Jump to ContentJump to Main Navigation
Populism and Liberal DemocracyA Comparative and Theoretical Analysis$

Takis S. Pappas

Print publication date: 2019

Print ISBN-13: 9780198837886

Published to Oxford Scholarship Online: May 2019

DOI: 10.1093/oso/9780198837886.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 24 August 2019

What is Populism?

What is Populism?

(p.13) 1 What is Populism?
Populism and Liberal Democracy

Takis S. Pappas

Oxford University Press

Abstract and Keywords

Chapter 1 begins with a concise overview of the ways earlier scholars have examined populism during consecutive waves of research on the topic, and goes on to expose common conceptual and methodological errors—such as obscuring the genus, overemphasizing essentialism while overlooking ontology, concept stretching, the incertitude about negative poles, and the difficulties with concept operationalization—before setting out to elaborate a truly minimal definition that can sufficiently account for the comparative study of modern populism. The chapter ends with a presentation of the book’s overall framework of analysis, which condenses the main methodological tools—such as indicators, variables, and concept properties—to be used later on during theory-building.

Keywords:   methodological errors, genus problem, essentialism, conceptual stretching, negative pole, operationalization, minimal definition, democratic illiberalism, variables, indicators

The study of populism is evocative of the story of the blind men and the elephant. This story, in which some blind men are asked to touch an elephant to determine what that thing is, has been told numerous times and in many versions, but never to its very end. The known part of the story is about each blind man feeling only one bit of the animal so that, on comparing their thoughts, they may decide on the object. In typical versions of the story, the blind man who feels the tusk says it is like a big tree branch; the one who touches a leg says it is like a big pillar; he who feels the belly says it is like a big wall; the one who touches the ear says it is like a big hand fan; and he who feels the trunk says it is like a big pipe. In this part of the story, the blind men cannot reach agreement about the thing examined.

But our men are blind, not deaf. Therefore, they know that what they touch is a living animal that breathes, moves, and emits sounds. Here comes the untold part of the story. In it, there exists another blind man who sits in a corner without touching the animal, only listening attentively to what the others have to say. He finally asserts: From what you are saying, this thing must be the biggest animal on earth! With that declaration, our man offers an ontological definition of the thing under examination based on size as its core characteristic (“the biggest animal”) while also contextualizing it (“among all fauna”). And, as none could argue against that conclusion, this is how since then the blind men identify their animal—“the biggest one on earth”—and tell it apart from all other fauna. End of story.

Now, the moral: As with the fabled elephant, and all its other features notwithstanding, populism is the major historical phenomenon of our times, currently posing an elephantine threat to many of the liberal democracies that became dominant after the end of World War Two. In this opening chapter it is therefore worth beginning our exploration into populism with an examination of the existing scholarship on the subject and then trying to contextualize it within modern-day democratic liberal politics. We are going to point out some of the major deficiencies in the study of populism, mostly due to limitations of theoretical perspective or methodological fallacies and other shortcomings. Like the attentive blind man in his corner, we may then be able (p.14) to arrive at a minimal definition that, hopefully, will tell us what the political animal we call “populism” really is. And, from there, we will also be better equipped to construct an overall theoretical framework to act as a guiding light in our explorations into modern populism.

1.1. The Epistemology of Populism

Since our first objective is to reconceptualize populism, let us commence with sage advice: “The rebuilding of a concept,” Sartori counsels (1984: 46), “begins with looking into its literature” and extracting its characteristics. Interestingly enough, when one looks into that literature, the first thing to observe is that it is relatively recent. The systematic study of populism in academia began only in the late 1960s and, to this date, has gone through three distinct waves, or “generations,” of scholarship, each corresponding to a particular time period and characterized by its own intent, agenda, and research design. In this chapter, I provide a concise overview of existing scholarly efforts to seize populism in analytic epistemological terms. Yet, rather than merely trying to catalog those efforts (for such a listing, see among others Moffitt 2016: 12–25), I am going to distinguish between cohorts of scholars who, in different periods and places, examined phenomena that they labeled “populist.” For each cohort of scholars and corresponding wave of scholarship, I will place emphasis particularly on three aspects: First, the concept’s intension, i.e., the collection of characteristics, or properties, which determines the empirical things to which populism applies and answers the question “What is populism?” Second, the concept’s extension, i.e., the empirical class of things to which populism refers while the point at issue becomes “Who are the populists?” And, third, the scholarly gains, but also particular difficulties, encountered in each of those endeavors, which raise questions of the type “Why is populism so important in contemporary politics?” For convenience, I will refer to the three groups of scholars roughly belonging to each of the three waves as the pioneers, the classics, and the contemporaries.

The Pioneers

The earliest wave of studies on populism originated in a conference held in 1967 at the London School of Economics, during which a multidisciplinary cohort of scholars explicitly set out to define the phenomenon.1 The (p.15) publication of the conference proceedings, which appeared in book form under the title Populism: Its Meaning and National Characteristics (Ionescu and Gellner 1969), effectively marks the beginning of the first wave of scholarship on populism, which then expanded into the 1970s and early 1980s.

The focus for those early scholars of populism was not their contemporary European societies or any other democratic and capitalist states. This is not surprising, of course, given that, by the time of the conference, postwar Europe’s only two political groups deserving the populist label had been the Italian Common Man’s Front2 and the French Poujadisme, initially a small shopkeepers’ movement against taxation.3 Both had proved stillborn.

What those early scholars on populism perceived instead was the rise of a “global populism” in pre-modern states located in the world periphery, including Africa, Asia, and the then communist Eastern Europe. Even so, lacking a clear concept of populism, the early research on the phenomenon looked, indeed, like a “fishing expedition without adequate nets” (Sartori 1970: 1039). The catch was impressive, as it included revolutionary movements seeking the end of feudalism, such as the Narodniki in imperial Russia; nondemocratic regimes, such as Latin America’s numerous autocracies; interwar peasant movements in Scandinavia, Eastern Europe, and the Balkans; anti-capitalist and anti-colonial crusaders in Africa and Southeast Asia. And when it came to more developed nations, like the United States or Canada, the point was not missed by the pioneering scholars that populism had only developed in the economically most backward and politically more underdeveloped rural areas of these countries. Such, for instance, were the cases of American populism, which flourished during the late nineteenth century in the agrarian southern and midwestern states among the “declining ‘liberal’ (p.16) classes living in declining areas” or of the Progressive Party of Saskatchewan, a political movement that developed in the prairies of Alberta in Canada from the 1920s to the mid-1930s (Lipset 1960: 173, 194). With reference to Western Europe, the only instances of populism mentioned by one conference participant were “a kind of proto-populism in the English Peasants’ Revolt and the Jacqueries of the fourteenth century, or in the Bundschuh and the peasant wars of the Reformation” (MacRae 1969: 154).

All in all, then, the pioneers in the study of populism were particularly keen to draw comparisons between disparate cases of “populism,” which they attributed to the lack of socio-political modernity and market economy. Since, during the 1960s and early 1970s, most societies around the world were pre-(socially)modern, non-(liberal)democratic, and pre-(advanced)capitalist, populism was seen to “bob up everywhere, but in many and contradictory shapes” (Ionescu and Gellner 1969: 1), its manifestations including, among many others,

the droolings of Tolstoy over muzhiks, the rationalizations of Eastern European resentments against alien traders, and the slogans in terms of which rulers of new nations legitimate themselves and subvert liberal institutions.

(Ionescu and Gellner 1969: 1)

On this realization, the question certainly arose about whether populism had any underlying unity or if one label covered disparate phenomena. Unfortunately, that early cohort of scholars was unable to reach a unanimous verdict on this matter—which is probably why their volume lacks a final chapter that would sum up findings and extract conclusions.

What prevented those pioneering scholars from defining populism and, eventually, explaining the phenomenon? The obvious reason is their stretching of the extension of the concept without a prior organization of its intension. The early-day scholars saw populism everywhere but had no time to reflect on its ontology; nor did they bother to distinguish between the many different historical and political contexts within which populism seemed capable of emerging. In reality, however, all this happened because modern populism had not yet emerged. Even by the early 1980s, when Margaret Canovan wrote her first book on the subject, populism was understood as a reactionary political occurrence, consisting of “a bewildering variety of phenomena” (Canovan 1981: 3) mostly related to various agrarian movements or associated with autarchic leaders, such as Argentina’s Juan Perón.

The earliest wave of scholarship on populism, in sum, failed to provide a commonly agreed definition of populism, causing instead conceptual stretching and empirical confusion. Concept ambiguity thus settled in. But there were gains, too. It sensitized the broader scholarly community to the importance of populism as a distinct—and certainly curious—phenomenon and put its study firmly into the agenda of historical analysis and comparative politics.

(p.17) The Classics

If the first wave of scholarship on populism was prompted mostly by European scholars out of academic curiosity, the second wave was occasioned by natives of Latin America (and, later, their students and disciples in North American universities) who were both highly politicized and endowed with first-hand knowledge of the countries whose tumultuous politics they set out to examine. It was this group of scholars who, even more than either their predecessors or their immediate successors, and by virtue of the pure quality of the cases they studied, reached conclusions of lasting worth—hence my labeling them as classics.

This group of scholars may be divided into two subgroups: the first of them focused attention on Latin American populism in the early postwar decades, which were characterized by the so-called policy of “import substitution industrialization” (ISI) (classically treated in Hirschman 1968 and Baer 1972); the second group became more preoccupied with explaining the unexpected compatibility of populist politics and neoliberal economics in Latin America in more recent decades. An abbreviated overview of both groups follows.

In the aftermath of World War Two, as Latin America entered a period of sustained economic growth, populism grew strong in the region. Almost the entire continent turned toward ISI aiming to create self-sufficient economies based on large industrial sectors and the expansive middle classes. Within that context, populism became the chief means of markedly autocratic leaders who, seeing the opportunity for extracting political gains, advocated free elections, the enfranchisement of women and the illiterate, and the entry of the masses (el pueblo) into politics (Germani 1978).4 So consequential was the role of individual leaders such as Getúlio Vargas in Brazil, Juan Perón in Argentina, or José María Velasco Ibarra in Ecuador that the populist systems they helped create are still remembered by their names: Varguismo, Peronismo, Velasquismo (and, in Peru, Aprismo, from the initials of Haya de la Torre’s party name).5

In such an historical and political context, and unlike the European pioneers who aimed at definitions, Latin American scholars saw populism as a phenomenon primarily related to the socioeconomic determinants of mass political movements that developed contemporaneously in their respective countries. As put by one of the most important representatives of the group, the chief concern of Latin American scholarship was to explain “the (p.18) conditions under which the political participation of the lower classes is channeled through a populist movement” (Germani 1978: 95). It was under this view, too, that many in this group of scholars considered populism as beneficial to Latin America’s democracies—in the words of Carlos de la Torre (2000: xiii), it was “a fundamental democratizing force that marked the entrance of common people into the political community.”

Progressively, there developed within this group of populism specialists two distinct approaches, one associated with the so-called “modernization theory” and another with structural Marxism. For the adherents of modernization theory, populism served as the means to incorporate into politics the newly mobilized urban working and middle classes that had emerged after the postwar breakdown of oligarchic regimes and the transition to capitalism and modernity (Germani et al. 1973, Malloy 1977, Collier 1979, Drake 1982). For the adherents of Marxism and the so-called Dependency School, populism was a multiclass political movement corresponding to the objective of import substitution (cf., O’Donnell 1973, Cardoso and Faletto 1979). It was precisely the statist and nationalist policies of ISI, according to this interpretation, that “allowed populist leaders to build cross-class alliances between urban labor, the middle sectors, and domestic industrialists” (Roberts 1995: 85). Notwithstanding their differences, both approaches saw populism as specific to the historical and political circumstances of development in the world semi-periphery. And both “agreed on the importance of defining it in social terms, rooted in relations of production and market conditions” (Jansen 2011: 79).

Thus having organized their research intension (i.e., populism as a means for mobilization in specific historical circumstances), the present group of scholars had no problem selecting their cases, which mostly included countries in Latin America but, occasionally, other countries in the semi-periphery, such as Greece (Mouzelis 1985). The high-profile cases were, of course, those of Argentina (Germani 1978, James 1988, Di Tella 1990, Horowitz 1999), Brazil (Conniff 1981, French 1989, Wolfe 1994, Conniff 1999), and Mexico under Lázaro Cárdenas (Knight 1990, Knight 1994, Basurto 1999). There were also several less well-known but still important cases (including, for instance, Haya de la Torre in Peru, Villaroel in Bolivia, Guzmán in Guatemala, or Gaitán in Colombia) that “may have been shorter-lived or less consequential than those of Perón, Vargas, and Cárdenas, but a theory of populist mobilization suggest[ed] that they share many meaningful similarities with [the] high-profile cases” (Jansen 2011: 88–9).

Despite its relatively good match between concept intension and research extension, however, the problem with this subgroup of classical scholarship on populism was its lack of comparability potential. In other words, it could only be used to explain authoritarian populist movements falling within the specific socioeconomic context of Latin America during ISI and, perhaps, a few other countries in the world semi-periphery. Therefore, this perception of (p.19) populism had little traveling capacity beyond Latin America or similarly conditioned world regions. As openly admitted by one of its proponents, this type of populism may materialize into “a mass movement only in societies where typical Western European leftist ideologies of the working class fail to develop into mass parties” (Germani 1978: 88). Even so, this specific literature yielded significant gains as it sensitized scholars to at least two important features of populism during its emergence phase: first, its mass movement character and, second, the role of individual agency and, specifically, the importance of charismatic leadership (see e.g., Dix 1978, de la Torre 2000). As will be shown later on, these two features are always present in any and all significant cases of modern populism analyzed in this book.

The eventual exhaustion of import substitution strategies in Latin America eroded the material foundations of state-based populism and led to the temporary demobilization of those actors “whose economic interests had defined traditional populist agendas” (Roberts 1995: 82, Conniff 1999). Yet, despite the fact that both modernization and dependency theories of populism gradually fell into disrepute, populism hardly disappeared. The spectacular emergence of Hugo Chávez in Venezuela was a reminder that old-style quasi-democratic populist politics in the region was far from dead. But that was not all. During the 1980s and 1990s, a new type of populism flourished in Latin America in a radically different socioeconomic and political environment than that of the earlier postwar decades.

Already by the late 1980s, younger Latin American scholars were astonished to find out that a new breed of populist politicians in the region were able to implement neoliberal policies while also enjoying remarkably high levels of popular support—a phenomenon termed neo-populism. The most prominent such cases were those of Peruvian Alan García and, subsequently, Alberto Fujimori, also in Peru; Carlos Menem in Argentina; Fernando Collor de Mello in Brazil; and Carlos Salinas de Gortari in Mexico. In all these countries, populism enjoyed a second political epiphany, which distinguished it from classical populism in two respects: The first was its social bases, which now consisted of members of the urban informal sector and the rural poor, as opposed to the organized working class that was the mainstay of earlier Latin American populism; the second difference was, of course, the implementation of neoliberal policies instead of import substitution and state corporatism (Roberts 1995, Weyland 1999).

What therefore became the central puzzle in this new scholarship was the antinomy between populism and neoliberalism, and, in particular, “the rise of personalist leaders with broad-based support, who follow neoliberal prescriptions for economic austerity and market-oriented structural adjustments” (Roberts 1995: 82). According to mainstream theory, the endeavor of most Latin American countries for a neoliberal adjustment of their economies presented upcoming political entrepreneurs with new opportunities for


the transformation and revival of populism under a new guise, one that is shaped by the breakdown of more institutionalized forms of political representation and the fiscal constraints that inhere in a context of public indebtedness and a diminished state apparatus.

(Roberts 1995: 114)

There have been several gains from this particular scholarship on populism. First, there was a renewed interest in charismatic leadership emerging in politically pluralist, rather than oligarchic, systems (Weyland 2001: 14, Hawkins 2003: 1138)—a subject to which we are going to return later in Chapter 3.2. Second, Latin America’s neoliberal populism exhibited much better comparability with other countries experimenting with similar solutions, such as those of Eastern Europe in the aftermath of communism (Weyland 1999, Gurov and Zankina 2014). Third, and perhaps most important of all, it offered extra proof that populism should rather be studied in a politically instrumental fashion—which, accordingly, would make its successes highly contingent on strategic political leadership and symbolic politics—rather than in a deterministic relationship to the state of the economy and the specific characteristics of the market.

The Contemporaries

In more recent years, the study of populism has grown—and is still growing—exponentially. In the field of comparative politics, in particular, populism has turned into a kind of “cottage industry” that, in coordination with often sentimentalist journalism and partisan punditry, has helped make it a modern buzzword. Largely based on the idea that a powerful populist zeitgeist, evident already by the early 1990s, has overwhelmed Western democracies (e.g., Mudde 2004), a global army of scholars has undertaken to analyze every nook and cranny of whatever one may term “populism.” All in all, contemporary research has made headway in four moves: searching for ever-innovative definitions with general applicability; expanding its comparative scope to include more and more country or political party cases; increasingly banking on quantitative methods, also displaying a certain obsession to “measure” populism; and, even more recently, trying to link the findings of empirical analysis with a normative discussion about liberalism’s decay and even democratic deconsolidation.

The search for definitions, first, is inexorably linked to how modern-day scholars approach populism by focusing on one or more of its essential characteristics. In a way strikingly similar to the blind men we met earlier, each of them touching just a part of the elephant’s mass, populist scholars choose to focus attention on one or more specific features of populism while neglecting others that are also essential. Those features include strategic (p.21) leadership, ideology, discourse and symbolic patterns, mass mobilization capacity, style, and many more. But as long as attempts for a universal definition privilege some of those features to the expense of others, such attempts are bound to fall short and fail their initial purpose.

The second move is to swell the field in order to include more and more cases for analysis. Initially conducted by close-knit communities of scholars located mostly in Europe and Latin America (Weyland 1999, Mudde and Kaltwasser 2012, Hawkins and Silva 2015), the study of populism has suddenly gone global (Judis 2016, Moffitt 2016, Müller 2016a). The appearance of Donald Trump, in particular, has propelled a fresh wave of US scholarship on populism. New research communities have emerged in various other places in Southeast Asia, Australia, and Africa, dedicating themselves to the study of populist phenomena. But, given the lack of a common conceptual understanding of populism, let alone a causal theory of its emergence, the situation today looks paradoxically like it was during the first wave of scholarship back in the 1960s and 1970s: “populism” is everywhere and everything is “populist.”

The third move of recent scholarship of populism is the emphasis on numbers and measurements. Perhaps because of the inability to settle the “what is” question, several students of populism took it upon themselves to assess degrees of populism utilizing either content analysis of populist discourse (e.g., Jagers and Walgrave 2007, Hawkins 2009) or textual analysis of manifestos and other party literature (e.g., Pauwels 2011, Rooduijn and Pauwels 2011, Aslanidis 2017). An even more recent area of growing interest is measuring populist attitudes at the mass social level (Akkerman et al. 2014, Rooduijn 2018). For all their sophistication, however, those efforts remain problematic as long as there is no agreement about what constitutes populism and what does not.

The fourth and more recent move of various scholars dealing with populism is to alert us about the growing threat of populism to liberal democracy and even to representative parliamentary politics (Urbinati 2014, Levitsky and Ziblatt 2018, Mounk 2018). In this particular perspective, populists represent—and, indeed, lead—insurgent anti-establishment forces that deny the legitimacy of mainstream parties and attack them as corrupt and elitist, detached from the ordinary voter. This may well be so, but it simply emphasizes the urgent need for a prior comparative empirical study of how populist parties behave, and succeed in their purposes, when they rise to power.

With so much water having flowed during past decades, does the bridge we stand on today still have firm foundations? Alas, they are far from solid. Despite the breathtaking expansion of output on populism, the basic problems encountered by the pioneers remain almost the same. Think of the concept’s many intensions that, in the face of ostentatious claims to the opposite, still prevent agreement about what populism is. No wonder, then, that the extension of the concept has become so overstretched as to allow in the populist (p.22) universe disparate countries, political parties, or individual leaders without sufficient discrimination. This is why, lacking a theory about populism, we are still unable to understand its causes and micro-mechanisms, which would enable the application of countermeasures to combat it and, hopefully, rescue liberal democracy.

Inevitably, our indeterminacy about populism leads to awkward outcomes, particularly as long as insurgent parties become increasingly successful and, sometimes, even rise to power. Labeling all those insurgents “populist” simply shows that the conceptual and theoretical mileage that we can achieve is still small—and, in fact, some of it has been afforded in reverse gear. For, not only do we try to compare different—which is to say, non-comparable—cases to no obvious theoretical avail, but, more disturbingly, we sometimes are not even able to recognize populism when we see it. Let me illuminate these points by using three examples from a single year, 2016, which was precisely the year in which “populism” became a phenomenon.

The first quite remarkable event of the year was the so-called Brexit, that is, the decision of the United Kingdom to leave the European Union (EU) after a referendum held on June 23, 2016, in which 52 percent of voters voted in favor of leaving the EU. That outcome was the result of increasing Euroskepticism in the UK, blended with growing anti-immigration sentiments, which were effectively expressed by the official campaign group for leaving the EU, aptly called Vote Leave. Without a doubt, the most important spokesman of the leave campaign was Nigel Farage, a former commodities trader and leader of the UK Independence Party (UKIP), who, repeatedly, had failed to win election to the British parliament. And yet, Farage, declaring himself the leader of the fight for “the people’s ‘revolution,’” was successful in carrying that out and, at the end, claimed Brexit to be “a victory for the real people.”

The second momentous event was prompted by a rather amateurish coup d’état that was attempted in Turkey on July 15, 2016 against the increasingly autocratic government of Recep Erdoğan. When the coup failed, mass arrests of soldiers and civilians followed, accompanied by an extensive purge of judges, university professors, school teachers, and public officials that some characterized as a “countercoup” (The Economist, July 23, 2016). When pro-government protesters demanded the reinstitution of the death penalty, Erdoğan retorted that he was ready to endorse it “because the people will get what they want,” to which he added, “sovereignty belongs to the people.”6 This, however, did not prevent the Turkish leader from declaring a state of emergency on July 20, that is, the suspension of parliamentary democracy and the introduction of rule by decree. When human and other social and individual rights were also suspended, Turkey ceased being an actual democracy.

(p.23) The third historic event of 2016 was the presidential election in the United States and Donald Trump’s unexpected victory. He was voted into office having promised to punish the domestic political establishment, to clean up unsavory aspects of government practice disliked by many American voters, and to reinstitute the American dream. In short, he was elected in order to “Make America great again,” a popular campaign slogan. As Trump further stated during his electoral campaign,

Our movement is about replacing the failed and corrupt political establishment by a new government controlled by you, the American people. … The political establishment that is trying to stop us is the same group responsible for our disastrous trade deals, massive illegal immigration, and economic and foreign policies that have bled our country dry.7

As should be clear by the brief precis of the three cases, which occurred in quick succession during one single year, all of them were effected in the name of the people. It is no less clear, though, that those cases speak about quite different phenomena leading to dissimilar outcomes: A Euroskeptic nativist campaigned successfully for his country’s departure from the EU; an autocratic leader suspended pluralist democracy by countercoup; and an egocentric candidate using pure populist discourse rallied against the political establishment to win a presidential election. Such obvious differences notwithstanding, many people—academics no less than journalists and pundits—still found it pertinent to label all three leaders populist. For instance, immediately after the conventional warning that “we should stop the inflationary use of the term ‘populism,’” Jan-Werner Müller advised putting Farage, Erdoğan, and Trump into the same populist basket because their “group claims exclusively to represent the one authentic people” (Müller 2016b).

What does this analytical muddle and theoretical bewilderment tell us about the present state of research on populism? They indicate that, for all our apparent progress, the original problems remain. This is why, as populism seems to grow stronger, we still don’t know how to deal with it. But to deal with it, we need to know what causes it in the first place. But to establish causality, we need to do meticulous empirical research on several significant cases of populist occurrence. But to do such empirical work, we need first to select our cases carefully and, if possible, universally. But to be able to select the cases, we must first proceed with a decent classificatory scheme that tells populists apart from non-populists. But to construct such a scheme, we need to have a clear and unambiguous definition of the object under consideration. It is only when we succeed with all the foregoing tasks that we may hope to end up with a sound theoretical framework for the comparative analysis of populism.

(p.24) 1.2. Dealing With Deficient Premises

In June 2016, The Economist published the following letter to the editor by one of its dismayed readers, closing with a valuable suggestion:

The Economist’s addiction to the epithet ‘populist’ has spun out of control. You put that label on pitches and policies as different as hard-wired xenophobes, plutocrato-phobes [sic], economic chancers, thoughtful progressives, trade protectionists and political opportunists. Your recent list includes (among many others) Pat Buchanan, Marine Le Pen, the Kirchners, Jeremy Corbyn, assorted middle-European cryptofascists, the [UK newspaper] Sun, a long-established centrist Irish political party, [and] Latin American presidents who cap the pay of senior civil servants. … What do the members of this vast, ever-growing universe actually have in common? First, they seek to appeal to the people (find me the politician that doesn’t). And second, The Economist doesn’t approve of them. I suggest restraint.

(The Economist, June 18, 2016)

Restraint, yes! But how is it achieved? This chapter continues with a brief discussion about the ways we have become accustomed to think about populism in general before it embarks on exposing specific conceptual and methodological deficiencies that must be overcome before arriving at what will be our major trophy in this part of the book—a minimal definition of modern populism.

Most of our thinking about populism stems from deductive reasoning. That is to say, it is based on general statements, or premises, which are thereafter used to reach a conclusion about whether some specific referent belongs in the populist class or not. By this logic, if the initial premise is true and its terms are clear, the conclusions reached are necessarily true. Yet, in practice, the premises set at the beginning are commonly intended to provide the sort of justification whereby, granting that the premises are true, the conclusion will be true as well.

Typically, in most of the literature about modern populism the argument is presented in the following form: Conditional on the premise that “Populism is something (e.g., a certain type of ideology, discourse, style, strategy, mobilization, economic policy, political logic, or whatever else one may choose as long as it seeks to appeal to ‘the people’) with characteristic(s) x,” it is claimed that when a certain leader, political party or, indeed, an entire political system displays that something, the conclusion is reached that that specific leader, party, or polity is in fact populist. The argument takes the form:

Populism is x;

Y features x;

therefore, Y is populist.

In this logic (known as modus ponens, or affirming the antecedent), if the antecedent condition (“populism is x”) is declared true, the consequent is also (p.25) true. The argument seems valid, but is it a sound argument as well? And how impervious is it to fallacy?

Deductive arguments are evaluated in terms of their validity and soundness. An argument is valid if and only if it is presented in a form that makes it impossible for the conclusion to be false if the premises are true. An argument is sound if it is both valid and its premises are actually true. Here is an example of an argument that is both valid and sound: On the premises that (a) all men are mortal and (b) John is a man, it is deduced that John is mortal. The problem with populism, however, is that, unlike “mortality” for which there can be no uncertainty, confusion reigns supreme when it comes to populism’s antecedent conditions. In the relevant literature there are too many requisites for populism to obtain, none of which is thought of as superior to all others. Under these conditions, a perfectly valid argument may turn unsound, either because its premises are challenged or, more commonly, because they are not specified in clear and unambiguous terms, as shown by the following example:

All birds have beaks;

this creature has a beak;

therefore, this creature is a bird.

Such a good argument, only to be instantly falsified once it is revealed that the creature under consideration is in fact an octopus—an animal also equipped by nature with a beak. For presenting the previous argument in a way that is both valid and sound, therefore, one must clearly define “beak” as the external anatomical feature birds use for eating, preening, or feeding their young, which is characterized by two bony projections—the upper and lower mandibles—covered by a thin epidermis. This kind of clarity is necessary for distinguishing birds from octopuses by dint of having different beaks.

The lesson learned with regard to the study of populism is this: in deductive reasoning about populism, the antecedent premises can be judged true only if stated in a clear and unambiguous way, which will then suggest those operations that are necessary to determine or prove it. Simply granting the premises to be true does not suffice for reaching an actually true conclusion.

The difficulty of specifying the antecedent conditions of populism often leads scholarship to the opposite direction of reasoning, that is, to affirm the consequent. In this common fallacy, the consequent is held to be a necessary and sufficient condition for the antecedent, as shown here:

If x, then populism;

Y is populist;

therefore, Y features x.

The argument here unfolds in the following way: On the premise that something x (be it some ideology, discourse, style, strategy, mobilization, etc.) is declared to indicate populism, and having already classified unit Y as populist, (p.26) research simply demonstrates that Y features the populist x. Only that, in the lack of real empirical and methodologically sound analysis, this often opens the door to erroneous conclusions. But, once the door has been opened, conceptual and methodological deficiencies abound, and further thwart our efforts to understand the populist phenomenon.

In this chapter, I am going to call attention to what I consider the five major such methodological drawbacks that have plagued the study of populism: (i) lack of genus specification; (ii) essentialism; (iii) conceptual stretching; (iv) unclear negative pole(s); and (v) poor concept operationalization.

The Genus Problem

One should naturally begin by asking: In our preoccupation with populism, what is actually being compared at the highest level of generality? Classical (i.e., Aristotelian) logic dictates that “every definition must be stated in terms of genus proximum and differentia specifica” (Hempel 1952: 5). The implication is, of course, that we must be comparing items that belong to the same genus, species, and subspecies—“in short, to the same class … [that] provides the ‘similarity element’ of comparability, while the ‘differences’ enter as the species of a genus, or the sub-species of a species” (Sartori 1970: 1036). Since in our analysis of populism we are bound to eventually engage with classifications per genus et differentiam, it is important that we are able to specify right from the beginning the general empirical universe to which this term applies.

Regrettably, at this point we are still far from a common agreement about which genus populism belongs to. As our previous glance at the literature has already shown, our empirical analysis of populism refers to several, most often diverse, types of phenomena (i.e., various genera) including political parties and social movements (Di Tella 1965: 47, Dix 1978, Germani 1978, Jansen 2011: 82); ideologies or “creeds” (Wiles 1969: 166, Laclau 1977: 172–3, Mudde 2004: 543, Stanley 2008); specific discursive patterns (Laclau 2005, Hawkins 2009: 1042, Pauwels 2011, Aslanidis 2016, Mouffe 2018); political strategies (Weyland 2001: 14, Betz 2002: 198, Ware 2002); representation “modes” (Roberts 2013: 39); or specific political styles (Kazin 1995, Knight 1998: 227, Moffitt and Tormey 2014, Moffitt 2016), some of which are particularly related to “communication techniques” (Jagers and Walgrave 2007, Urbinati 2013: 137), and many more.

The problem with this richness is that we are not certain about the family and genus of phenomena populism belongs to. A convenient method to deal with this problem is to minimize its importance and gloss over the maze. For example, after acknowledging that “there is some debate about what exactly populism is a type of: a discourse, a thin-centered ideology, or something else,” scholars may simply decide to consider “these differences as minor [and] (p.27) use the terms ‘discourse,’ ‘ideology,’ ‘outlook,’ and ‘worldview’ somewhat interchangeably” (Hawkins and Rovira Kaltwasser 2017: 514). This is a self-defeating and eventually futile strategy. As will be shown shortly, a better way out of the genus maze is to envisage a high-level conceptualization of populism that would serve as the general analytical framework for our research purposes, within which all populist occurrences (the species and subspecies) can be conveniently classified.


Based on the view that populism is characterized by a set of attributes, properties, or other characteristics that are necessary to pin down the phenomenon, most conceptualizations of populism are essentialist rather than ontological—in other words, they try to capture its external characteristics (the form) while overlooking the actual stuff populism is made of, that is to say, its matter. This is, however, frustrating given that, although “attempts to capture the essence of populism have sprung up at different times and in different places, … it is very difficult to see a consistent pattern” (Taggart 2000: 10).

To give just one example, by canvassing the relatively recent literature on populism, Rooduijn (2014: 578) recorded no less than twelve characteristics mentioned in it: people-centrism; anti-elitism; homogeneity of the people; direct democracy; exclusionism; proclamation of a crisis; simplistic language; direct communication style; polarization; image of outsider; centralization of leader; and a loosely mediated relationship between leader and followers. This, clearly, is a far from complete list as further perusal of the literature reveals additional characteristics such as a strong moral element, charismatic leadership, and much more. In this view, populism ends up being treated as a family resemblance category—one, that is, whose members share a set of commonalities or defining attributes, not all of which are to be found in every instance (Collier and Mahon 1993).

This is a broadly used approach, one, however, that is beset with problems. First, as Hempel (1952: 6) warned, the notion of essential nature of some entity is so vague “as to render this characterization useless for the purposes of rigorous inquiry.” Second, the situation is made worse with the accumulation of several such features and other properties lacking theoretical elaboration. What is “ideology”? What constitutes “morality” (and how do we perceive immoral in contemporary politics)? When does charismatic leadership obtain? Third, and perhaps most crucially, the same features that are recognized as essential in the various definitions of populism are often identical with properties declared necessary in other mass phenomena that are quite distinct from populism such as, for instance, millenarian and—whether ancient or modern—religious movements, charismatically led political communities, (p.28) regional or separatist parties, militant republicanism, contemporary right-wing or left-wing extremisms, nativism, nationalism, and even fascism. Take, for instance, “people-centrism” and its corollary notion of the supremacy of the “will of the people” which is central in several definitions of populism. Granted! But is there a single political system, whether democratic or not, that is not people-centered? “Even the most inhumane and authoritarian regimes in modern times,” notes Worsley (1969: 245), “have, at least verbally, rationalized their authority in terms of some reference to the will of the people.” The implication is clear: our analytical focus should shift from the concept’s various “essential” properties to populism’s general ontology; from the parts of the elephant to the animal itself.

Conceptual Stretching

The most visible result of augmenting the number of essential attributes that a phenomenon must possess in order to classify as “populism” is stretching the boundaries of the concept so that it may accommodate an ever-increasing number of empirical referents, albeit at the loss of connotative precision. Conceptual stretching causes indefiniteness and is conducive to undelimited conceptualizations. As astute Canovan (1981: 6) remarks, “[t]he more flexible the word has become, the more tempted political scientists have been to label ‘populist’ any movement or outlook that does not fit into any established category.”

Indeed, as we have seen in our concise review of previous waves in the study of populism, conceptual stretching has been the ancestral sin. Witness the plethora of different collective and individual units lumped together under the rubric “populism” by one of the pioneering scholars:

The Levellers; the Diggers; the Chartists; the Narodniki; the US populists; the Socialist-Revolutionaries; Gandhi; Sinn Fein; the Iron Guard; Social Credit in Alberta; Cárdenas; Haya de la Torre; the CCF in Saskatchewan; Poujade; Belaunde; Nyerere.

(Wiles 1969: 178)

The situation has hardly improved since, leading many, like the reader of The Economist earlier in this chapter, to exasperation. Not only has “populism” been used to describe such qualitatively disparate phenomena as “Maoism, Nazism, Peronism, Nasserism” (Laclau 1977: 143–4), but, as the following example shows, the list of populists may expand at whim and in quite surprising ways:

The two most successful populist leaders in late-twentieth and early twenty-first century Britain were Margaret Thatcher and Tony Blair. There was a touch of populism in Gladstone and a stronger one in Lloyd George. … In continental (p.29) Europe, Hitler was the most evil populist leader of our time, and perhaps of all time … In the United States, Franklin Roosevelt … thought he had a special relationship with the American people. In an earlier generation, William Jennings Bryan … [also] was a populist.

(Marquand 2014)

Evidently, to lump together under the same conceptual roof dissimilar political parties across the globe—medieval or pre-capitalist movements, totalitarian systems, and nationalist ideologies—is certainly no help in our attempts to make sense of modern populism. We must avoid stretching the concept of populism to a point of meaninglessness and get instead a general concept with global applicability. To this purpose, we need to reduce the number of our concept’s necessary, or defining, properties. It should not be forgotten that, assigned as we are with the task of constructing a mousetrap, it is mice after all that we want trapped and not other small mammals, reptiles, or unlucky birds of the field.

Negative Pole

Given that any definition serves to convey the meaning (what is?) of a concept, it “must embrace the whole of what it defines, but no more” (Sartori 1987: 182). Whatever remains outside the boundaries of the concept as delimited by the definition, forms an ontological contrary—what the concept is not. Therefore, when addressing the question “What is populism?” one should be able to also clearly posit its opposite: What is populism not? Only then will it become possible to decide whether a given entity is—ontologically speaking—populist or not.

As things stand, two opposites of populism have been suggested, elitism and pluralism, (cf., among others, Worsley 1993, Hayward 1996, Mudde and Kaltwasser 2013, Müller 2015), but both are problematic.

What is elitism? The term may refer to either “elite parties” or elitist ideas. The first option should be rejected for the mere lack of suitable empirical referents in contemporary politics. Elite parties, first described by Edmund Burke in the early 1770s, and best conceived as parliamentary assemblies of notables pursuing their political goals unsupported by the people (Gunther and Diamond 2003: 175–7), gradually evolved into so-called cadre parties, until they too became eclipsed by the modern mass (Duverger 1954) and later the catch-all party type (Kirchheimer 1966). Since the elite party is historically obsolete, a relic of the past, we run out of cases and the term in this sense must be dropped. The second option, elitist ideas, alludes to the interwar elitist school of thought and, in particular, Mosca’s disdain for majorities, Pareto’s circulation-of-elites theory, and Michels’ famous “iron law of oligarchy.” But, in modern liberal democracy, there are hardly any significant parties openly espousing such ideas.

(p.30) Still, “elitism” may be salvaged as a mirror of populism if (and only if) one considers political liberalism as an elitist project. In this view, as nicely put by Hochschild (1984), “elitism is perfectly compatible with liberal democracy” as liberalism “has always relied on elites to save it from itself” (cited in Lasch 1991: 567). This critique is not a new one. It originates in early debates about Joseph Schumpeter’s very notion of democracy8 and continued with Bachrach’s (1980) assessment of political liberalism as “democratic elitism.” It logically leads, though, to juxtaposing populism not with elitism but with liberal democracy qua elite-devised project—a point to which I am going to return later on.

Besides being anti-elitists, populists are also considered to be “always antipluralist” on account of their “claim that they, and they alone, represent the people” (Müller 2016a: 3, 20). This of course goes as far as it may as a claim but cannot undo the fact that modern populism always develops in pluralist political systems, where a plurality of political ideas clash, and the number of parties is always in the plural (i.e., at least two parties).9 When we say “pluralism,” what we have in mind is manyness. And when pluralism is negated or otherwise discontinued, what we end up is monism (just one sole unit), which is the real opposite of pluralism. Plainly, then, if we insist on considering pluralism as an opposite to populism, we assign it the same conceptual meaning with monism and thus end up with a false synonymy.


Although not all concepts can be made operational (think, for instance, of beauty, love, virtue, or morality), clear conceptualization should ideally be accompanied by some kind of operationalization, that is, stating “the conditions, indeed the operations, by means of which a concept can be verified and, ultimately, measured” (Sartori 1970: 1045). Such operations necessitate the use of clear indicators or, in Hempel’s own words (1952: 41), “criteria of application couched in terms of observational or experimental procedure.” Concept operationalization should take the form: let x be what can be defined (and verified or falsified) via the indicators a, b, and c (Sartori 2009: 89). The problem is that the feasibility, let alone validity, of operationalization is (p.31) inversely proportional to the number of characteristics, or properties (i.e. the intension), inherent in our concept definition.

This alerts us further in two respects. First, the more its properties, the larger the number of operations we must undergo to verify the concept. Second, we must perform our operations with all, not only some, properties. It is probably because of such difficulties that operationalizing populism is still underdeveloped. Most efforts towards operationalization concern specific variables of the concept (the nature of its ideology, its discursive style, the characteristics of its leadership) rather than the concept itself. Such efforts include, among others, the distinction between “thin” and “thick” ideology (Freeden 1996, Stanley 2008), the content analysis of various populist manifestations, whether verbal or textual (Jagers and Walgrave 2007, Hawkins 2009, Pauwels 2011), and the status of populist leaders on the basis of various indicators (Levitsky and Loxton 2013).

A final note in this section is in order. Since modern populism, like any other political concept, is anchored in specific historical contexts and bound by distinct political and cultural modalities, it is important that our efforts to operationalize it must be attentive to precisely those contexts and modalities as they, and they alone, may provide the conditions and facilitate the necessary micro-mechanisms of populism to emerge and develop.

1.3. The Minimal Definition

The search for an apt definition of populism is not new and, sure enough, the difficulties associated with this enterprise are not small in number. As early as the very first conference on comparative populism held in 1967 at the LSE, Isaiah Berlin, serving as chairman of the concluding section, spoke to the rest of participants with these words:

Now we must address ourselves to the biting of the sour apple, a difficult part of our proceedings, which is the attempt to formulate some kind of model or definition or formula into which we can fit all the various types and nuances of populism which have been discussed; or, if we think that we cannot do it, to give reasons for our failure to do so, which might be equally fruitful.

(Berlin 1967: 114)

As shown in the previous sections, the failure to agree on a common understanding of modern populism is related to our mixing up, and studying together, different kinds of populism, which in turn has caused several methodological pitfalls. As we are now confronted with the steady increase of populism in the liberal democratic world and with an ever-increasing production of studies on populism in the academic world, we simply cannot afford, as Taguieff (1995: 25) implicitly suggests, to abandon attempts at a (p.32) rigorous definition. Now more than ever, we need a minimal (i.e., per genus et differentiam) definition of populism in the hope that, ultimately, we will be able to both classify our empirical cases and compare them in theoretically meaningful ways. This section undertakes the task to offer such a minimal definition of populism, which, by addressing the previous conceptual and methodological shortcomings, will make our concept theoretically more “general” and at the same time better able to cover greater empirical mileage.

A minimal definition is one that includes only the core, or constant, properties of the concept’s referents while excluding the variable (or accompanying, or contingent) ones (Sartori 1984: 79). Since the core properties should alone be able to bound the concept extensionally, any variable properties are to be “treated as the focus of empirical investigation rather than as a matter of definition” (Collier and Gerring 2009: 5). Ideally, as Gerring (2001: 78) remarks, a minimal definition should be “perfectly substitutable”—which is to say that one ought to be able to replace “populism” in any sentence with the minimal definition with no loss of intended meaning.

That minimal definitions as just described are difficult to achieve hardly needs stressing. As Kurt Weyland (2001: 4) concedes, “[i]t is often difficult to identify characteristics that are necessary and jointly sufficient in classifying a case as an instance of a phenomenon and to distinguish those definitional attributes from background conditions, causes, functional requirements, and consequences, as Sartori demands. … The particularly confusing concept of populism provides a best case for this skeptical position.” Keeping such difficulties in mind, here is another piece of expert advice to follow while engaging in reconceptualization:

Make sure that the definiens of a concept is adequate and parsimonious: adequate in that it contains enough characteristics to identify the referents and their boundaries; parsimonious in that no accompanying property is included among the necessary, defining properties.

(Sartori 1984: 56)

But this would hardly be the end of our difficulties. “Concepts,” writes Goertz (2006: 27), “are about ontology. To develop a concept is more than providing a definition; it is deciding what is important about an entity.” A concept’s ontology is intimately related to causality. Good concepts have causal theories and hypotheses embedded in them, which is to say that they “must also possess theoretical, or systematic, import; i.e., they must permit the establishment of explanatory and predictive principles in the form of general laws and theories” (Hempel 1952: 46). It is from those theories and hypotheses that we set out to understand how a concept presents in the real world, interacts with its environment and, through various mechanisms, which vie for our explanation, causes change.

So, what is populism in the context of our contemporary liberal democratic world? Under this innocuous formulation we are compelled to treat populism (p.33) as an object concept (Sartori 1984: 84), that is, a concept used to identify distinct, albeit quite diverse, referents. On that account, and stripped to its essentials, populism could well be defined as the idea that political sovereignty belongs to and should be exercised by “the people” without regard to institutions. Obviously, the key term in the foregoing definition is “the people”—the meaning of which may differ from one place to another, and thus can only become clear after empirical and comparative research. Whatever the case, the foregoing conceptualization of populism entails four essential attributes of “the people”: (a) its potential to form a political majority; (b) its allegedly homogeneous, oversoul nature; (c) its subservience to impersonal institutions; and (d) its belief of holding the moral right. All these attributes draw from early republicanism and, while not extraneous to the democratic principle, are fundamentally inimical to basic precepts of postwar liberal democracy, which demand (aa) preventing the tyranny of majority; (bb) promoting social plurality; (cc) protecting checks and balances by elevating them above individual interest; and (dd) advocating the rationality of ends rather than the morality of means.

That being so, and by taking our reconceptualizing effort only one little but brave step further, we may eventually achieve a truly minimal, as well as perfectly substitutable, definition as follows (where =Df stands for equivalently defined as):

Populism =Df Democratic illiberalism

This definition characterizes a whole class of political phenomena that develop in liberal democratic systems (including but not limited to individual leaders; movements and parties, whether in opposition or in government; party systems; political systems; ideologies or simpler sets of ideas; political discourses, rhetorical forms, and symbolic actions; and political strategies and policy packages) whose members combine just two characteristics: democraticness and illiberalism. Accordingly, the terms “populism” and “democratic illiberalism” will be used in this book interchangeably in our efforts to study modern populism—they carry exactly the same meaning and denote exactly the same thing.

Such a novel conceptualization of modern populism qua democratic illiberalism offers major methodological advantages since it adequately sorts out each and every of the shortcomings identified in the literature about populism.

More specifically, first, at the highest level of generality our minimal definition points to two different genera, democracy and illiberalism, which, however, may be combined perfectly well, as shown in Figure 1.1. Since parliamentary democracy can be differentiated into liberal (say, the Netherlands) and illiberal (say, Hungary), while illiberalism can be differentiated into democratic (say, Hungary) and nondemocratic (say, Francoist Spain), (p.34) populism is the integrated result of “illiberal democracy” and “democratic illiberalism.”10

What is Populism?

Figure 1.1. Populism per genus et differentiam

Second, by reducing the concept’s properties to bare bones, we both avoid the essentialist trap and achieve a clear grasp of what constitutes the ontology of populism. A concept’s ontology rests on its core characteristics, or properties, which alone constitute what the phenomenon actually is. In this sense, our definition is indeed a minimal one in that it contains only two necessary properties—democraticness and illiberalism—which, when combined, yield the constitutive—i.e., necessary and sufficient—dimensions of modern populism, that is to say, the concept’s very intension.

Third, precisely as a result of having established its intension, our concept is also well bounded extensionally with regard to the ensemble of referents to which it applies. In operational terms (shown in Table 1.1), only phenomena that score positively in both democraticness and illiberalism (1,1) may be classified as populist (e.g., Hungary). In contrast, the absence of illiberalism (p.35) (1,0) makes for a referent that is liberal (e.g., the Netherlands) while the absence of democracy (0,1) signifies a nondemocratic referent (e.g., Francoist Spain). Regarding the absence of both dimensions (0,0), this is simply a contradiction in terms, a theoretical absurdity.11

Table 1.1. Populism’s necessary-and-sufficient-conditions concept structure




















Fourth, perhaps even more crucially, our minimal definition of modern populism points directly to its two negative poles, that is, political liberalism and nondemocratic autocracy. These distinctions establish what modern populism is not, at the same time providing us with a clear dichotomous view of our object: Populism is always democratic but never liberal. Such a conception points to two clear cleavage lines that may open up in modern politics and which are essential for further understanding the populist phenomenon: One cleavage dividing democratic from nondemocratic forces (which effectively pits liberals and populists jointly against autocratic nondemocrats), and another dividing liberal from broadly illiberal forces (which pits liberals against populists and nondemocrats jointly). I will return shortly to these distinctions (see Chapter 2.3) and use them to provide a novel, robust classification of parties, party systems, and regimes in which populism is present.

Fifth, the two core attributes of newly conceptualized populism bear a very traceable relation to sets of specific indicators, and therefore lend themselves to operationalization and empirical testing. Democraticness has two indicator variables—electoral contestation and constitutional legality; illiberalism involves three such variables—singular cleavage, adversarial politics, and majoritarianism (in contrast to liberalism’s acceptance of plural cleavages, the pursuit of political moderation, and the protection of minority rights). When it comes to empirical analysis, then, if any of the indicator-level variables has value zero then its core dimension is also zero. For the latter to take a positive value, then, it is required that all indicators score positively.

So far, having established what populism is (democratic illiberalism) and what it is not (either political liberalism or autocracy), our concept has already become specific enough to attempt its empirical measurement and to try to comparatively assess its degrees in time and space. We are also fit at this point to code our populism-related variables in contradistinction to the variables related to either liberalism or democraticness, and make them the basis for quantitative analysis.

(p.36) 1.4. The Overall Framework of Analysis

We need an overall classificatory and typological scheme that will allow us to seize the bewildering complexity and fluidity of contemporary liberal world politics, and place populist parties, or movements, which are the subject matter of this book, squarely into that maze. Such a scheme, besides of course its methodological firmness and analytical parsimony, should satisfy two research aims: First, to serve a mapping purpose, in a way that will convey useful and theoretically relevant information about populism in contemporary politics; and, second, to serve an explanatory purpose, so as to provide us with causally relevant insights for a more detailed empirical and comparative science. The framework presented in Table 1.2 intends to satisfy these aims and will be used as a comprehensive guide for analysis in the rest of the volume.

Table 1.2. The overall framework of analysis for populist parties and their (core and variable) properties

What is Populism?

The table is based on, and begins causally with two variables (column A), democraticness and liberalism, which, depending on how they amalgamate, may constitute either liberal democracy or modern populism. Democraticness is herein treated in a binary way: A political party is democratic if and as long as it meets certain criteria, and it is nondemocratic if it fails to meet them. Therefore, unless a party is examined during a transitory state of its development, there can be no halfway situation (i.e., democratic hybrids).12 When it comes to liberalism, the distinction is binary, too, but it also allows significant gradation for the illiberal parties since they may range from pro-democratic to antidemocratic and even to nondemocratic (in which latter case we have already entered the realm of autocratic politics).

From variables, our next causally logical step brings us to the indicators (column B in Table 1.2) that determine whether democraticness and illiberalism obtain so that a populist party may classify as such. Each indicator is broad enough but still easily operationalized and empirically observed in a number of meaningful ways. With respect to democraticness, it obtains when a party decides to contest non-violent or non-voter-intimidating elections and is able to—vocally or otherwise—demonstrate its allegiance to parliamentary democracy. In contrast, democraticness does not exist in systems with no electoral contestation. At single party level, it is declared null and void when the party (p.37) (p.38) openly declares its opposition to parliamentarism. With respect to liberalism, it obtains when a party observes all the indicators that amount to it, namely, the acknowledgment of a plurality of divisions, and hence interests, in society; a clear preference for political moderation and consensual politics; and adherence to the rule of law, including the protection of minority rights. Inversely, we are in the domain of illiberalism when a party presents society as an antagonistic duel between “the people” and the “elites”; opts for polarization while shunning political compromise; and is willing to curb the law and violate minority rights to serve majoritarianism.

Evidently, in close interaction with the indicators, our two principle variables take different values along which individual political parties may now be located according to their democraticness (or its lack) in combination with their liberalism (or illiberalism). From variables and indicators, we can now move on to classification proper (column C in Table 1.2), that is, in our case, the ordering of political parties, or movements, in classes that, at least in ideal-typical form, can be said to be mutually exclusive and jointly exhaustive.

In this way, the ordering that results from the democraticness criterion, first, yields three broad classes of parties:

  1. 1. Parties that contest competitive elections and are loyal to the democratic rules of game.

  2. 2. Parties that contest competitive elections but are disloyal to democratic politics, openly declaring their opposition to it.

  3. 3. Parties in nondemocratic systems that preclude contestation.

The ordering that results from the liberalism-to-illiberalism criterion, produces four additional classes of parties:

  1. 4. Liberal parties, those meeting all the criteria set for liberalism.

  2. 5. Illiberal parties that contest competitive elections and are loyal to democracy.

  3. 6. Illiberal parties that contest competitive elections but are disloyal to democracy.

  4. 7. Illiberal parties in nondemocratic systems; most often these have a monopoly on state violence and use political repression.

From classification we can then move to typology (column D in Table 1.2), which is a rather more complicated matter since, unlike classification, it “is an ordering of ‘attributive compounds,’ i.e., an ordering resulting from more than one criterion” (Sartori 1976: 125). By combining the core attributes of democraticness and liberalism, we come up with a simple but exhaustive fourfold typology of political parties for modern-day politics. Two of these types, liberal and populist, contain democratic parties, which are, however, (p.39) distinguished by their different stances towards political liberalism. The other two types, antidemocratic and nondemocratic, contain illiberal parties that are also inimical to parliamentary democracy. Their difference is that, while antidemocratic parties still contest democratic elections in the hope of overthrowing parliamentarism, nondemocratic ones belong to autocracies. Populist parties are clearly distinguished in our typology from both antidemocratic illiberal and nondemocratic parties, but also from democratic liberal ones. Notice, in passing, that the requirement of variable operationalization holds as one proceeds across the columns of the chart and examines the typology resulting from the classification.

From this point on, our analysis of populist parties parts ways with other party types, and their empirical referents, and becomes specific enough and specifically focused on modern populism, which, according to our minimal definition, is the result of only two constant properties—democraticness and illiberalism (column E in Table 1.2). Just put them together and you get populism.

A host of other secondary or variable properties (column F in Table 1.2) that also characterize populist parties—such as their distinct mobilization patterns; the strategic use of polarization, populist discourse, and particular communication styles; their susceptibility to patronage politics; the importance of charismatic leadership; the contagious effects they may have especially on liberal parties, and many others—also emerge in variable form and must be fruitfully analyzed at the empirical level for a better understanding of the populist phenomenon.

To recapitulate and conclude this short but crucial section, we have constructed a solid analytical framework that recommends itself on the grounds of being readily intelligible, theoretically parsimonious, and resting on variables that are easy to operationalize. The objection may be raised, perhaps, that, since the classes and types seem to occupy fixed locations, our framework is rather static and lacks process dynamics. This is, however, true only if one considers the individual party cases as static rather than dynamic political entities—which is anything but true. As noted earlier, parties are opportunistic organizations that develop in ever-shifting political contexts and for this reason are required to constantly adjust their strategic priorities and modify their political (and policy) positions. In short, they are expected to change positions along the two variable axes of democraticness and liberalism, and so move from one party-type box to another. This realization is particularly important for populist parties since they sit midway between liberalism and autocracy. The idea I want to convey is that, if a populist party loosens up its democratic component, it approaches autocracy; and that, vice versa, if it plays down its illiberal component, it approaches the liberal party ideal. This realization is of the utmost importance for politicians and policymakers alike. But more about that later in Chapter 7.


(1) For the true aficionados, a summary of the conference proceedings appeared under the title “To Define Populism,” in Government and Opposition 3 no. 2 (1968), 137–79, while a complete transcript of the proceedings is to be found in the library of the LSE under the title “London School of Economics Conference on Populism, May 20–21, 1967: Verbatim Report,” shelf mark HN 17 C74.

(2) The Fronte dell’ Uomo Qualunque was an ultra-conservative and monarchist movement with no confirmed leadership appealing predominantly to Italy’s southern peasants (see e.g., Tarchi 2003). It participated in the 1946 legislative elections (winning 5.3 percent of the national vote) only to be absorbed soon thereafter by the rising Christian Democrats.

(3) In 1953, Pierre Poujade, a small bookstore owner, created the Union for the Defense of Traders and Artisans (UDCA) to express the economic interests and social resentment of shopkeepers and other low- and middle-class sectors in French society facing rapid economic and social change (Hoffman 1956, Bouclier 2006, Souillac, 2007). Besides its opposition to state centralism, industrialism, urbanization, and political liberalism, UDCA stood against French decolonization in Africa, while, in internal politics, it presented a xenophobic and anti-Semitic platform. In the 1956 general elections, it won 12.6 percent of the total vote and fifty-two seats in the French parliament. The youngest of its deputies was Jean-Marie Le Pen. By 1958, when Charles de Gaulle inaugurated the French Fifth Republic, and to a large extent because of Poujade’s own “uncharismatic” leadership, Poujadisme had practically disappeared. Still, several of its ideas persisted in French society. Jean-Marie Le Pen also survived politically. Some years later he was to reappear in the French political scene as the founder and leader of the National Front.

(4) Brazil was the first country in Latin America to give women the ballot in 1932. Argentina introduced women’s suffrage in 1947, Chile in 1949, Colombia in 1954, Peru in 1955, and Bolivia in 1956. Illiterates were enfranchised in 1979 in Peru and only as late as 1988 in Brazil.

(5) Only in Colombia was the rise of populism thwarted with the assassination in 1948 of leftist populist presidential candidate Jorge Eliécer Gaitán, an event that provoked an almost decade-long, and particularly violent, civil war between the supporters of Colombia’s two major political forces, the Liberals and the Conservatives (Braun 1985).

(7) Team Trump: “Donald Trump’s Argument for America”: https://www.youtube.com/watch%3Fv=vST61W4bGm8.

(8) As famously put in a nutshell: “[t]he democratic method is that institutional arrangement for arriving at political decisions in which individual [leaders] acquire the power to decide by means of a competitive struggle for the people’s vote” (Shumpeter 1942: xxx).

(9) Note, in passing, the markedly anti-elitist element in pluralism. The Britannica Concise Encyclopedia defines “pluralism” as follows: “In political science, the view that in liberal democracies power is (or should be) dispersed among a variety of economic and ideological pressure groups and is not (or should not be) held by a single elite or group of elites” (cited in Plattner 2010: 89).

(10) Accordingly, if we stick with Orbán’s Hungary as our example, both of the following definitions refer to the same definiendum: (a) Hungarian populism is a democratic phenomenon which is illiberal or (b) Hungarian populism is an illiberal phenomenon also presenting the attributes of parliamentary democracy.

(11) Interestingly, and by some stretching of “liberalism,” one could claim that there is at least one country case that falls within this category: Singapore. It recognizes multiple cleavages in society, pursues moderate and consensual politics, and respects the rule of law (Cherian 2000). But since it provides no full protection of political liberties, Singapore is hardly democratic. In a sense, Singapore resembles what one author has dubbed “undemocratic liberalism,” or rights without democracy (Mounk 2018: 14).

(12) To be sure, parties may often switch from one category, or party type, to another. For instance, several modern-day democratic nativist parties, like the Austrian FPÖ, the French National Front, the Italian National Alliance (dissolved in 2009) or, more recently, the Hungarian Jobbik, have nondemocratic roots. The FPÖ originates in Austrian Nazism (its first party leader, Anton Reinthaller, was a former Nazi Minister of Agriculture and SS officer), the Action Française was monarchist, the National Alliance sprang out of Italian fascism, while Jobbik was initially distinguished by Hungarian anti-Semitism and a spirit reminiscent of interwar authoritarianism. For the opposite outcome, i.e., a formerly democratic party switching to the nondemocratic camp, both Venezuela’s United Socialist Party (PSUV) under President Maduro and Turkey’s Justice and Development Party (AKP) after the 2016 counter-coup of President Erdoğan are perfect examples of the trend.