Populism: Key Dimensions and Debates

Definitions of populism, and identification of its key dimensions, are nearly as varied as the phenomenon itself. Indeed, scholars have variously described populism as a category of political party or movement, as a distinct ideology, as a political or organizational strategy, as a characteristic of leaders or leadership styles, and as a type of political communication or performance (Diehl, Reference Diehl2022). From among these many vantage points, three principal approaches have prevailed.

A first approach sees populism as a “thin-centered” ideology consisting of two main claims: that the interests of a pure and disadvantaged people are confronted by those of an illegitimate and corrupt elite, and that “politics should be an expression of the volonté générale (general will) of the people” (Mudde, Reference Mudde2004: 543). While it proclaims to share democracy’s emphasis on popular power, populism as a thin-centered ideology often threatens the democratic process by equating the people’s will with the will of the majority. However, populism’s precise democratic implications differ depending on the thick ideological constructs with which it is paired. Right-wing populisms are presumed to be especially susceptible to majoritarianism, because of the additional emphasis they place on threats posed by domestic or foreign others (Mudde & Rovira Kaltwasser, Reference Mudde and Rovira Kaltwasser2018). This expectation is borne out by studies of European party competition, which find that a large proportion of right-wing parties—such as the Rassemblement National (RN, previously the Front National) in France, the Alternative for Germany (AfD), and the UK Independence Party (U.K.I.P.) to name a few examples—exhibit high scores on the “authoritarianism” index and display a propensity toward anti-immigrant, ultra-nationalist and eurosceptic stances (Norris, Reference Norris2019).

A second approach conceives populism as an organizational strategy marked by an anti-establishment orientation and tendency to reject intermediate bodies and representative institutions, including existing parties and party systems, to forge direct, unmediated relationships with the people (Weyland Reference Weyland2001; Reference Weyland, Rovira Kaltwasser, Taggart, Ochoa Espejo and Ostiguy2017). When in office, leaders informed by a populist organizational strategy face dilemmas squaring this organizational orientation with their own institutionalization, which they often do by centering power in the executive and delegitimating other branches of the state, including the judiciary (Blokker, Reference Blokker2019). Although this approach is deeply rooted in the study of Latin American populisms (Barr, Reference Barr2019), it has been applied to other geographies. For instance, a key example in Europe can be found in Hungary, where Orban’s government drew on populist discourses to justify revising the constitution in ways that erode meaningful checks on the power of the executive (Bugaric and Kuhelj, Reference Bugaric and Kuhelj2018).

In this study, we subscribe to a third approach, treating populism as a communication style, which manifests in attempts to gain resonance through performative appeals to voters and other audiences (Jagers and Walgrave, Reference Jagers and Stefaan2007; Moffitt and Tormey, Reference Moffitt and Tormey2014). Although the nature of these appeals differs across cases, common features include the abundant use of frames that mobilize the people against the elites, expressions of “bad manners,” simplified explanations of complex phenomena, references to “us versus them,” promotion of “common sense” against expert knowledge and use of emotional or aggressive language (Moffitt, Reference Moffitt2016; Diehl, Reference Diehl2022).

Although presented as an alternative to the ideological approach, operationalizing populism as a communication style also builds, at least partially, on definitional dimensions offered by the literature on populism as a thin-centered ideology (Aslanidis, Reference Aslanidis2016: 92–93). These strategies vary, but a common approach is to focus on three key discourses: people-centrism, which consists of positive references to a people; “anti-elitism,” which highlights vertical discrepancies in the power of the people versus political, economic or cultural elites; and exclusivity, which highlights differences on a horizontal plane, distinguishing the people from the variously defined foreign and domestic others (Brubaker, Reference Brubaker2020; Jagers and Walgrave, Reference Jagers and Stefaan2007; Meijers and Zaslove, Reference Meijers and Zaslove2021; Zulianello et al., Reference Zulianello, Albertini and Ceccobelli2018).

Because it focuses on the contingent and shifting nature of political discourse, the stylistic approach to populism is less susceptible than competing perspectives to categorical uses of “populist” to label and typologize political leaders, parties and movements. The ideological approach is especially prone to such categorization “since it presupposes a ‘core’ of attributes which are ‘necessary and sufficient’ to delineate the concept” (Diehl, Reference Diehl2022: 21; see also Aslanidis, Reference Aslanidis2016: 92, 101). While they may have their own merits, these categorical distinctions among “populists” and “non-populists” are ineffective in recognizing that individual politicians—including those who are not traditionally considered populist (Brown et al., Reference Brown, Mondon and Winter2021)—may regularly deploy populist frames in a manner that is unstable and dependent on the political environment (Fahey, Reference Fahey2021). These findings underscore the advantages of theorizing populism not as a characteristic of parties, leaders and movements, but rather as a way of “doing politics” that varies by degree (Diehl, Reference Diehl2022; Meijers and Zaslove, Reference Meijers and Zaslove2021).

Despite appeals to avoid categorical treatments of populism, however, research remains susceptible to generalizing claims about what differentiates the populist style deployed by right- versus left-wing parties. In this regard, comparisons of European and Latin American cases have been influential. Based on systematic comparison of party-political populisms in Austria and France versus Bolivia and Venezuela, Mudde and Kaltwasser (Reference Mudde and Rovira Kaltwasser2013) reached a generalized conclusion: in Europe, a relatively high degree of economic development has enabled (primarily right-wing) politicians to foreground post-materialist issues of identity, resulting in discourses of exclusion, while in Latin America, economic disparity and poverty have resulted in primarily materialist manifestations of populism, with (primarily left-wing) parties mobilizing “inclusive” discourses to promote socio-economic redistribution. This continental comparison informs the general hypothesis that left-wing populist repertoires promote a dyadic antagonism among the people and elite, whereas right-wing populism is triadic and also includes attacks on others (March, Reference March2017: 284–85).

The kinds of people and elite imagined by right and left populisms are also presumed to differ. In most European scholarship, right-wing populism is characterized as focusing on the cultural threat of immigration and its implications for national sovereignty (Berezin, Reference Berezin2009; Rydgren, Reference Rydgren2005). The people in this formulation are an ethnic or cultural group denied access to the collective goods of the state by both minority others—who siphon resources and recognition—and elites—who support liberal (often secular) ideas that erode the cultural traditions of the majority. Left-wing populisms, on the other hand, are presumed to be primarily concerned with rectifying economic inequality. They portray the people as having been denied economic resources by primarily corporate elites responsible for free trade, globalization and Western imperialism (Bonikowski et al., Reference Bonikowski, Daphe Halikiopoulou and Rooduijn2019: 67).

Empirical studies of populism in Europe broadly support these hypotheses, revealing that right-wing politicians are more likely to target political elites and the media, while those on the left more often target corporate elites (Maurer and Diehl, Reference Maurer and Diehl2020). They further confirm that right-wing politicians are more likely to discuss, and exhibit negative sentiments about, immigration (Berezin, Reference Berezin2009; Maurer and Diehl, Reference Maurer and Diehl2020; Rydgren, Reference Rydgren2005), whereas parties that deploy populism in the service of left-wing ideological projects are more likely to adopt pluralistic definitions of the people (Meijers and Zaslove, Reference Meijers and Zaslove2021).

Observations of European cases thus render a series of expectations about populism’s key dimensions and relationship to ideology. First, right-wing politicians deploying people-centrism as a discursive style are expected to define the people in cultural terms, whereas their left-wing counterparts will address the people in primarily economic terms. Second, there is an expectation that right-wing parties engaged in anti-elitism will primarily target political and cultural elites, whereas left-wing parties will focus their criticisms on economic or corporate elites. Third, the European literature renders the expectation that exclusion of others is a built-in feature of right-wing populisms, whereas left-wing politicians will avoid this tendency in favour of a more inclusive discursive style.

There are several reasons to question the validity and reliability of these expectations. Parties and leaders deploying populist ideas, strategies and/or styles are often ambiguous about what they plan to do, making their expressed ideologies less predictive of their policy decisions once in office. Relatedly, parties classified as populist in the literature often strategically eschew left-right ideological boxes, in favour of grand narratives capable of attracting an array of constituencies (Norris, Reference Norris2019: 985). How generalizable, then, are expectations of right- versus left-wing populisms derived from the study of European cases? To what extent do observations of European populisms serve as a useful blueprint for conceptualizing and operationalizing populist political performances elsewhere?

In proposing to tackle these questions via an examination of politicians’ social media activity in Canada, our study contributes to research investigating the role of shifting digital landscapes in the rise of populism. Since the early days of the internet, scholars have predicted novel opportunities and challenges for democratic participation and debate (Bimber, Reference Bimber1998). In the political scene, studies have found that social media might contribute to a tendency towards the personalization of politics, where charismatic leadership is emphasized (Kriesi, Reference Kriesi2014; Zamir, Reference Zamir2024). Moreover, works on political communication analyzed the specificities of the populist phenomenon in the age of information technologies, scrutinized through concepts such as “social media populism” (Bobba, Reference Bobba2019), “populism 2.0” (Momoc, Reference Momoc2018) or “digital populism” (Bartlett et al., Reference Bartlett, Birdwell and Littler2011), among others. However they may define and operationalize populism, contemporary examinations of the phenomenon increasingly make use of the digital rhetoric and performances of politicians and parties to assess their populist tendencies. Analyzing the online discourses of Canada’s federal leaders through the lens of populism, our article joins this burgeoning research field.

Canadian Populisms: Still the Exception?

At the same time as the election of Donald Trump and victory of the “Brexit” campaign were drawing worldwide attention as examples of rising populism in 2016, Canada was being touted as an exception to populism’s surge. Indeed, as suggested by a 2016 Economist article entitled “The last liberals: Why Canada is still at ease with openness,” Canada was being praised as a stalwart of openness and inclusion in an era evidently marked by rising closure and exclusion (The Economist, 2016). Yet, historical evidence shows that populism has long been, and continues to be, a key feature in Canadian political culture, particularly at the regional level. In Alberta, for instance, populism has been a prominent vehicle for articulating “Western alienation” and demanding autonomy from the federal government (Wesley, Reference Wesley2011). In Saskatchewan, a stronger emphasis on collectivism has led populism to be, until recently, mainly a tool of the left, particularly the Cooperative Commonwealth Federation and its successor, the New Democratic Party (Wesley, Reference Wesley2011). The term “populist” has also been used to describe the rhetoric advanced by the Doug Ford government in Ontario (Budd, Reference Budd2020), to diagnose changes in Québec’s nationalist project (Tanguay, Reference Tanguay, Guillaume Dufour and Peker2023), and to describe the strategic efforts of pro-Anglophone parties and movements in New Brunswick (Chouinard and Gordon, Reference Chouinard, Gordon and Boily2021).

While certainly present in Canadian politics, there is an expectation that populism in Canada is less likely to feature nativist characterizations of the people for reasons tied to both demand and supply. On the demand side, desire for nativist populism is expected to be tempered by Canadians’ relatively strong and stable popular support for large-scale immigration, even in periods of relatively high unemployment (Ambrose and Mudde, Reference Ambrose and Mudde2015). Although the consensus around immigration has recently shown signs of fraying (Lundy, Reference Lundy2023), it was for many years bolstered by the emphasis on skills in Canada’s immigration system. Perhaps more important to Canadian exceptionalism, though, are the obstacles to populist mobilizing on the supply side. One such obstacle is the presence of a “first-past-the-post” electoral system, which benefits large parties with regional concentrations and discourages fringe parties’ success (Triadafilopoulos and Taylor, Reference Triadafilopoulos, Taylor, Samy and Duncan2021). Another important impediment to nativist populism is the role of linguistic and regional cleavages in limiting the resonance of pan-Canadian conceptions of the people (Gordon et al., Reference Gordon, Jeram and Van der Linden2020).

Yet, the exceptionalism of Canadian populism with respect to nativism is also worthy of reconsideration. Research suggests that while Canadians generally score lower on the nativism scale, populist attitudes are prevalent in the public, whether or not supply-side factors are present (Medeiros, Reference Medeiros2021). Moreover, implicit discourses of exclusion have long permeated Canadian politics. For instance, in the discourse of Conservative Prime Minister Stephen Harper (2006–2015), the boundaries of the people were purposely kept discursively ambiguous and a wedge was created between “good” and “bad” immigrants. This enabled the party to increase its vote share in immigrant communities and achieve a minimum winning coalition lasting two terms, while simultaneously limiting the scope of policies related to immigration and multiculturalism (Abu-Laban, Reference Abu-Laban and Jedwab2014). When such discourses and policies began to take a more explicit form—for instance, through references to “old stock Canadians,” plans to target “barbaric cultural practices,” attacks on “fraudulent refugee claims” and a proposed niqab ban in citizenship ceremonies—the Harper government lost its hold over the immigrant vote and was ultimately defeated (Carlaw, Reference Carlaw2017; Friesen, Reference Friesen2021; Kwak, Reference Kwak2020).

The exclusion of immigrant minorities has been a far more explicit aim of the PPC, which Maxime Bernier, a former Conservative cabinet minister, founded in 2018 following a failed bid—with a very close margin—to lead the Conservative Party of Canada. A comparison of the party’s website, public statements and other written materials with those of other right-wing parties by Peker and Winter (Reference Peker and Winter2024) shows that the PPC pledges to reduce immigration and end Canadian multiculturalism, thereby casting aside the “cult of diversity” established by Justin Trudeau. The party’s official materials also explicitly refer to Canada’s immigration policy as a mechanism to “forcibly change the cultural character and social fabric of our country” (cited in Peker and Winter, Reference Peker and Winter2024: 13). The right-wing populist political rhetoric of exclusion also increasingly targets feminists and 2SLGBTQ+ minorities. Erl’s (Reference Erl2021) study of election survey data showed that PPC supporters showed hostility to feminism and opposed so-called special privileges for women. The fight against radical gender ideology is also a key tenet of the PPC’s current platform (People’s Party of Canada, September 4, 2021). While the party’s relative success in the 2021 election might be more directly related to opposition to COVID-19-related restrictions, nativist and populist attitudes are also influential predictors for supporting the PPC (Medeiros and Gravelle, Reference Medeiros and Gravelle2023).

Contrary to the exceptionalism literature, therefore, populism has long been a force in Canadian politics, manifested in both right- and left-wing ideological projects. Moreover, there is evidence that exclusion of others features both implicitly and explicitly in Canadian political rhetoric, although parties engaged in explicit discourses of exclusion along nativist lines (for example, the PPC) have had limited success at the ballot box. To what extent are the hypotheses and categories elicited by the European experience capable of capturing the contours of Canada’s evolving populist zeitgeist? To answer this question, we employ a methodological approach that estimates the prevalence and intensity of three populist political discourses—“people-centrism,” “anti-elitism” and “exclusion of others”—in original tweets published by federal party leaders across the political spectrum.

Methodology

Researchers have devised several strategies to operationalize and measure populism. One approach uses expert surveys to estimate the presence of populist rhetoric within a political party, often in cross-national perspective (Meijers and Zaslove, Reference Meijers and Zaslove2021; Norris, Reference Norris2019, Reference Norris2020). In a prominent study based on this technique, Norris (Reference Norris2019) uses the Chapel Hill Expert Survey to compare populist rhetoric—operationalized as anticorruption and anti-elitism—across a wide range of European countries. The author then considers how populism interacts with other dimensions of politics, including cultural cleavages (authoritarianism, liberalism) and economic ones (left, right). While they provide a useful snapshot of the prevalence of select populist themes across parties, and are therefore beneficial to comparative research, studies based on expert surveys suffer significant drawbacks. Based as they are on experts’ holistic evaluations of parties, they tend to freeze those parties in time, thus precluding attention to populism’s continuous or contingent features. Expert-based research is also unlikely to consider the kinds of people and elites that parties construct to gain resonance with the electorate. This leaves unanswered questions about whether and how party ideology (left vs right) corresponds with populist political communication styles in different settings.

A second approach to measuring populism is based on holistic grading, wherein researchers devise coding schemes to rank whole documents (for example, speeches, manifestos or press releases), and occasionally paragraphs (Rooduijn and Pauwels, Reference Rooduijn and Pauwels2011; Rooduijn et al., Reference Rooduijn, De Lange and Van Der Brug2014), and grade them on a populism scale (Bernhard and Kriesi, Reference Bernhard and Hanspeter2019; Fahey, Reference Fahey2021; Hawkins, Reference Hawkins2009; Pauwels, Reference Pauwels2011). In Hawkins’s (Reference Hawkins2009) foundational application of the method, a document counts as populist if it divides the political field into good people versus the evil elite and understands the elite as subverting the peoples’ interests. A text is graded 0 if not all elements of populism are present, a 1 if all elements are present but not consistently used and a 2 if all elements are present and consistently used (Hawkins, Reference Hawkins2009).Footnote ² While the simplicity of this approach may be advantageous, holistic grading in its basic form precludes clear understanding of the ways that populism’s multiple dimensions interact (Meijers and Zaslove, Reference Meijers and Zaslove2021: 377).

A third approach to measuring populism, which we adopt in this study, uses social media data to ascertain the presence of populism as a communication style (Engesser et al., Reference Engesser, Ernst, Esser and Büchel2017; Ernst et al., Reference Ernst, Sina Blassnig, Büchel and Esser2019; Jagers and Walgrave, Reference Jagers and Stefaan2007; Maurer and Diehl, Reference Maurer and Diehl2020; Zulianello et al., Reference Zulianello, Albertini and Ceccobelli2018). Studies in this vein frequently consider how different dimensions of populism (such as people-centrism, anti-elitism and exclusion of others) interact with parties’ other characteristics (that is, right/left, challenger/incumbent, extreme/mainstream). The focus on social media is also advantageous, given its popularity as a tool for reaching large audiences in a relatively unmediated way. Social media algorithms are also known to funnel users to partisan information with which they are already likely to agree, thus strengthening the base of political movements (Pajnik, Reference Pajnik, Diehl and Bargetz2024). And, importantly when it comes to populism, social media are often perceived as a “voice for the underdog and the unrepresented,” providing means for “crowd building” among the politically disaffected (Gerbaudo, Reference Gerbaudo2018).

However, there are drawbacks to existing strategies for using social media to estimate populism. While types of elites are frequently given attention, rarely are the types of people invoked by political actors explored in these studies (for example, Jagers and Walgrave, Reference Jagers and Stefaan2007; Maurer and Diehl, Reference Maurer and Diehl2020). This is a significant oversight, given that different brands of populism identify the people’s interests in substantively different ways (Brubaker, Reference Brubaker2020). Similar blind spots exist with respect to understanding the exclusion of others. Perhaps because of a focus on the European experience, several studies single out and quantify negative references to immigrants in social media speech (for example, Maurer and Diehl, Reference Maurer and Diehl2020). However, the targeting of other minority groups—including 2SLGBTQ+ communities—to demarcate the people is rarely explored. Given evidence that exclusion of these communities is of growing significance to populism in North America (Stein, Reference Stein2023), this may lead to underestimating the exclusionary dimension of certain politicians’ populist rhetoric.

We aim to address these oversights in our quantitative analysis of 5,845 original tweets posted by federal party leaders on X (formerly Twitter) between January 1 and December 31, 2022. By that year, pandemic-related frustrations and public divisions occupied an increasingly prominent place in Canadian politics, which, by some accounts (McCullough, Reference McCullough2022), strengthened the role of parties, leaders and movements adopting right-wing populist discourses. Just a few months earlier, in the federal election of September 2021, the far-right PPC tripled its 2019 vote based on an anti-multiculturalist and anti-COVID mandate platform. Shortly thereafter, in early 2022, the Freedom Convoy opposing vaccines and other mandates drew international attention by shutting down the city of Ottawa for three weeks. Coinciding with these events was a marked spike in Google searches of the phrase “populism in Canada” (Peker et al., Reference Peker, Rémi and Laxer2023).

Our analysis focuses on the X activity of six federal party leaders. Justin Trudeau (1,607 tweets) was Prime Minister and leader of the centrist Liberal Party of Canada at the time of the study. Pierre Poilievre (1,541 tweets) announced his candidacy for the leadership of the right-wing Conservative Party of Canada at the height of the “Freedom Convoy” and was elected leader in September 2022. Jagmeet Singh (737 tweets) was leader of the left-wing New Democratic Party. Yves-François Blanchet (700 tweets) leads the Bloc Québécois, a party that advocates for Québec nationalism and interests at the federal level with a view to promoting Québec sovereignty, and only runs candidates in that province. Maxime Bernier (967 tweets) is leader of the far-right PPC, and Elizabeth May (293 tweets) led the Green Party of Canada in 2022. These tweets were collected using the X Application Programming Interface (API).Footnote ³ We focus on original tweets (we exclude retweets, for example) in order to prioritize content produced directly by the federal leaders.

Tweets were coded manually by a team of research assistants based on a codebook developed by the authors, which operationalizes the three main discourses associated with populism as a communication style: people-centrism, anti-elitism and exclusion of others. In a first round of coding, the exact terminology constituting each discourse was recorded. We then conducted a second round, grouping statements belonging to each discourse into common categories. We defined people-centrism as any statement referring positively to a group or community whose interests the leader in question claims to represent. This operationalization follows the approach adopted by Jaggers and Walgrave (Reference Jagers and Stefaan2007) and reproduced by others (Pauwels, Reference Pauwels2011; Maurer and Diehl, Reference Maurer and Diehl2020; Zulianello et al., Reference Zulianello, Albertini and Ceccobelli2018).Footnote ⁴ Depending on the characteristics assigned to the people, attributes of this variable were categorized as general, economic (that is, communities defined by their material interests), political (that is, communities defined in terms of their political/rights-based interests), cultural (that is, communities defined by their cultural interests) and other.Footnote ⁵ We defined “anti-elitism” as any statement referring negatively to an individual, organization or institution in a position of power, whose actions were deemed to undermine the interests of the people. Categories of this variable include elites who are political, economic, cultural, foreign and other.Footnote ⁶ Finally, we defined “exclusion of others” as any negative statement about a third out-group, besides the elite. Given our interest in estimating the value of hypotheses derived from European cases, where the focus is on immigrant out-groups, we prioritized negative statements addressing immigrants, and other ethno-cultural or linguistic minorities in our operationalization of others. We also demarcated negative statements pertaining to 2SLGBTQ+ communities, as they are an increasingly targeted identity group in right-wing populisms in North America (Stein, Reference Stein2023).Footnote ⁷

Based on these variables, we tested three hypotheses about populism’s key discourses and their relationships to party ideology.

To estimate the significance of each discourse in the leaders’ tweets, we adapted the index developed by Jagers and Walgrave (Reference Jagers and Stefaan2007). This index combines two measures: proportion (that is, the number of characters associated with each discourse divided by the total number of characters) and intensity (that is, the total number of times a discourse is mentioned in the dataset). Our index is a simple multiplication of proportion by intensity, divided by 100.

Two points about our sampling strategy should be noted. First, recognizing that there were significant discrepancies in the overall X activity of each leader during the period under study (with Trudeau being the most active, at 1,607 original tweets, and May being the least active, at 293 original tweets), we devised two different versions of the indexes of people-centrism, anti-elitism and exclusion of others. The first version involved measuring the significance of each variable controlling for differences in the overall X activity of the six leaders. To achieve this, we administered a weight to the intensity scores. Specifically, we multiplied the intensity scores by 700 (the number of tweets posted by Blanchet during the period examined), divided by the total number of tweets posted by each politician. For example, the weights applied to May, Singh and Poilievre were 700/293, 700/737 and 700/1541, respectively.Footnote ⁸ The second version seeks to capture actual differences in the content published by each party leader without controlling for overall X activity. In this series of analyses, we estimate the significance of each variable for the entire sample of tweets. Although it relies on unweighted data, this method allows us to ascertain the real magnitude of populist discourses put out by each leader on X. Below, we present results based on both the weighted and unweighted samples, allowing us to comparatively evaluate both sets of measures. Second, as Poilievre did not become party leader until more than halfway through the year, in September 2022, we ran the analyses separately to compare the results based on the full sample of tweets for the year 2022 to those based on the subperiod from September 10 to December 31, 2022, during which Poilievre became leader of the CPC. Finding no significant differences in the overall patterns among politicians between the two time periods, we proceeded to analyze results for the year as a whole.