” Ethnic and Racial Studies 25(1):120–49. 116 Lynch, Julia. 2019. “Populism, Partisan Convergence, and Mobilization in Western Europe.” Polity 51(4):668–77. Lyrintzis, Christos. 1987. “The Power of Populism: The Greek Case.” European Journal of Political Research 15(6):667–86. Magni, Gabriele. 2017. “It’s the Emotions, Stupid! Anger about the Economic Crisis, Low Political Efficacy, and Support for Populist Parties.” Electoral Studies 50:91–102. March, Luke. 2011. Radical Left Parties in Europe. New York: Routledge. Mavrogordatos, George Th. 1997. “From Traditional Clientelism to Machine Politics: The Impact of PASOK Populism in Greece.” South European Society and Politics 2(3):1–26. McQuarrie, Michael. 2017. “The Revolt of the Rust Belt: Place and Politics in the Age of Anger.” The British Journal of Sociology 68(S1):S120–52. Moffitt, Benjamin. 2016. The Global Rise of Populism: Performance, Political Style, and Representation. Stanford, CA: Stanford University Press. Mols, Frank, and Jolanda Jetten. 2016. “Explaining the Appeal of Populist Right- Wing Parties in Times of Economic Prosperity: Economic Prosperity and Populist Right-Wing Parties.” Political Psychology 37(2):275–92. Morgan, Stephen L. 2018. “Status Threat, Material Interests, and the 2016 Presidential Vote.” Socius: Sociological Research for a Dynamic World 4. Morgan, Stephen, and Jiwon Lee. 2018. “Trump Voters and the White Working Class.” Sociological Science 5:234–45. Mudde, Cas. 2004. “The Populist Zeitgeist.” Government and Opposition 39(4):541– 63. Mudde, Cas. 2007. Populist Radical Right Parties in Europe. Cambridge, UK; New York: Cambridge University Press. Mudde, Cas. 2010. “The Populist Radical Right: A Pathological Normalcy.” West European Politics 33(6):1167–86. Mudde, Cas. 2016. On Extremism and Democracy in Europe. London; New York: Routledge, Taylor&Francis Group. Mudde, Cas, and Cristóbal Rovira Kaltwasser. 2013. “Exclusionary vs. Inclusionary Populism: Comparing Contemporary Europe and Latin America.” Government and Opposition 48(2):147–74. 117 Mudde, Cas, and Cristóbal Rovira Kaltwasser. 2017. Populism: A Very Short Introduction. New York, NY: Oxford University Press. Mudde, Cas, and Cristóbal Rovira Kaltwasser. 2018. “Studying Populism in Comparative Perspective: Reflections on the Contemporary and Future Research Agenda.” Comparative Political Studies 51(13):1667–93. Müller, Jan-Werner. 2016. What Is Populism? Philadelphia: University of Pennsylvania Press. Muro, Mark, and Sifan Liu. 2016. Another Clinton-Trump Divide: High-Output Amer- Ica vs Low-Output America. Brookings. Norris, Pippa. 2005. Radical Right: Voters and Parties in the Electoral Market. New York, NY: Cambridge University Press. Norris, Pippa, and Ronald Inglehart. 2019. Cultural Backlash: Trump, Brexit, and Authoritarian Populism. New York: Cambridge University Press. Noury, Abdul, and Gerard Roland. 2020. “Identity Politics and Populism in Europe.” Annual Review of Political Science 23(1):421–39. Oliver, J. Eric, and Wendy M. Rahn. 2016. “Rise of the Trumpenvolk: Populism in the 2016 Election” edited by L. M. Bartels. The ANNALS of the American Academy of Political and Social Science 667(1):189–206. Ostiguy, Pierre. 2017. “Populism: A Socio-Cultural Approach.” Pp. 73–97 in The Oxford Handbook of Populism, edited by C. Rovira Kaltwasser, P. A. Taggart, P. O. Espejo, and P. Ostiguy. Oxford, UK; New York: Oxford University Press. Papke, Leslie E., and Jeffrey M. Wooldridge. 1996. “Econometric Methods for Fractional Response Variables with an Application to 401(k) Plan Participation Rates.” Journal of Applied Econometrics 11(6):619–32. Pappas, Takis S. 2014. “Populist Democracies: Post-Authoritarian Greece and Post- Communist Hungary.” Government and Opposition 49(1):1–23. Powell, G. Bingham. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven: Yale University Press. Quillian, Lincoln. 1995. “Prejudice as a Response to Perceived Group Threat: Population Composition and Anti-Immigrant and Racial Prejudice in Europe.” American Sociological Review 60(4):586. 118 Ramiro, Luis. 2016. “Support for Radical Left Parties in Western Europe: Social Background, Ideology and Political Orientations.” European Political Science Review 8(1):1–23. Ramiro, Luis, and Raul Gomez. 2017. “Radical-Left Populism during the Great Recession: Podemos and Its Competition with the Established Radical Left.” Political Studies 65(1_suppl):108–26. Roberts, Kenneth M. 2006. “Populism, Political Conflict, and Grass-Roots Organization in Latin America.” Comparative Politics 38(2):127. Roberts, Kenneth M. 2015. “Populism, Political Mobilizations, and Crises of Political Representation.” Pp. 140–58 in The Promise and Perils of Populism, Global Perspectives, edited by C. de la Torre. University Press of Kentucky. Roberts, Kenneth M. 2019. “Bipolar Disorders: Varieties of Capitalism and Populist Out-Flanking on the Left and Right.” Polity 51(4):641–53. Rodrik, Dani. 2018. “Populism and the Economics of Globalization.” Journal of International Business Policy 1(1–2):12–33. Rooduijn, Matthijs, and Tjitske Akkerman. 2017. “Flank Attacks: Populism and Left- Right Radicalism in Western Europe.” Party Politics 23(3):193–204. Rooduijn, Matthijs, and Brian Burgoon. 2018. “The Paradox of Well-Being: Do Unfavorable Socioeconomic and Sociocultural Contexts Deepen or Dampen Radical Left and Right Voting Among the Less Well-Off?” Comparative Political Studies 51(13):1720–53. Rooduijn, Matthijs, Stjin Van Kessel, Caterina Froio, Andrea Pirro, Sarah De Lange, Daphne Halikiopoulou, Paul Lewis, Cas Mudde, and Paul A. Taggart. 2019. The PopuList: An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. Rovira Kaltwasser, Cristóbal, Paul A. Taggart, Paulina Ochoa Espejo, and Pierre Ostiguy. 2017. “Populism: An Overview of the Concept and the State of the Art.” Pp. 1–24 in The Oxford handbook of populism, edited by C. Rovira Kaltwasser, P. A. Taggart, P. O. Espejo, and P. Ostiguy. Oxford, UK; New York: Oxford University Press. Ruth-Lovell, Saskia P., Anna Lührmann, and Sandra Grahn. 2019. “Democracy and Populism: Testing a Contentious Relationship.” SSRN Electronic Journal. Rydgren, Jens. 2007. “The Sociology of the Radical Right.” Annual Review of Sociology 33(1):241–62. 119 Schumacher, Gijs, and Matthijs Rooduijn. 2013. “Sympathy for the ‘Devil’? Voting for Populists in the 2006 and 2010 Dutch General Elections.” Electoral Studies 32(1):124–33. Semyonov, Moshe, Rebeca Raijman, and Anastasia Gorodzeisky. 2006. “The Rise of Anti-Foreigner Sentiment in European Societies, 1988-2000.” American Sociological Review 71(3):426–49. Smith, David Norman, and Eric Hanley. 2018. “The Anger Games: Who Voted for Donald Trump in the 2016 Election, and Why?” Critical Sociology 44(2):195– 212. Solt, Frederick. 2008. “Economic Inequality and Democratic Political Engagement.” American Journal of Political Science 52(1):48–60. Stavrakakis, Yannis, and Giorgos Katsambekis. 2014. “Left-Wing Populism in the European Periphery: The Case of SYRIZA.” Journal of Political Ideologies 19(2):119–42. Swank, Duane, and Hans-Georg Betz. 2003. “Globalization, the Welfare State and Right-Wing Populism in Western Europe.” Socio-Economic Review 1(2):215– 45. Taggart, Paul. 1995. “New Populist Parties in Western Europe.” West European Politics 18(1):34–51. Taggart, Paul. 1998. “A Touchstone of Dissent: Euroscepticism in Contemporary Western European Party Systems.” European Journal of Political Research 33(3):363–88. Taggart, Paul A. 2000. Populism. Buckingham, UK; Philadelphia: Open University Press. Taggart, Paul, and Aleks Szczerbiak. 2004. “Contemporary Euroscepticism in the Party Systems of the European Union Candidate States of Central and Eastern Europe.” European Journal of Political Research 43(1):1–27. Tavits, Margit. 2012. “Organizing for Success: Party Organizational Strength and Electoral Performance in Postcommunist Europe.” The Journal of Politics 74(1):83–97. Tomz, Michael, Joshua A. Tucker, and Jason Wittenberg. 2002. “An Easy and Accurate Regression Model for Multiparty Electoral Data.” Political Analysis 10(1):66–83. Urbinati, Nadia. 1998. “Democracy and Populism.” Constellations 5(1):110–24. 120 Van Der Brug, Wouter, Meindert Fennema, and Jean Tillie. 2000. “Anti-Immigrant Parties in Europe: Ideological or Protest Vote?” European Journal of Political Research 37(1):77–102. Vlandas, Tim, and Daphne Halikiopoulou. 2019. “Does Unemployment Matter? Economic Insecurity, Labour Market Policies and the Far-Right Vote in Europe.” European Political Science 18(3):421–38. Werts, Han, Peer Scheepers, and Marcel Lubbers. 2013. “Euro-Scepticism and Radical Right-Wing Voting in Europe, 2002–2008: Social Cleavages, Socio-Political Attitudes and Contextual Characteristics Determining Voting for the Radical Right.” European Union Politics 14(2):183–205. Weyland, Kurt. 2001. “Clarifying a Contested Concept: Populism in the Study of Latin American Politics.” Comparative Politics 34(1):1–22. Zmerli, Sonja, and Juan Carlos Castillo. 2015. “Income Inequality, Distributive Fairness and Political Trust in Latin America.” Social Science Research 52:179–92. Zulianello, Mattia. 2020. “Varieties of Populist Parties and Party Systems in Europe: From State-of-the-Art to the Application of a Novel Classification Scheme to 66 Parties in 33 Countries.” Government and Opposition 55(2):327–47. 121 APPENDIX Mean SD Min Max 𝑵 Source Dependent variables, measured at election year 𝒕† Far-right populist vote share (%) 9.8 13.8 0 69.4 114 PopuList; ParlGov Far-left populist vote share (%) 3.1 7.0 0 35.9 114 PopuList; ParlGov Non-radical populist vote share (%) 6.8 13.1 0 54.1 114 PopuList; ParlGov Independent variable, measured at 𝒕 − 𝟏 Unemployment rate (%) 9.1 4.7 3.7 26.5 114 Eurostat Poverty gap (% of poverty line) 22.2 5.2 13.8 38.2 108‡ Eurostat Gini (%) 29.9 4.0 23.2 38.8 108‡ Eurostat GDP growth 1.8 4.1 -12.9 23.9 114 AMECO (real, %) Interest rate (%) -.02 3.2 -7.4 25.2 114 AMECO FDI (% of GDP) 6.2 12.1 -12.0 80.7 114 AMECO Social spending (% of GDP) 14.4 2.8 8.4 19.8 114 AMECO Political distrust (factor score) .1 1.0 -2.0 1.7 114 Eurobarometer‡† Public concern about immigration 11.3 10.3 .3 50.4 114 Eurobarometer‡† (% of r) Euroscepticism (% of r) 14.3 7.4 4.1 34.1 114 Eurobarometer‡† Table A1. Descriptive statistics and data sources †The total vote share won by each populist family in a parliamentary election in year 𝑡. In cases where election was held more than once within a year (e.g., Greece in 2012 and 2015), the yearly averages of the vote shares were taken. ‡Missing values for Belgium (2002), Germany (2004), UK (2004, 2018), Spain (2003), Poland (2004) ‡†Standard Eurobarometer 47.0; 47.1; 47.2, 48.0, 49, 50.0; 50.1; 51.0; 52.0; 53; 54.1; 55.1; 56.2; 57.1; 57.2; 58.1; 59.1; 60.1; 61; 62.0; 63.4; 64.2; 65.2; 66.1; 67.2; 68.1; 69.2; 70.1; 71.1; 71.3; 72.4; 73.4; 74.2; 75.3; 76.3; 77.3; 77.4; 78.1; 79.3; 79.5; 80.1; 81.2; 81.4; 82.3; 82.4; 83.1; 83.3; 84.1; 84.3; 85.2; 86.1; 86.2; 87.3; 88.1; 88.3; 89.1; 90.1; 90.3. 122 Country Far-right Far-left Non-radical Austria FPÖ Martin BZÖ TS Belgium VB LDD FN Pp Bulgaria NSFB GERB Volya BBZ/BBT Ataka BBB RZS NDSV IMRO Croatia HDSSB HL-SR/CL-LP MOST Zivi Zid HGS Cyprus SYM/SYPOL Czech Rep. SPD ANO SPR-RSC VV S-JB Dawn Denmark DF NB FrP Estonia EKRE EKo Finland Ps SIN France DLR|DLF FI FN/RN Germany AfD Linke Greece EL MR25 ANEL LAOS SYRIZA POLAN DIKKI SYN Hungary FIDEZS Jobbik Fi+KDNP MIÉP MH Ireland SF Italy Fdl FI/PdL LN M5S LAM PdL LV Latvia KPV LV ZRP Lithuania JL LLaS DP LCP TT DK TPP Luxembourg ADR 123 Netherlands FvD SP LN PVV LPF SP CD Poland Kukiz’15 SRP PiS LPR X Portugal CH Romania PRM PPDD PRU PUNR Slovakia SNS OLaNO SR Smer PSNS ANO SOP ZRS Slovenia SDS L LMS SNS ZdLe/L Spain Vox ECP Podemos EM Sweden SD NyD UK UKIP SF R Table A2. Populist parties (acronyms) from each party family in 27 European countries Sources. PopuList data Note: All parties competed in election at least once during 2003-2019. 124 DV: Far-right populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Immigrants!!! .0009** .001** -.03* .001** .001** .001 .001** .001** -.03** (in thousands) (.0003) (.0003) (.01) (.0003) (.0003) (.004) (.0003) (.0004) (.007) Unemp!!! .0002* ×Immigrants!!! (.00009) Pov Gap!!! <.0001 ×Immigrants!!! (.0001) Gini!!! .001** ×Immigrants!!! (.0002) 𝑁 59 59 59 59 59 59 59 59 59 DV: Far-left populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Immigrants!!! .001 .0009 .0003 .003* .003* .002 .001 .001 .0006 (in thousands) (.0006) (.0006) (.001) (.001) (.001) (.001) (.0007) (.0007) (.01) Unemp!!! .0002 ×Immigrants!!! (.0001) Pov Gap!!! <.0001 ×Immigrants!!! (.00008) Gini!!! .0001 ×Immigrants!!! (.0003) 𝑁 30 30 30 30 30 30 30 30 30 DV: Non-radical populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Immigrants!!! -.001 -.001 -.002 -.0007 -.0008 .01 -.0008 -.0008 -.02 (in thousands) (.001) (.002) (.002) (.001) (.001) (.01) (.001) (.001) (.04) Unemp!!! -.001 ×Immigrants!!! (.0006) Pov Gap!!! -.0006 ×Immigrants!!! (.0005) Gini!!! .0002 ×Immigrants!!! (.001) 𝑁 32 32 32 32 32 32 32 32 32 Table A3. Models estimated with the measure of immigration inflows: selected outcomes Note: Each model controls for the same variables as the corresponding model in Table 1, 2, and 3. The measure of immigration inflows was drawn from Eurostat data of long-term immigrants. *p<.05; **p<.01 (two-tailed significance levels). 125 DV: Far-right populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Immigration concern!!! .01** .01** .005 .01** .01** .01 .01** .01** .005 (% of r) (.003) (.003) (.006) (.003) (.003) (.01) (.003) (.003) (.02) Unemp!!! .001 ×Immigration!!! (.0008) Pov Gap!!! .001 ×Immigration!!! (.0008) Gini!!! .0003 ×Immigration!!! (.0009) DV: Far-left populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Unemployment!!! .02 -.09* .02 .06 .19 .06 -.03 .05 -.03 (%) (.02) (.03) (.02) (.04) (.37) (.04) (.06) (.06) (.06) Poverty Gap!!! (% of poverty line) Gini!!! (%) Political Distrust!!! .91** .44* .84** .68** -.19 .62 .94** - .90** (% of r) (.29) (.20) (.23) (.23) (.37) (.31) (.16) 3.27** (.21) (.57) Unemp!!!×Distrust!!! .07** (.01) Pov Gap!!!×Distrust!!! .04** (.01) Gini!!!×Distrust!!! .14** (.02) DV: Non-radical populist vote share M1 M2 M3 M4 M5 M6 M7 M8 M9 Immigration concern!!! .006 .005 .03* .001 .002 .06 .01 .01 .08* (% of r) (.01) (.01) (.01) (.01) (.01) (.03) (.01) (.01) (.04) Unemp!!! -.004* ×Immigration!!! (.001) Pov Gap!!! -.002 ×Immigration!!! (.001) Gini!!! -.002 ×Immigration!!! (.001) 𝑁 114 114 114 108 108 108 108 108 108 Table A4. Models estimated with the entire sample: selected outcomes Note: Each model controls for the same variables as the corresponding model in Table 1, 2, and 3. *p<.05; **p<.01 (two-tailed significance levels). 126 Chapter 3. Toward A New Theory of Social Integration: Durkheim’s Theory of Solidarity Revisited Abstract The economic and political crises that materialized in the global financial crisis and the rise of populism and illiberal politics over the last few decades accompanied ever- growing conflicts and antagonism between social groups. While this inspired an explosion of literature on social conflicts and political polarization, however, research on how such conflicts and polarization can be managed has seldom been developed since Émile Durkheim (1858-1917) first proposed his theory of social integration. In this chapter, I first provide a critical review of Durkheim’s theory of social integration and then propose a set of directions to renew his theory to tackle the novel challenges raised in our contemporary society. 127 Where the economic and political crises materialized in the global financial crisis and the rise of populism and illiberal politics over the last few decades, we have witnessed ever-growing conflicts and antagonism between diverse social groups. Not surprisingly, this inspired an explosion of literature on social conflicts and political polarization. Interestingly, however, research on how such conflicts and polarization can be managed has seldom been developed since Émile Durkheim (1858-1917) first proposed his theory of social integration. Durkheim’s theory of solidarity provides a firm ground on which to develop a new approach to social integration, as it addresses the key characteristics of the modern industrial society – capitalism and division of labor – that remain relevant for our contemporary society. This chapter assesses the validity of Durkheim’s account against the backdrop of the major social transformations in the late twentieth century and early twenty-first century and explores the directions for its renewal to elucidate the new social problems that have undermine social cohesion. This work would help to shed light on the continuities in the modern industrial society for which Durkheim’s theory still has explanatory leverage, while revealing the discontinuities between the early capitalist society and the late capitalist society, which calls for building a new theory of solidarity on the ground of Durkheim’s legacy. DURKHEIM’S THEORY OF SOCIAL INTEGRATION As widely known, Durkheim developed his theory of solidarity to make sense of how a society maintains itself while undergoing the fundamental changes (and challenges) that stem from the transition from the primitive society to the modern industrial 128 society. Key to understanding the maintenance of a society is the sets of conditions that characterize the two different types of society, as those conditions would determine the way in which a society retains its status quo. Mechanical Solidarity In The Division of Labor in Society (1893, DL), Durkheim characterizes the primitive society as one in which citizens and social groups are highly similar with each other with respect to the fundamental social characteristics such as roles, occupations, and family types. As such, “society insists upon its citizens displaying all these basic resemblances” (DL:61). Such fundamental similarities in turn determine the form of social cohesion that unites citizens and social groups in the primitive society. Durkheim’s famous notion of mechanical solidarity stands for this form of social cohesion. In the primitive society where citizens remain less differentiated from each other, Durkheim argues, each individual’s consciousness that pursues her personal goals and preferences and the consciousness that serves the collective causes do not contradict each other. Instead, the “two consciousnesses are linked to each other, since in the end they constitute only one entity, for both have one and the same organic basis. Thus they are solidly joined together.” (DL: 61) This is why Durkheim characterizes the form of social cohesion prevalent in the primitive society as mechanical: since the pursuit of collective goals and causes is not distinct from the pursuit of individual goals and causes, it turns out that the former is drawn from the latter, or vice versa. 129 While Durkheim’s account of mechanical solidarity in DL is primarily concerned with the resemblances between individuals, addressing social cohesion in the primitive society as one driven by the structural identity of individuals, he also formulates a set of concrete ideas and concepts to account for how the individual-level resemblances in the end develop into the collective cohesion. In The Elementary Forms of Religious Life (1912, EFR), Durkheim draws on what he calls conceptual categories and collectivity to explain how the resemblances between individuals lead to social cohesion. According to Durkheim, conceptual categories “are essentially collective representations” and “[t]hey depend upon the way in which the collectivity is organized” (EFR:15). What is then collectivity? Durkheim defines it as “[a]ll consensus among minds” based upon which individuals all share “a homogeneous conception of time, space, cause, number, and so on.” (EFR:16) That is, collectivity is a whole set of commonly agreed ideas of the society, such as how the society works and how it is (de-)stabilized, and it is socially formed and forged by collective behavior. In other words, it is a basic consensus that provides individuals with a shared understanding of their society as it is. Collectivity alone, however, is not enough for individuals to fully account for the society they live in. To do so, individuals need specific tools that allow them to comprehend the complexity of the society. That is, individuals need to conceptualize each part of the society in different manners – for instance, the way in which time is measured is different from the way in which space divided; while some social events can be properly analyzed through a causal mechanism, others cannot. Hence, in order for such differentiated conceptualizations to work out, various categories are required 130 to come into being. These conceptual categories allow individuals to understand the society more concretely in accordance with the collectivity. Durkheim gives several examples of conceptual categories: categories of “time, space, number, cause, substance, personality” (EFR:8). He argues that all these existing categories must be understood as “social beings, products of collective thought” (EFR:9) that help to conceive the society. For instance, the category of time (“days, weeks, months, years, etc.”) is a collective invention that represents the “collective thought” and provides “all men of the same civilization” with a means to communicate and cooperate with each other based upon the same mode of recognition of time without which they could hardly behave collectively (EFR: 10). The same logic applies to the other categories. That is, all the existing categories, as “products of collective thought” (EFR: 10), are conceptual tools that enable individuals to conceive the society in mutually shared ways, which in turn paves the way for social integration. Organic Solidarity Durkheim recognized that the existing forms of solidarity and social integration do not last permanently but change as the society undergoes fundamental transformations. Here comes his famous notion of organic solidarity. As discussed above, mechanical solidarity unites individuals and social groups by a common set of values and perceived realities. This commonality underpins social interactions between individuals and in turn forms identity of individuals in the primitive society. Solidarity driven by identity, however, diminishes, as the formations of the society radically 131 change and stop being characterized by interactions between individuals with homogenous interests, values, and views. Unlike the primitive society, the modern, industrialized society consists of individuals with heterogeneous identities. Nor do individuals share a common set of interests, values, and views – that is, conceptual categories and collectivity that have established the basis of social cohesion in the primitive society seize to work out. The transition from the primitive society to the modern industrialized society entails a profound change in the social lives of individuals and the ways in which they interact with each other. Central to the transition is an increase in socio-economic complexities. Durkheim’s emphasis on the division of labor addresses increasing complexities in the modern society, capturing a growing degree of occupational specialization in industrialized workplaces. The division of labor in the workplace, Durkheim elaborates, has broad consequences for individuals, altering personality and social interactions: “on the one hand each one of us depends intimately upon society the more labour is divided up, and one the other, the activity of each one of us is correspondingly more specialised, the more personal it is.” (DL:85) The division of labor in the modern industrialized society, therefore, accompanies the two (conflicting) major changes in the social world: increasing individualism and interdependence between citizens. Organic solidarity stems from the interdependence between citizens with greater individualistic sentiment. The idea that individualism goes in tandem with interdependence might sound counterintuitive. With the expansion of the division of labor, however, the harmonious dynamics of individualism and interdependence 132 become a functional necessity, as individuals with diverse occupational specialties are motivated to complement each other. To account for solidarity driven by the differences between individuals, Durkheim provides a famous analogy: “(t)his solidarity resembles that observed in the higher animals. In fact each organ has its own special characteristics and autonomy, yet the greater the unity of the organism, the more marked the individualisation of the parts. Using this analogy, we propose to call “organic” the solidarity that is due to the division of labour.” (DL:85) Lessons from Durkheim’s Theory of Solidarity While there might be many different ways to construe Durkheim’s theory of solidarity, two general lessons seem to be drawn. First, one may understand Durkheim’s two types of solidarity as Weberian ideal types, i.e., conceptual devises that help to approximate a particular type of social phenomena regularly observed in many different social settings (Swedberg 2005). None of the two types of solidarity would perfectly represent the progression of social cohesion in any given society. And yet, each type of solidarity would give a conceptual framework that broadly captures the process of social integration commonly witnessed in the primitive society or in the modern industrial society (needless to say, the two types of society can also be best understood as ideal types). Next, as the transition from the primitive society to the modern industrialized society stems from the rise of capitalism (both the development of mass production and the transformation of social relations between occupational groups) (Giddens 1971), it is organic solidarity that should help to account for social cohesion in the 133 contemporary social world . If organic solidarity, increasing social integration driven by differences, turns out to fit with the process of social integration in the contemporary societies, it may hold empirical validity; if not, however, we need to develop an alternative theory of solidarity that better accounts for the social dynamics underlying the process of social integration in the contemporary societies. PROBLEMS: WEAKENING INTERDEPENDENCE AND GROWING POLARIZATION At the heart of organic solidarity is the reinforcing dynamics of difference and interdependence between individuals, functionally necessitated by the division of labor. As often addressed by critiques (Giddens 1971, 1986), however, growing individualism in the modern society may end up weakening the role of the division of labor, so that the overwhelming differences between individuals come to undermine the mutual dependence of individuals. In effect, the major changes that have occurred in the post-war social world provide good reason to suspect that organic solidarity no longer represents the principles of social integration in the contemporary society. Here I shall discuss two interrelated social processes that countervails the reinforcing dynamics of difference and interdependence and therefore weakens social integration. One is weakening interdependence between social groups, and the other is growing polarization. 134 Weakening Interdependence One important factor that undermines the logic of organic solidarity is weakening interdependence between citizens in the modern world. Somewhat paradoxically, this stemmed primarily from increasing economic dependence. That is to say, what characterizes the world economy in the late twentieth century and onwards is economic globalization: especially, the openness of global economy spurred by increasing cross-border flows of commodities and human capital (Gereffi 1996, 2001; Held et al. 1999; Sassen 2000, 2003). The massive weave of globalization resulted in ever-growing dependence between countries, as transnational movements of capital facilitated the economic specialization of each country. In the age of global economy, therefore, prosperity and growth originate in mutual economic dependence between countries: no country could thrive alone. Where the rise of global economy reinforced mutual economic dependence and reciprocity between countries in the international economy, however, it also seriously weakened mutual economic dependence and reciprocity in the domestic economy. Such conflicting movements in the two different domains of economy took place due to the unintended consequences of globalization for the domestic economy (Sassen 1998; Stiglitz 2002). Those consequences were most salient in advanced economies (especially, those in North America and Western Europe) where a large share of cross- border capital flows resulted from an influx of immigrants. Although the actual economic effect of increasing immigration on the receiving countries has yet to be validated, it was the perceived threat from immigrant workers that has formed a consensus on the negative impact of globalization on the advanced economies (Norris 135 2019; Rodrik 2018). Among others, blue-collar workers in traditional manufacturing industries appeared to be hostile to the inflow of immigrant workers, as those jobs turned out to significantly decrease due to automation as well as the availability of cheap labor from developing economies (Milner 2021; Morgan 2018). As discussed in Chapter 2, anti-globalization and anti-immigrant sentiment turned many of blue-collar workers (largely from the ethnic/racial majority group) into “the losers of globalization”, who are more likely than the other social groups to support nativism, often at the expense of democratic principles. Recent studies show that those perceived victims of globalization have channeled their nativist and anti-globalization sentiment into grass-roots campaigns and institutional politics (Jardina 2019; Norris 2019; Swank and Betz 2003). This movement led to an unprecedented success of rightwing populist leaders and parties across the world: as described in detail in Chapter 2, far-right populist parties with the nativist and authoritarian ideology made remarkable electoral success in many European countries and the US over the last decades. While the proponents of far-right populism often described themselves as victims of globalization or elite politics, their political movements also generated a large group of victims: that is, racial/ethnic minorities and immigrants they have proclaimed as aliens or enemies. The hostility toward globalization and immigrants raises a serious challenge to the logic of organic solidarity, which raises social cohesion by the reinforcing dynamics of difference and interdependence between citizens. Ad discussed earlier, Durkheim stressed that increasing differences between individuals could strengthen social cohesion in the industrialized society where those differences spur 136 interdependence between individuals. At the heart of the spurring role of difference in strengthening interdependence is the division of labor, which functionally necessitates the complementary relationship between individuals. The unintended consequences of globalization for the domestic economy, however, casts doubt on the role of the division of labor in turning difference into interdependence between individuals. It is interesting to see that, like the development of mass production systems in the earlier capitalist economy, globalization promotes the division of labor, albeit in more complex forms: the increasing openness of global economy and ever-growing cross-border flows of capital lead to further specialization across the world, motivating the global north to specialize in high-tech and service industries and the global south to specialize in traditional manufacturing industries (Cardoso 1972, 1977; Wallerstein 1976). As such, the global division of labor across countries resulted in the decline of manufacturing sector in the global north, generating the victims of such industrial changes (Milner 2021; Swank and Betz 2003). As rage and grievances among those victims inspired by globalization turn into increasing hostility toward social groups that have shown growing presence in the era of globalization, the global division of labor seems to wind up weakening interdependence between citizens. If the global division of labor exacerbates conflicts rather than raises solidarity, therefore, this calls for a fundamental revision of the theory of solidarity. Central to organic solidarity, as Durkheim states, is the harmonious organization of differentiated individuals orchestrated by the division of labor. This form of organization transcends a mere coexistence, as it achieves an “organic unity” like one 137 witnessed in the organs of the body of the higher animals. Increasing conflicts and antagonism structured by the global division of labor, therefore, not only challenges the empirical validity of organic solidarity in the contemporary world but also fundamentally erodes its logic addressing the division of labor as a causal mechanism linking difference up to interdependence between individuals. Growing Polarization The other challenge to the logic of organic solidarity stems from growing political polarization in the contemporary society. As many scholars of political culture and public opinion have stressed, the past several decades witnessed ever-growing polarization of public opinion over a wide range of issues encompassing cultural, ethical, economic, and political problems. Polarization refers to “the simultaneous presence of opposing or conflicting principles, tendencies, or points of view” (Fiorina and Abrams 2008:566). In many cases, it designates a bimodal distribution of public opinion, displaying a state in which many individuals have extremely opposing views over a given set of issues. Obviously, polarization raises a serious challenge to the logic of organic solidarity. Above all, it suggests that differences between individuals need to be considered seriously. As a sizable volume of literature on the “culture war” in America (Baldassarri and Bearman 2007; Baldassarri and Gelman 2008; Fiorina, Abrams, and Pope 2005) has long investigated, exceedingly opposing attitudes among American citizens toward a set of moral and cultural issues (e.g., abortion, drug use, gun regulation, and law enforcement) addressed nearly irreconcilable conflicts in 138 various spheres of the social world, such as families, schools, colleges, towns, and firms (Bolce and De Maio 1999; Jacoby 2014). Not surprisingly, those attitudinal differences often lead to serious conflicts that accompany substantial social and economic costs. As the political scientist Adam Pzreworski pointed out, growing attitudinal polarization among the public makes it very difficult to find out a solution for the issues at stake: “(w)hen conflicts are intense and a society is highly polarized, finding policies acceptable to all major political forces is difficult and may be impossible. There are limits to what even the best-intentioned and competent governments can do” (Przeworski 2019:9). Where Durkheim stressed the “organic unity” of individuals with their “own characteristics”, he held that individual differences in the modern society would end up constituting an ordered whole as those differences addresses the need for harmonious coexistence between differentiated individuals. Like in the case of decreasing interdependence, however, growing polarization suggests that Durkheim’s argument on the causal mechanism linking difference to interdependence need to be reconsidered and revised. In effect, it seems unfeasible for individuals with highly conflicting or opposing attitudes toward major social issues to reach any equilibrium in the short term: as well-documented by the existing literature on polarization and the culture war, political polarization has been deepened over the last several decades and largely shaped by the long-lasting historical conflicts (Fiorina et al. 2005). Attitudinal differences between individuals characterized by growing polarization, therefore, would weaken social bonds and exacerbate social disintegration. Like in the case of globalization, recent research on populism has 139 reported that polarization tends to promote populist politics and is highly associated with voting for populist leaders and parties (Mudde and Rovira Kaltwasser 2018). This might not be surprising, as the populist ideology, as discussed in Chapter 2, works by the antagonistic division of social groups, and growing polarization over major social issues would further strengthen the division by providing sources of antagonism and conflicts. This would in turn reinforce the populist voter base, contributing to electoral success of populists, who in many cases propose policy alternatives geared toward only a faction of voters. In addition to the difficulty of reconciling conflicting preferences, growing polarization provides a social psychological reason to believe that solidarity would diminish instead of becoming stronger in the contemporary social world. This reasoning is drawn from mutual hatred among partisans inspired by ideological polarization. Often referred to as “affective polarization” (Hutter and Vliegenthart 2018; Iyengar, Sood, and Lelkes 2012; Iyengar and Westwood 2015; Levendusky and Malhotra 2016), partisan animosity has been addressed as one of the major characteristics of contemporary politics over the last few decades. While much of prior study on political polarization focused on the extent to which partisans disagree on particular issues, a growing volume of recent studies has shifted attention to emotional aspects of polarization. Studies in the latter group highlight the belief among partisans that “the other party’s members are hypocritical, selfish, and closed-minded, and they are unwilling to socialize across party lines, or even to partner with opponents in a variety of other activities” (Iyengar et al. 2019:7.2). 140 Where ideological differences between partisans translate into strong mutual hatred, it becomes harder for partisans to build a basis of strong solidarity. Instead of pursuing the peaceful coexistence grounded in mutual respect, partisans would seek to eradicate the other party’s influences on policies and the livelihoods of citizens. This is what is often witnessed when the ruling party and opposition parties refuse to put collective effort into reaching a compromise. Further, as the political scientists Steven Levitsky and Daniel Ziblatt pointed out in their influential book, How Democracies Die (2018), strong affective polarization and the resulting erosion of mutual tolerance among partisans would pave the way for a crisis of democracy by incentivizing parties to dismiss the other parties in pursuit of their own interests at the expense of democratic rules and norms. I will discuss this issue in greater detail below. TOWARD AN ALTERNATIVE THEORY OF SOLIDARITY The theory of solidarity, therefore, needs to be revisited to address the problems of decreasing interdependence and growing polarization. As discussed above, Durkheim’s logic of organic solidarity that marries difference with interdependence between individuals and makes the case for the legitimacy of solidarity on the basis of this marriage turns out to lose much of validity against the backdrop of the global division of labor and the highly polarized society. Before ending the paper, I shall propose a set of directions for an alternative theory of solidarity in the contemporary society. The first direction requires refining the theory of the division of labor, which is key to the transition from the primitive society to the modern industrial society. The 141 division of labor appears in Durkheim’s theory of organic solidarity as a mechanism that legitimates the rise of interdependence from increasing individualism and interpersonal differences. Throughout his discussion of the division of labor, Durkheim pays overwhelming attention to the division of labor at the domestic level, which was probably a limit caused by the stage of capitalism he had witnessed during his lifetime. Yet, the progress of globalization and increasing cross-border flows of capital not only expanded the scope of economic activities but also rearranged the places of economic production on the globe. This in turn led to the specialization of production at the global scale, resulting in the new division of labor in the global north and south. As discussed in detail in the previous section, the global division of labor had a profound impact on the domestic economy in developed economies, inspiring the backlash from the victims of globalization. The inter-group conflicts and animosity caused by globalization certainly weakened social cohesion while breaking the existing social bonds. Clearly, Durkheim’s account of the division of labor does not appear to be capable of addressing these issues, which in turn calls for a fundamental revision of his thesis on the division of labor in ways that account for the consequences of global economy and explore possible social processes that reinforce social (re)integration in the age of globalization. The second direction for an alternative theory of solidarity is concerned with the multidimensionality of differences and conflicts between citizens in the contemporary society. In DL, Durkheim correctly stressed the rise of individualism in the modern industrialized society, pointing out that increasing heterogeneity and the pursuit of self-interests caused the decline of collective consciousness that had 142 embodied the commonality of individuals in the primitive society. And yet, it is important to note that the heterogeneity and differences between individuals have become more complex to the extent that the thesis that they cannot be resolved through the interpersonal interdependence functionally drawn from the division of labor. Not only does research on political polarization demonstrate a consistent tendency among citizens to have highly conflicting or opposing views over many issues, but it also shows that polarized attitudes toward diverse issues are highly associated with each other, meaning that citizens who have opposing views on some issue are very likely to have opposing views on the other issues as well. As such, growing polarization and its multidimensionality point to deep-seated antagonism among citizens in the contemporary society. Accordingly, an alternative approach to solidarity need to tackle the multidimensional nature of attitudinal differences and conflicts to explore how those differences and conflicts can be reconciled. The last direction I shall propose is related to politics. The previous section stressed that both the global division of labor and growing polarization have inspired the rise of populist leaders and parties over the last few decades. As discussed in Chapter 2, populists seek to capitalize on the divisive ideology, advocating the rights of their in-groups while excluding the out-groups, and thrive over and reinforce the existing antagonism and conflicts. Throughout his discussion of mechanical solidarity and organic solidarity, however, Durkheim tends to pay very little attention to the agents of solidarity: this appears to be salient in his discussion of organic solidarity grounded in the functional necessity of cooperation of individuals. A sizable volume of literature on populism, however, highlights the role of political leaders and parties 143 in spreading the populist ideology and, needless to say, winning voter support. Clearly, divisive campaigns and messages from the agents of populism should not be overlooked, because they are major countervailing forces against social cohesion. Accordingly, the theory of solidarity should shed light on the agent side of social cohesion. 144 REFERENCES Baldassarri, Delia, and Peter Bearman. 2007. “Dynamics of Political Polarization.” American Sociological Review 72(5):784–811. Baldassarri, Delia, and Andrew Gelman. 2008. “Partisans without Constraint: Political Polarization and Trends in American Public Opinion.” American Journal of Sociology 114(2):408–46. Bolce, Louis, and Gerald De Maio. 1999. “Religious Outlook, Culture War Politics, and Antipathy Toward Christian Fundamentalists.” The Public Opinion Quarterly 63(1):29–61. Cardoso, Fernando Henrique. 1972. “INDUSTRIALIZATION, DEPENDENCY AND POWER IN LATIN AMERICA.” Berkeley Journal of Sociology 17:79–95. Cardoso, Fernando Henrique. 1977. “The Consumption of Dependency Theory in the United States.” Latin American Research Review 12(3):7–24. Durkheim, Émile. 1984. The Division of Labor in Society. New York: Free Press. Durkheim, Emile. 1995. The Elementary Forms of Religious Life. New York: Free Press. Fiorina, Morris P., and Samuel J. Abrams. 2008. “Political Polarization in the American Public.” Annual Review of Political Science 11(1):563–88. Fiorina, Morris P., Samuel J. Abrams, and Jeremy Pope. 2005. Culture War? The Myth of a Polarized America. Boston, MA: Longman. Gereffi, Gary. 1996. “Global Commodity Chains: New Forms of Coordination and Control among Nations and Firms in International Industries.” Competition & Change 1(4):427–39. Gereffi, Gary. 2001. “Shifting Governance Structures in Global Commodity Chains, With Special Reference to the Internet.” American Behavioral Scientist 44(10):1616–37. Giddens, Anthony. 1971. Capitalism and Modern Social Theory: An Analysis of the Writings of Marx, Durkheim and Max Weber. Cambridge, UK: Cambridge University Press. Giddens, Anthony. 1986. The Constitution of Society: Outline of the Theory of Structuration. 1. paperback ed. Berkeley: Univ. of California Press. 145 Held, David, Anthony G. McGrew, David Goldblatt, and Jonathan Perraton. 1999. Global Transformations: Politics, Economics and Culture. Stanford, CA: Stanford University Press. Hutter, Swen, and Rens Vliegenthart. 2018. “Who Responds to Protest? Protest Politics and Party Responsiveness in Western Europe.” Party Politics 24(4):358–69. Iyengar, Shanto, Yphtach Lelkes, Matthew Levendusky, Neil Malhotra, and Sean J. Westwood. 2019. “The Origins and Consequences of Affective Polarization in the United States.” Annual Review of Political Science 22(1):129–46. Iyengar, Shanto, Gaurav Sood, and Yphtach Lelkes. 2012. “Affect, Not Ideology.” Public Opinion Quarterly 76(3):405–31. Iyengar, Shanto, and Sean J. Westwood. 2015. “Fear and Loathing across Party Lines: New Evidence on Group Polarization: FEAR AND LOATHING ACROSS PARTY LINES.” American Journal of Political Science 59(3):690–707. Jacoby, William G. 2014. “Is There a Culture War? Conflicting Value Structures in American Public Opinion.” American Political Science Review 108(4):754–71. Jardina, Ashley. 2019. White Identity Politics. Cambridge, United Kingdom: New York, NY, USA : Cambridge University Press. Levendusky, Matthew, and Neil Malhotra. 2016. “Does Media Coverage of Partisan Polarization Affect Political Attitudes?” Political Communication 33(2):283– 301. Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. New York: Crown. Milner, Helen V. 2021. “Voting for Populism in Europe: Globalization, Technological Change, and the Extreme Right.” Comparative Political Studies 54(13):2286– 2320. Morgan, Stephen L. 2018. “Status Threat, Material Interests, and the 2016 Presidential Vote.” Socius: Sociological Research for a Dynamic World 4. Mudde, Cas, and Cristóbal Rovira Kaltwasser. 2018. “Studying Populism in Comparative Perspective: Reflections on the Contemporary and Future Research Agenda.” Comparative Political Studies 51(13):1667–93. Norris, Pippa. 2019. Cultural Backlash: Trump, Brexit, and Authoritarian Populism. New York: Cambridge University Press. Przeworski, Adam. 2019. Crises of Democracy. 1st ed. Cambridge University Press. 146 Rodrik, Dani. 2018. “Populism and the Economics of Globalization.” Journal of International Business Policy 1(1–2):12–33. Sassen, Saskia. 1998. Globalization and Its Discontents: Essays on the New Mobility of People and Money. New York: New Press. Sassen, Saskia. 2000. “Territory and Territoriality in the Global Economy.” International Sociology 15(2):372–93. Sassen, Saskia. 2003. “Globalization or Denationalization?” Review of International Political Economy 10(1):1–22. Stiglitz, Joseph E. 2002. Globalization and Its Discontents. New York: W. W. Norton. Swank, Duane, and Hans-Georg Betz. 2003. “Globalization, the Welfare State and Right-Wing Populism in Western Europe.” Socio-Economic Review 1(2):215– 45. Swedberg, Richard. 2005. The Max Weber Dictionary: Key Words and Central Concepts. Stanford, CA: Stanford University Press. Wallerstein, Immanuel. 1976. “A World-System Perspective on the Social Sciences.” The British Journal of Sociology 27(3):343. 147 Concluding Remarks and Avenues for Future Research My dissertation research revealed the consequences of the economic and political crises that broke out in the first two decades of the twenty-first century, demonstrating how individuals and social groups reacted to the financial crises, how a set of fundamental shifts in the economy and institutional politics leveraged the rise of populism of various kinds, and how the global economic transformations and crisis and polarization undermined the conventional process of social integration. Combined together, the three chapters sought to shed light on the consequences of the economic and political crises for both social actors and political systems. Sociologists have tried to understand the occurrence of crises as a critical moment that reshapes the ways in which a society maintains itself (Abbott 2016; Giddens 1986; Swidler 1986). As we have just lived the first two decades of the twenty-first century and a lot of crises, whether in known forms or in unknown forms, may have yet to come in upcoming years, our knowledge still remains very limited. I hope my dissertation research helps to build a framework for understanding the possible changes the economic and political crises – with the case of global financial crisis and populism – bring to social actors and institutions in the twenty-first century. Clearly, there are many issues that need to be tackled to account for the social dynamics of crises. Above all, the global pandemic crisis that broke out in the late 2010s arose as one of the most important themes for the sociological study of crises. Perhaps, the COVID-19 had the greatest consequences for economic, political and social lives in the contemporary social world over the last few years. Not only do many expect its impact to last long, but also believe that another pandemic may come 148 in the near future and have even more serious consequences. Scholars of crisis dynamics are thus encouraged to delve into the pandemic more than ever. There are other important avenues for future research on crises. One relates to a long-term impact of political crises on our society. Mt future research project examines how the rise of populism and our divided society would affect democratic governance. The other is concerned with how the economic crisis can be politically managed. Especially, I am planning to explore whether and to what extent two major doctrines for the management of economic crisis, stimulus and austerity, exert persuasive leverage. I shall close by briefly introducing my future research in each area. FUTURE RESEARCH AGENDA (1): POPULISM, POLARIZATION, AND DEMOCRACY In connection with my dissertation research on populism, the crisis of democracy and democratic governance constitute an important pillar of my future research. I am currently designing a project that examines the long-term effect of affective polarization (partisan animosity) on the state of democracy. Affective polarization and its supposedly negative impact on democracy have inspired an explosion of research articles and debates in recent years, especially among scholars of American politics and political culture. And yet, little is known about specific conditions under which affective polarization drives democratic decay. Especially, the existing literature does not give generalizable evidence for the concrete social mechanisms connecting affective polarization to varying degrees of democratic decay. 149 I will undertake a rigorous causal analysis of the (potentially co-integrated) dynamics of polarization and the state of democracy, analyzing large-scale dynamic panel data compiled from multiple harmonized public opinion and cross-country political/economic data. The finalized dataset covers more than 30 countries over 50 years. Combining the insights from my research on populism and polarization, my long-term goal (set as a book project) is to unveil the social processes that have profoundly undermined the workings of representative democracy and hopefully to propose a set of preliminary agendas for the renewal of democratic governance. At the heart of the populist surge and increasing polarization is the unprecedented level of social divides that have prevailed in many democracies across the globe over the last few decades. I argue that understanding the nature of those social divides requires a synthetic account of populism and polarization, the two central pillars of the social divides that tend to be considered separately in the existing literature despite the structural isomorphism underlying the ways in which populist discourses and partisan animosity become reinforced over time. Revealing the relational dynamics of populism and polarization leading to growing social divides would certainly offer a concrete social mechanism that eventually aroused democratic decay. It would also offer new research agendas for the renewal of democracy, revealing the structural forces behind the ruptures in democratic governance and in turn shedding new light on possible steps of repair. 150 FUTURE RESEARCH AGENDA (2): POLITICAL MANAGEMENT OF ECONOMIC CRISIS Next, in connection to my previous research on economic crisis and redistributive justice, my future research seeks to give an extensive insight into mass politics in the age of global economic crisis. I am currently designing a collaborative project that assesses the persuasive power of austerity vis-à-vis that of its rival doctrine: fiscal stimulus. Popularized by economists and policy-makers in the post-war era (often dubbed the “Hayek versus Keynes” debate), the conflicting relations between the two doctrines have been a recurring theme in times of financial crisis, since they offer opposite programs for recovery as well as welfare provision and redistribution amid the crisis. And yet, their respective persuasive power has seldom been tested. With a group of colleagues, I will conduct a conjoint Bayesian analysis of multiple dimensions of public preferences over fiscal tightening and stimulus to estimate the communicative efficacy of the two doctrines simultaneously proposed in response to growing fiscal imbalances. This research is expected to shed new light on the potential mechanisms that reconcile the conflicting relations between social welfare and crisis management and enable policy-makers to pursue redistributive agendas in hard times. 151 REFERENCES Abbott, Andrew. 2016. Processual Sociology. Chicago; London: The University of Chicago Press. Giddens, Anthony. 1986. The Constitution of Society: Outline of the Theory of Structuration. 1. paperback ed. Berkeley: Univ. of California Press. Swidler, Ann. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(2):273–86. 152