This is a critical time for the study of populism. After several decades of languishing, the study of populism has revived. The emergence of new populist politicians, parties, and movements in several regions—Chavismo and its allies in Latin America, the Tea Party and Occupy movements in the United States, the populist radical right in Europe—has attracted the attention of a large number of scholars. In returning to this topic, they have brought to bear innovative methods and a comparative focus, both of which were scarcely imagined by those who first studied it in the 1960s. The initial studies of populism were largely conceptual and descriptive, debating how to define populism and describing the populist parties and movements of their day, as well as a few historical cases (Ionescu and Gellner 1969; Roberts 1995). Today, a growing number of scholars across disciplines and regions agrees on a common ideational definition (Mudde and Rovira Kaltwasser 2013a). Furthermore, a number of studies purport to measure populism across countries and time periods, using related techniques of textual analysis at the elite level (Armony and Armony 2005; Hawkins 2009; Jagers and Walgrave 2007; Rooduijn and Pauwels 2011; Rooduijn et al 2012) and survey research at the mass level (Akkerman et al 2014; Hawkins at al 2012; Stanley 2011), not to mention through traditional qualitative techniques (de la Torre 2010; Mudde 2007).
However, scholars have yet to seriously tackle the next most basic question in the study of populism: what causes it? So far, there has been little comparative research that could offer a generalized answer to this question, and many existing theories are at odds. We propose a collaborative project to answer the question “What causes populism?” This framework is designed to catalyze the kind of iterative work that is more characteristic of productive individual research in the field, one in which scholars repeatedly test competing theories. The collaboration is multi-stage, seeking to produce an edited volume in the first two years and a series of big-data projects in the subsequent 5 years. We aim to secure funding for the second stage through an Horizons 2020 grant and an NSF grant, for which we aim to develop proposals during the ECPR Research Sessions.
The Project’s Contribution
The point of departure for the project is an ideational (discursive, ideological) definition of populism as a Manichaean discourse that sees politics as a struggle between a reified “will of the people” and a conspiring elite (Hawkins 2009; Mudde 2004). This minimal definition allows actual populist movements to assume different ideologies and issue positions, with radical leftist and radical right versions being the most common. Indeed, we agree with the current view in some of the literature which sees populism as a set of ideas that always requires a host ideology (Stanley 2008; Mudde and Rovira Kaltwasser 2013b). Given this, our main question is to explain what makes varying levels of populism appear on the political stage of any country.
Although attempts at explaining the causes of populism are still rare, there are several emerging lines of argument. These can be categorized roughly as macro-level explanations, which come from political sociology and political science; and micro-level explanations, which typically come from psychology or communications studies. While macro-level explanations are often surprisingly ideographic, focusing on current global trends, micro-level explanations are nearly always framed in broad, nomethetic terms. And while macro-level explanations rely on surveys and party- or country-level comparisons to test their arguments, micro-level explanations tend to rely on surveys and experiments.
At the macro-level, explanations began with early structuralist work on Latin America that explained populism through a sociological lens and saw populist movements as a response to delayed modernization (Cardoso 1972; Di Tella 1965; Germani 1978; Ianni 1975; Weffort 1978). A stream of newer studies, with a focus on particular regional manifestations of populism and also with a sociological cornerstone, has emerged. For Europe, this work has emphasized the transformations wrought by economic globalization, which has left working class groups in Europe open to the appeals of radical right parties, most of which match a populist discourse with anti-immigrant appeals and a celebration of traditional family and religious values (Bornschier and Kriesi 2012; Kitschelt 1994; Kriesi et al. 2012). For Latin America, scholars have emphasized the role of economic crisis and the worsening of economic inequality after the onset of market-oriented reforms in the 1980s and a lack of political representation on the left during the experience of market reform, in generating support for populist alternatives (Roberts forthcoming). Others point to the weakness of democratic governance in developing countries, where corruption is widespread and systematic, which provides fertile ground for populist movements (de la Torre 2010; Hawkins 2010). In addition, there are also macro-level institutional arguments that focus on the ability of formal institutions or democratic consolidation to mediate the effects of external forces (globalization, economic crisis, weak governance) and prevent the easy emergence of third-party populist options (Mudde and Rovira Kaltwasser 2012).
In contrast, micro-level arguments tend to be more generalizable and focus less on the broad context that feeds the demand for populist parties, and more on the attitudes of voters and the cognitive mechanisms that allow these attitudes to become activated or mobilized. Most work focuses on the framing effects of populist discourse (Bos et al 2013). Related work from political psychology suggests that traits such as authoritarian disposition, dogmatism, and need for closure, are often pre-existing and similarly distributed across most populations. What matters is how these latent dispositions are activated. Most studies suggest that the key trigger is a perceived normative threat to the community (Feldman 2003; Hetherington and Weiler 2009; Stenner 2005). According to this view, the language or frame matters less than the material reality of current issues, which are only partly susceptible to reframing. Some initial support for these arguments as applied to populism comes from studies using survey data (Hawkins et al 2012).
Current research therefore, offers a number of obvious and interesting possibilities for explaining populism. However, a general theory of populism has yet to be developed. The development of such a theory has been hampered by at least three factors. First, there has been little comparative research across regions that could lead to a general explanation for the resurgence of populism in recent years. Most studies at the individual level take place in just one country, and even the most ambitious studies of sociological theories focus on a single region, e.g., Western Europe or Latin America. Cross-regional studies are rare (c.f. Mudde and Rovira Kaltwasser 2012).
Secondly, many of the theories that have been presented here focus on explaining the success of particular kinds of populists, where the emphasis is on the ancillary, thick ideologies that define these party families, rather than on their populism per se. In the literature on the populist radical right, for example, a heavy emphasis is placed on the role played by these parties’ anti-immigrant stances and by voters xenophobic attitudes (Bornschier and Kriesi 2012; Kitschelt 1997; Mudde 2007). However, these factors are unlikely to explain the rise of social populist parties like Die Linke or the SP in Europe or that of radical left populist leaders in Latin America.
Finally, most studies that are empirically ambitious focus on just one or a few theories and use only one method of analysis (Hawkins 2010). Scholars who work at different levels of analysis rarely refer to each other’s work or attempt to speak to one another’s findings. And it is not unusual for volumes to test just one preferred theory without regard for alternative hypotheses.
In short, there is both a greater need and an opportunity for data collection and collaboration. This can help us develop a proper comparative theory of populism for the first time. This is the purpose of and contribution of this project.
Our project seeks to answer the question of what causes populism through a common theoretical framework that simultaneously test theories across regions and levels of analysis. Based on current research, we will work from the following general model:
• Most citizens of democracies already have strong populist attitudes. The discourse of populism is already present in the minds of citizens everywhere, a product of democratic values and personality traits, shaped somewhat by later socialization.
• These attitudes lie dormant until they are activated by political context. The context consists of the framing used by politicians and activists, but also the nature of the issues at hand.
• Political leadership, especially charismatic leadership, catalyzes successful populist movements. Aroused citizens will not support a party/movement unless they trust that its leaders sincerely embody populist discourse. Moreover, citizens know that success depends on having a large, coordinated movement. Charismatic leaders resolve these problems by offering a focal point for organization into which individual citizens can project their individual wills.
• The ideological flavor depends on the socioeconomic context. The host ideology of successful populist parties and movements varies independent of their level of populist discourse, but generally responds to the level of economic development and the size of the largest social class. Advanced industrial democracies feature large propertied classes that are more concerned about maintaining or recovering a favorable social order, while developing countries feature poor, propertyless classes that favor economic redistribution.
Together, these elements suggest an approach that combines methodological tools. We argue for simultaneously testing these propositions at the individual level, through surveys and experiments, and at the aggregate level, through studies of parties and party systems.
Methods and data
The project is oriented around work packages rather than a set of country cases. The work packages are organized around the project’s goals: 1) developing a comprehensive and coherent theory that can explain the rise of populism; 2) developing instruments that can measure the presence of populism in, for example, in party documents, elite discourse or citizens’ attitudes; 3) testing explanations for the rise of populism. Thus, the work packages define a series of teams, which organize around theory formation and particular methods of data collection and analysis. For this reason, the inclusion of the country partners is mostly dictated by 1) their knowledge of and experience with researching populism; 2) their methodological expertise; 3) their ability to develop and test our measurement instruments in the country in which they are based; and 4) funding requirements, as the EU will require representatives from all sub-regions of Europe.
We do not discount the need for country based case studies. However, it is our ambition to work truly comparatively and thus first collect comparative data using the same measurement instrument on in all countries in the selected regions. To achieve this objective, the structure of the project is not country-based, but expertise-based. Of course, coordinating and executing all of these efforts at data collection is an ambitious goal, especially in such linguistically and geographically diverse set of countries However, the state of the art in the field makes it possible and individual members of the project team have already developed instruments on the basis of case studies that could potentially be used in a comparative context (see for example Akkerman et al 2014; Rooduijn at al 2015 and Hawkins 2009).
The work packages or research teams are the following:
The development of a true theory of populism is the eventual goal of the project. To this end, relevant theories from different disciplines will be integrated. However, the project is build on the idea that theorizing cannot be done in isolation from data and cases. Theories will therefore also be developed on the basis of existing and new studies of particular parties and movements.
Team content analysis
One of the most important goals, and also one of the most ambitious ones of the project is to measure populism. In order to establish how populist parties, politicians, movements and leaders are the level of populism in texts needs to be measured. Although many content analysis techniques are available (e.g. computerized content analysis, human-coding), it needs to be investigated which techniques produce reliable and valid populism measures and which texts are most relevant to analyse (e.g. party programs, speeches, websites)
Team expert surveys
Although expert surveys are commonly used to measure party positions, they have not yet been used extensively to measure how populist parties or movements are. Thus, in the context of the project expert surveys form an alternative way of measuring populism. Moreover, they can also be used for benchmarking purposes, that is, to assess the reliability and validity of measures created by the context analyses.
Team mass surveys
Mass surveys will be the projects most important means for measuring citizen attitudes. Especially the development of survey questions that measure the extent to which citizens adhere to populist ideas and the extent to which they support populist parties and movements is crucial for the project. Moreover, through mass surveys micro-level support for these parties and movements can also be explained. Multiple regional surveys exist and the goal is to develop reliable and valid items that can be included in these surveys.
Team parliamentary surveys
In addition to mass surveys, parliamentary surveys are an important means to measure the extent to which parliamentarian have populist attitudes. Moreover, in the parliamentary structured surveys supply-side explanations for the success of populist parties can be explored as well.
Experiments will be an essential innovative tool for testing causal mechanisms at the individual level. Online and lab-based experiments will be used to assess how individual-level populist tendencies are activated, for example through framing. The use of experiments in the field of populism studies is relatively new, and the challenge is to develop good treatments that can be used in multiple countries.
Team semi-structured elite interviews
Semi-structured elite interviews will be an important way of testing supply-side explanations for the success of populist parties and movements. In the interviews, focus will be on the role of leaders in creating demand for populism. As in the other teams, the focus is on the development of a semi-structured questionnaire that can be administered in the different regions.
Team database management
It is the project’s ambition to make the collected data publicly available, for example through an interactive website. This is mostly likely to generate a lasting product for various teams and the effort will require an expensive and extensive infrastructure (e.g. to make all collected documents from populist parties available or enable others to replicate the project’s findings).
Milestones and Research Steps
The project consists of two distinct phases. The first phase, which is comprised of two conferences leading towards an edited volume, is a preparatory one. Each collaborator has joined one or more of the functionally, methodologically oriented teams and is currently developing a paper outlining methodological challenges or presenting the results of pilots in which potential measurement instruments have been tested. These papers will be presented at the workshops, in which initial methodological and theoretical questions will be addressed and a truly collaborative, long-term scholarly network will be build. The first conference, to be held in London on 30 April – 2 May 2015 at the London Centre of Brigham Young University, will give all collaborators the chance to finally meet each other and to present initial ideas and preliminary results. At this workshop the contours of our funding proposals for the second phase (see below) will also be discussed. The second conference, to be held in at Brigham Young University in Utah on 7 – 9 January 2016, will enable the collaborators to present the results of our research and finalize funding proposals. The revised papers will be published in an edited volume to will be completed in 2016.
The second phase consists of the development of a series of large grant proposals for the EU Horizons 2020 program and the U.S. National Science Foundation in order to conduct cross-regional studies using multiple methods. Participation in the ECPR Research Sessions is an important step to achieve this goal. The team leaders will draft the proposals during the sessions, which are to be submitted to the funding organizations in 2016.
Team leaders attending:
Cristobal Rovira Kaltwasser (Universidad Diego Portales) - theory
Dave Doyle (Oxford University) – database management
Kirk Hawkins (Brigham Young University) – content analysis
Levi Littvay (Central European University) – mass survey
Nina Wiesehomeier (University of Swansea) – expert survey
Ryan Carlin (Georgia State University) – experiments
Sarah de Lange (University of Amsterdam) – interviews
Saskia Ruth (University of Zurich) – parliamentary survey
The Research Project and the ECPR
The research project is embedded with the ECPR network. The project team had its first meeting at the ECPR General Conference 2014 in Glasgow and will meet again at the ECPR General Conference 2015 in Montreal. Moreover, its proposal to participate in the ECPR Research Sessions 2015 in Nijmegen is supported by the ECPR Standing Group on Extremism and Democracy and the ECPR Standing Group on Latin American Politics. The members of the project come from a wide range of ECPR member institutions in countries such as Austria, Belgium, Greece, Hungary, Italy, the Netherlands, Portugal, Serbia, Spain, Switzerland, United Kingdom. In addition, the project also includes members from various Latin American countries and the United States.