Mapping Populism and Nationalism in Leader Rhetoric Across North America and Europe

Abstract

We conceptualize populism and nationalism as vertical and horizontal discursive frames of sovereignty, and we investigate the prevalence of these frames in the speeches of chief executives (presidents and prime ministers) in Europe and North America to assess whether these discourses are on the rise at the highest levels of government. To do so, we compile an original database of leader speeches, measuring both discourses using a technique called holistic grading. We find that neither populism nor nationalism is on the rise across Europe and North America over the past twenty years; instead, the rise is concentrated in sub-regions and specific countries. We also find that populism and nationalism are highly but imperfectly correlated in leaders’ speeches in the corpus as a whole, but that populism is far less common in the speeches of western leaders. In the penultimate section, we use a selection of speech vignettes to demonstrate that state leaders employ populism to counter political opponents, nationalism to counter hostile nations, and a combination to mobilize against conjoined threats from above and beyond the “people-nation.”

Introduction

In the early twenty-first century, mass politics have come to be seen as a threat to liberal democracy and the global liberal order. We are said to be living in a populist zeitgeist, meaning that the spirit of our times is characterized by the widespread belief that “the people” rather than “elites” should rule (Mudde 2004; Müller 2016). In Europe at least, mainstream parties have increasingly incorporated populist elites-versus-people messaging into their platforms (Mudde 2004; see also Mair 2002), and the past decade has given rise to a smattering of populist heads of state. This apparent shift toward populist rhetoric is thought to signal a decline in support for liberal democratic institutions as well as party-based representation. Others have argued that we are living in a nationalist zeitgeist, characterized by the widespread belief that the ethnonational majority should enjoy a privileged status in the state relative to ethnic minorities (Cederman 2019; Snyder 2019). Finally, a growing number of state leaders appear to be integrating nationalist and populist framing in their public addresses. In this conjoined ethnopopulist framing, the ingroup is the core ethnopolitical group and the outgroups are political elites who are conspiring with enemy nations to exploit the “people-nation.”¹

In view of this, social scientists have turned their attention to the coincidence of populism and nationalism in political communication (Bonikowski et al. 2018; Brubaker 2017, 2020; De Cleen 2017; De Cleen and Stavrakakis 2017, 2020; Blokker 2005). This hyper-exclusionary style of politics is understood to be on the ascendant in the industrialized West (Mudde 2007; Inglehart and Norris, 2019; Kriesi and Pappas 2015; Betz 1994). North America and Europe in particular are seen as major sites of a nationalist, “right-wing” variety of populism, as opposed to the “left-wing” populism of Latin America (Mudde and Kaltwasser 2013; Betz 1994; Rooduijn and Akkerman 2017; Ivarsflaten 2008 Kaltwasser and Taggart 2015; Oesch 2008; Mudde 2004).

That populism assumes an ethnonationalist cast in Europe is generally attributed to two things. First, exclusionary ethnic nationalism is far more common in Europe than the inclusionary variety, due in part to the greater role that linguistic nationalism played in the nineteenth century state-building process in Europe (Hobsbawm 1992). Second, exclusionary versions of nationalism are closely connected to narratives of demographic anxiety, which resonate strongly in European countries today due to their declining national birth rates and rising immigration and refugee flows from non-European countries (Offe 2016). These anxieties are believed to have combined with “democratic fatigue” (Rupnik 2007) and rising grievances due to globalization (Rodrik 2018; Kriesi and Pappas 2015) to create a groundswell of support in the west for exclusionary populism.

However, a closer look at party politics in Europe and North America complicates this picture. Although to a lesser degree than Latin America, these regions are also the site of left-wing, frequently non-nationalist populism, as seen in Spain’s Podemos movement (Gómez-Reino and Llamazares 2019), Italy’s Five-Star Movement (Ivaldi et al. 2017), the Syriza government in Greece (Katsambekis 2016; Stavrakakis and Katsambekis 2014), and Democratic presidential candidates such as Bernie Sanders in the USA (Hawkins and Littvay 2019).² Europe is also the site of non-populist nationalism, where state leaders inscribe an imagined sovereign nation united against external threats, as in the case of former UK Prime Minister David Cameron, former Austrian Chancellor Christian Kern, former Armenian President Robert Kocharyan, and Belarussian President Lukashenko’s first term in office in the 1990s.

Why do chief executives sometimes articulate nationalist and populist discourses together, but other times separately? More specifically, how frequently does nationalism accompany populism, and are these discourses on the rise in the rhetoric of state leaders in Europe and North America? To answer these questions, we first argue for recasting nationalism and populism as discursive frames of sovereignty that state elites use to inscribe the boundaries of the idealized community more restrictively. As frames of sovereignty, populist discourse is used to inscribe the demos more restrictively (excluding “elites” or “the establishment” from the good “people”), whereas nationalist discourse is used to inscribe the ethnos more restrictively (excluding non-national “others” from the “nation”). Having conceptualized these discursive frames, we develop coding rubrics using the holistic textual analysis created by Hawkins (2009) and work with a team of graduate students with expertise in the coded countries to collect and code four types of speeches delivered by the head of state of 30 European countries, with two North American countries included for comparison. The resulting Nationalism Populism Database contains quantitative measures of both nationalist and populist discourse across 128 leader terms.

The biggest surprise is that we find little support for the contention that either populism or nationalism is on the rise across Europe and North America; most of the increase can be found in selected countries and sub-regions. Furthermore, although populism is strongly and positively correlated with nationalism across leader speeches, this association is far stronger in Eastern Europe. In Western Europe and North America, there is only a weak association between populism and nationalism. We also find an aggregate secular decline in both nationalism and populism in the former Soviet Union, which may be due to a shift toward different strategies of legitimation by competitive authoritarian regimes. In the remainder of this article, we explain our conceptualizations, describe the research design, and present quantitative results of our dataset. We also use a selection of speech vignettes to demonstrate that state leaders employ populism to counter political opponents, nationalism to counter hostile nations, and a combination to mobilize against conjoined threats from above and beyond the “people-nation.”

Populism and Nationalism as Frames of Sovereignty

Studies of exclusionary or right-wing populism have traditionally viewed the “right wing” in programmatic or issue-centric terms; it is conceptualized as a combination of authoritarian disposition, conservatism on social issues, and xenophobia or anti-immigrant views (Mudde 2007). Some scholars have referred to this combination as “national populism,” because of how it invokes nationalism’s focus on the dominant ethnonational group (Eatwell and Goodwin 2018; Stanley 2011). These stances are seen as components of a traditional form of ideology—as discrete, manifest political ideas that can be translated into policy. At the mass level, they are usually measured through survey items asking respondents about their stance on issues. At the elite level, they are measured through expert surveys and sentence-level analyses of party manifestos (Marks, et al. 2006; Volkens et al. 2019).

While these are appropriate conceptual and methodological approaches for identifying the programmatic characteristics of contemporary party families, we think that they misidentify nationalism and populism in particular and fail to provide us with valid techniques for measuring their presence, especially in leader rhetoric. A more helpful approach is one informed by framing theory and discourse analysis, according to which nationalism and populism are orthogonal discursive frames³ used by political leaders to project a more restrictive image of the sovereign community in the political sphere. Frames are cognitive structures that “enable individuals to locate, perceive, identify and label occurrences” (Snow and Benford 1988, p. 464). During moments of interpretive ambiguity, movement entrepreneurs engage in “strategic framing” of an issue in order to “turn the heads” of movement participants (Snow and Benford 1992). Master frames are relatively fixed templates that contain a set of inner logics that can be applied across different political contexts (Snow and Benford 1992; see also Schank and Abelson).

Populism is an effective master frame because it is a “thin-centered ideology” that lacks significant ideological content, depth and ambition—it is not so much a set of issue positions as a way of interpreting how issues relate to broader questions about the sovereign community (Mudde and Kaltwasser 2013, p. 150). As testament to its flexibility, populism has been employed on both the left and the right—by democrats as well as authoritarians (Enyedi 2016). Nationalism, too, is best described as a modular form (Anderson 1983) or discursive practice (Brubaker 1996, 2006) than a fixed identity or philosophy of governance. When deployed at the rhetorical level, nationalism can be just as thin and flexible as populism, with little to say about key distributional questions in society (Freeden 1996). To paraphrase Barth (1969, p. 15), nationalist and populist frames are more important for inscribing boundaries of the idealized sovereign than they are for their policy content. Leaders everywhere avail themselves of these discursive technologies to reframe imagined sovereign boundaries across a variety of cultural and ideological contexts. They differ in that they inscribe different sovereign spaces. While populism is a vertical, “up-down” frame that calls for popular sovereignty, nationalism is a horizontal, “in–out” frame that demands national sovereignty (De Cleen 2017; De Cleen and Stavrakakis 2017, 2020; Jenne 2018, 2021; Heiskanen 2020).

So influential is the chief executive in these discursive performances that they function as “securitizing actors” whose speech produces a “discontinuous reconfiguration of a social state” by invoking threats to the core identity or ontological security of the sovereign community (Waever 2011, p. 468). As observed by Hawkins (2009, p. 1048), “populism is so often associated with the leaders who create and galvanize the movement.” Mols (2012, p. 332) likewise writes that “influential politicians are crafty ‘identity entrepreneurs’, whose social influence depends, not so much on their ability to frame issues, but on their ability to redefine the audience’s collective self-understanding.” In invoking these discursive frames, the speaker offers her audience a lens through which to understand a given political event or situation. On a practical level, the speaker “interpellates” or “transforms” concrete individuals or groups into “subjects” at play in her interpretive frame (Althusser 1971; see also De Cleen 2017 and Glynos 2008).

Although we do not test for this, we expect that chief executives are more likely to reframe the sovereign community in the wake of destabilizing events such as financial crises or international war that marginalize “normal” or status quo political narratives (Subotić 2016). Legro (2005) says that politicians adopt new ideas when exogenous shocks undermine the old orthodoxy and when competing frames appear as viable alternatives. This resembles a state of the world that Krebs (2014) dubbed “unsettled narratives,” a period during which the dominant sovereign frame is no longer credible to a large or influential portion of the population. At such times, pressure grows from the base for a leader who can articulate an adequate response to such crises by reinscribing political space—expanding it during periods of confidence and restricting it during periods of crisis.

Populism and Internal Sovereignty

Populism comprises a set of arguments about threats to the demos or internal sovereignty.⁴ “The people” is understood here as having a homogeneous general will; it is the job of the leader to discern this will and ensure that it is consistently translated into policy (Mudde 2004; Canovan 1999). There is an implied underdog dimensionality to “the people” (Moffitt 2016; Ostiguy 2009), although the “oppressed people” is often the middle class (Brubaker 2020). Either way, the underlying logic of populism is revolutionary; the imperative is to achieve popular sovereignty by restoring sovereignty to “the people.”

“The people” is conceptualized in different ways depending on the speaker. Non-nationalist populists tend to frame the demos in economic or class-based terms. Examples include Bernie Sanders’ 2016 campaign in which “the people” were defined as the “99 percent” of Americans who were not in the millionaire or billionaire class that corrupts democracy. For Hugo Chávez, “the people” were the poor and indigenous majorities of Venezuela. Intriguingly, some left-wing populists, such as supporters of the Democracy in Europe Movement 2015 (DiEM25), reject national boundaries altogether—their idealized demos is explicitly transnational (De Cleen et al. 2019) . What counts as “elites” also varies by speaker and political context. As a rule, however, they are the “power-holders,” including government bureaucrats, members of opposition parties, hedge fund managers, multinational corporations, the EU, the IMF, journalists, intellectuals or the judiciary (Roodujin 2014).

Following Jansen (2011), we argue that populism serves as a master frame to mobilize popular support for the leader’s fight against the establishment, both foreign and domestic. Populism is used to target one’s political opponents, and it is more likely to resonate in a climate of mounting mass grievances when people are engaged in motivating searches for scapegoats on which to blame their woes. According to populist discourse, “elites” must be excluded from the demos, along with the institutions and foreign backers that permit them to exercise their illegitimate influence. This exclusion is based on the populist premise that “‘corrupt elites’ have unjustly usurped the sovereign authority of the ‘noble People’” (Aslanidis, 2015, p. 12). For populists, the return of control of the sovereign from elites to the people is the most urgent task of state leaders, sometimes requiring extra-legal means.

Nationalism and External Sovereignty

Nationalism, by contrast, is a set of arguments about threats to the ethnos. Ernest Gellner (1983, p. 1) defined nationalism as a principle that holds that communities of common descent have the right to territorial or external sovereignty. At the same time, nations have only a virtual existence because no single member can know the millions of others who are part of the same community. Benedict Anderson (1991, p. 7) wrote that nations were merely “imagined” communities and imagined specifically as horizontal sovereign communities with “finite, if elastic, boundaries beyond which lie other nations.”

Nationalism has played a key role in the process of building, integrating and consolidating modern states (Gellner 1983; Greenfeld 1992; Hobsbawm 1992; Wimmer 2013) as well as justifying state-led campaigns of irredentism (Saideman 1998, 2001). In these projects, the ethnos, the imagined national core of the state, can be inscribed broadly enough to include not just the dominant nation, but also national minorities and even foreigners—each having equivalent status in the state under a kind of civic or multicultural identity.⁵ Conversely, it may be inscribed so narrowly as to include only the dominant ethnonational group, downgrading ethnic minorities who are citizens and sometimes excluding foreigners (for a discussion of different models of ethnonational dominance across states, see Wimmer 2004, 2018; Mylonas 2012).

By reshaping or reinforcing collective understandings of the nation, nationalist frames provide a blueprint for policy action. Besides elevating the “national self,” they call for the identification of threats to the nation both at home and abroad. The nation may be framed as under siege by enemy nations, meaning that one must unify the national community to a fight against hostile nations and outwit or outperform competitor nations. International conflict, in this view, is understood as the natural outgrowth of national competition: Russians against the Poles, Turks against Greeks, Japanese against Koreans (Billig 1995). Finally, nationalists can use nativist frames against immigrants or other internal groups deemed to be hostile to, or undermining, the integrity of the nation.

Ethnopopulism and “Dual” Sovereignty

The fact that populism and nationalism are articulated around different sovereign concerns does not rule out their combination—quite the opposite. Numerous scholars have remarked on the extent to which the two are “stitched together” or “co-articulated” in what we call ethnopopulism, where the speaker inscribes the demos as coterminous with the ethnos.⁶ Here, the authentic sovereign is the core ethnopolitical community—the “people-nation” (De Cleen 2017) concentrated in the idealized heartland (Taggart 2000). Hints of this can be seen in ethnopopulist slogans such as “Real Americans” and “True Finns.” Ethnopopulists argue that the people-nation are threatened by “national others” such as immigrants, as well as “political enemies” such as opposition groups and their foreign allies. In the ethnopopulist view, elites are fellow members of the core ethnonational community who have, for selfish reasons, allowed enemy ethnicities to infiltrate and endanger the pure nation. They have betrayed both their democratic duties as the people’s representatives and their filial duties to protect the national community.

As Canovan (1984, p. 323) suggests, ethnopopulists imagine conspiracies everywhere. The threat to the people-nation comes not only from hostile national “others,” but also from political opponents, national “others” and rival nations, who are purported to be working together to subvert the dominant ethnopolitical group. Here, one is reminded of Nazi Party conspiracy theories that the Social Democrats had secretly sold out to the Great Powers and international Jewish financiers, or contemporary radical right theories that leaders of traditional parties are in cahoots with the Eurocrats in Brussels or the Troika. Similar claims about international linkages are also common among left-wing populists, who draw from Marxist rhetoric about capitalist imperialism and (the same) international financial institutions, although these claims are not always cast in nationalist terms. The key element for any form of populism is the demonization of a domestic elite, internal to the country’s demos. For ethnopopulists, this elite is also internal to (and hostile to) the ethnos.

For particularly embattled state leaders, ethnopopulism offers an ideological justification for aggressive, even eliminationist strategies to defend the “people-nation.” If hostile nations are believed to have aligned with one’s domestic political enemies, then disarming both becomes a matter of national emergency. Fifth column fantasies and conspiracy theories aid in securing popular support for discriminatory policies that serve the leader’s personal or political agenda.

Measuring Populism and Nationalism

To test our ideas about the coexistence of these two discourses and whether they have become more prevalent over time, we measured the populist and nationalist discourse of heads of state in 30 European and North American countries over the period 1998–2018. Specifically, we analyzed speeches by the chief executive (prime minister or president; both in semi-presidential systems) for every term in office longer than one year. We selected these regions because we want to see how much populist and nationalist discourses are co-articulated, following Bonikowski et al. (2018), in the part of the world that boasts the heaviest concentration of ethnopopulist leaders. We were also constrained by the availability of coders who speak each language at our universities. Despite these limitations, the result is a large database covering 128 leader terms.

If nationalism and populism function as sovereign discursive frames rather than programmatic ideologies, they must be measured using different techniques than those normally used to measure issue positions. Political discourses are embedded in texts as latent, diffuse sets of meaning, and are best captured by reading words and phrases in context, usually in longer passages in the original language. Hence, for both discourses we employ a method known as holistic grading, a technique of content analysis drawn from educational psychology in which coders read a text and assigning a grade to the entire document, based on a comparison with anchor texts exemplifying key values in a scoring rubric (White 1985; Sudweeks et al. 2004) .

Our coding was conducted based on a holistic grading technique developed by Hawkins (2009) to measure populism. Under the populist rubric, a speech counts as populist if it divides the political field into a good people and an evil establishment, the latter understood as actively subverting the interests of the people. Based on this criterion, the coder classifies a speech into one of three grades: 0, if not all elements of populism are present; 1, if all necessary elements are present but not consistently used; and 2, if all necessary elements are present and used consistently throughout the text. Coders then provide another three pieces of information: (1) a decimal grade for the speech, where 0.5 rounds up to a categorical 1, and 1.5 rounds up to a categorical 2; (2) representative quotes to justify the grade selected, categorized according to the aspect of populist discourse the quote represents; and (3) a summary paragraph justifying their grade.⁷

The nationalism rubric was designed along the lines of the populism rubric. Its content draws heavily on Billig (1995), who argued that nationalist rhetoric divides the political field into a binary frame of “homeland” and “foreigners” (the “national self” and “national others”), which are understood to be locked in a zero-sum competition. Coders score each speech separately for the glorification or elevation of the “national self” and the defense against, separation from or competition with “national others.” Although the two usually go together, we wanted to ascertain whether some nationalist discourse has a greater focus on the “self” than on the “other.” The rubric therefore asks coders to code each speech separately for these two components. The two scores are then added up to form a composite score, which like populism ranges from 0 to 2 (see Online Appendix 1 and 2 for the full rubrics).

Using these rubrics, we trained 58 graduate students on the discursive structure of populism and nationalism. The coders were taught the concepts of populism and nationalism and practiced coding speeches.⁸ We aimed to recruit at least two students per country who could code in the original language to ensure intercoder reliability as well as context-sensitivity of our measures. During the training, coders practiced on a set of anchor texts: speeches selected by the project leaders to represent the cut-points in the scales. All anchor texts were in English, and the scores were discussed extensively with the coders.⁹ Coders then searched for speeches from the leaders they were assigned to code following the sampling criteria discussed in the next paragraph. During the coding itself, there were weekly meetings between each country team and project leaders to discuss the scores given to each speech.

The unit of analysis is one president or prime minister term. For each term, we locate four speeches.¹⁰ Hawkins (2009) shows that using four speeches per term gives a reliable approximation of populist discourse in a single leader term, and we expect the same to apply for nationalism. The four speeches are a quota sample drawn using the sampling criteria in Hawkins (2009): one famous, one international, one from a ribbon-cutting event, and one from an electoral campaign (for a detailed description of each kind, see Online Appendix 3). The four types provide a glimpse of politicians’ discourse under different circumstances with different audiences; random samples draw too heavily from the most frequent speeches, which are not always the most important or the most well-known. The list of countries and leader terms can be found in Table 3 of the Appendix.

Findings

Our quantitative results indicate that neither populism nor nationalism is particularly salient in the speeches of European and North American leaders. Furthermore, while nationalism coincides with populism in a few expected cases (enhancing our confidence in the new rubric), ethnopopulism is mostly found in Eastern Europe.

We first consider the separate results for each discourse. The average level of populism in leader speeches is 0.37 (on the 0–2 scale), meaning that the use of this type of frame is rare among chief executives across Europe and North America.¹¹ These numbers confirm previous findings (Hawkins 2009, 2010; Rooduijn and Pauwels 2011). The second panel in Fig. 1 breaks the average down by type of speech. Campaign speeches score the highest, averaging 0.62. Since scoring a 0.5 (which rounds up to a 1) requires that a speech feature all three elements of populism, this means that the average campaign speech is moderately populist. By contrast, chief executives are least likely to use populist discourse in their international and ribbon-cutting speeches. The bars indicate 95% confidence intervals.

Fig. 1

Average levels of populism and nationalism across types of speech

For nationalism, the distribution is similar, with an overall average of 0.35. Famous—rather than campaign—speeches tend to be the most nationalist. We think this makes sense: famous speeches are often delivered on the anniversaries of national holidays or as state-of-the-union addresses that aim to unify the national community. By contrast, campaign speeches are more likely to focus on partisan competition using populist rhetoric. Unsurprisingly, international speeches have the lowest level of both nationalist and populist discourse, as state leaders are more likely to deliver speeches to foreign audiences that stress common global interests while downplaying national and political divisions.

Table 1 lists the ten most populist politicians in our sample. The highest is former Slovak President Vladimír Mečiar, with an average of 1.67 during his term in the 1990s. Next we see Belarus’ Alexander Lukashenko, for the second of his five terms in office, together with former Moldovan President Vladimir Voronin, followed closely by current Turkish President Recep T. Erdoğan. At the bottom of the list, although with averages still near the middle of the scale, are former Czech Prime Minister Mirek Topolánek, former Czech President Václav Klaus and current Hungarian Prime Minister Viktor Orbán. Results for Orbán may seem surprising, since he is usually hailed as an example of populism in CEE; however, most of the leaders above Orbán are from non-EU countries, which means that he still stands out from other contemporary EU leaders. This might explain why Orbán’s populism draws much more attention from media and academics.

Table 1 Ten most populist leaders

For comparison, Table 2 lists the ten most nationalist leaders in Europe and North America, based on a sum of national “self” and “other” dimensions. We observe some overlap with the most populist leaders, and also some expected results that give face validity to our measures. Erdoğan and Lukashenko once again figure in the list, with high levels of praising the national self and excluding national others. Orbán moves up the list noticeably, suggesting that nationalism may be more prevalent in his rhetoric. Together, these three leaders constitute the clearest cases of ethnopopulism in the dataset. The table also shows that, although highly correlated, the two dimensions of nationalism do not always go together. Ukraine’s Petro Poroshenko, for instance, is one of the highest overall in referring to the national self, but a round 0 when it comes to excluding national others. For most other leaders, however, the two dimensions of nationalism appear more balanced. Table 3 in the Appendix contains the full list of leaders’ nationalism and populism scores.

Table 2 Ten most nationalist leaders

The Macro-level View

We now explore the correlations between the dimensions of nationalism, and between populism and nationalism, across our corpus of speeches. The first panel in Fig. 2 shows that the two dimensions of nationalism itself are highly correlated, at Pearson’s r = 0.61, as expected by Billig (1995), who argues that national identity is two-sided—it defines who is in the nation as well as those who are out—and that we rarely see references to the national self without references to national others. That said, the correlation is imperfect, as leaders tend to emphasize the national self with greater frequency than national others.

Fig. 2

Correlations between self and other nationalism; populism and nationalism

The second panel shows the correlation between nationalism and populism at Pearson’s r = 0.49. Although not as high as the correlation between the two dimensions of nationalism, it confirms the conventional observation that populist and nationalist frames are often articulated together, particularly in the industrialized West. The three leaders mentioned already—Lukashenko, Erdoğan, and Orbán—are the clearest examples, but if we search a bit lower on both scales, we find for example Jarosław Kaczyński from Poland, with 0.90 on nationalism and 0.75 on populism. Still, ethnopopulism is not quite the norm. Some speeches are highly populist but non-nationalist; an even greater number are highly nationalist but non-populist. For example, the dot in the lower-right corner in the second panel of Fig. 2 represents Slovakia’s Vladimír Mečiar. Although he is the most populist leader in this sample, he has a mean nationalism score of just 0.13. Similar leaders include Václav Klaus and Mirek Topolánek, both from the Czech Republic. Leaders in the upper-left quadrant (high on nationalism but low on populism) include Cameron of the UK, Stephen Harper of Canada, two former presidents of Armenia and Serbia (Kocharyan and Aleksandar Vučić, respectively), as well as the third term of Nikola Gruevski as Prime Minister of what is now North Macedonia.

To assess the aggregate prevalence of populism and nationalism in speeches over time, we display the results as a time series in Figs. 3 and 4. Let us first consider overall trends. Figure 3 indicates that the presence of these discourses is fairly stable over time, with a small uptick in nationalism and a concomitant slight decline in populism over the past few years. Thus, talk of the rise of populism and nationalism across Europe is somewhat exaggerated, at least in chief executive rhetoric. However, when we break these trends down by sub-region in Fig. 4, we see that the overall trend masks important differences. Among post-Soviet states, there has been an overall decline in both nationalism and populism from the early 2000s to present. Given the pattern of democratic backsliding in post-Soviet states, this may suggest that the leaders of authoritarian states rely less on antagonistic mobilization than on repression or cooptation to remain in power (Gerschewski 2013). By contrast, the sub-regions where we see the greatest increase in nationalist and populism are Central and Eastern Europe and Southern Europe. Populism increases or is high early in the period, while nationalism has ticked up slightly since 2014, perhaps reflecting the nativist upsurge since the 2014–15 refugee crisis in Europe. In Western Europe and North America, not only are levels of populism relatively low, but the uptick in nationalism (still at low levels) is not matched by a similar increase in populism.

Fig. 3

Yearly averages of populism and nationalism

Fig. 4

Yearly averages of populism and nationalism by sub-region

We can only speculate as to the reasons for these sub-regional differences. On its face, the pattern seems to conform to the expectations of Kohn (1944), who drew a sharp line between the virulent nationalisms of the East and the more calm and reasonable nationalisms of the West. However, these differences may also reflect the fragility of sovereign boundaries of post-communist European states, which are still only a few decades post-independence. Some states experienced internecine civil war and boundary disputes, and this framing is reflected in leader rhetoric in Georgia, Ukraine, Moldova, and Armenia. Likewise, the fact that Western leaders are so much less likely to use populist frames may reflect how highly embedded their societies are in the international liberal order and how much more institutionalized and effective their states are at providing public goods (Hawkins et al. 2019a, b). Along these lines, it is worth noting that in Southern Europe, populism is much more prevalent than in the other sub-regions. We think the stronger presence of (mostly left-wing) populist leaders in these countries reflects the relatively weak quality of governance and, more recently, the deeper impact of the Great Recession.

The Micro-level View

To get a sense of how these frames function at the micro-level, we present a selection of populist, nationalist and ethnopopulist speech vignettes from our speech corpus. These not only reveal the different ways in which in- and outgroups are constructed in each frame of sovereignty, but also how each frame is articulated across different political contexts and with what immediate effects.

Populist Speech Vignettes

The first two vignettes were selected to illustrate the function of populist frames in leaders’ speeches. These are leaders whose four speeches in a given term averaged 1 or higher on the populist rubric, but less than 0.5 on the nationalism rubric. What these populist speeches have in common is an injunction to protect the regime and thus the demos from self-dealing elites, who are sometimes given to be colluding with external enablers. To combat these shady “forces,” the leader justifies placing restrictions on the media, political parties and civil society organizations, which have the effect of preserving or enhancing the leader’s discretionary power (Batory 2016).

The first vignette comes from Slovak Prime Minister Vladimír Mečiar in 1994, who had just risen to power on the heels of the 1993 Czech-Slovak Velvet Divorce, which he had helped orchestrate. In one speech, Mečiar railed against the previous Slovak leader, Ján Čarnogúrský, who had purportedly committed abuses of power against “the people.” He further questioned “the right to leadership for those who have been failing for years for the sake of people who have been failing for years in programs, who have built up the whole program only on hatred, disowning people, lying, provoking affairs, afraid, many unsubstantiated raids.” Finally, he accused the opposition of “burdening other political circles and clusters in Europe” with “pseudo-problems hiding their own incompetence and waiting for foreign representations or foreign media, or foreign business sphere to come to Slovakia to win the election for them.” Slovakia had need of “only one” Movement for a Democratic Slovakia (HZDS), according to Mečiar, which was “better suited to historical tasks” of running the country than other parties due to its “inner life” and “internal democracy.”

To understand the utility of the populist master frame in this context, it is important to know that Mečiar himself had profited immensely from the privatization process, which had been tilted heavily toward him and members of his inner circle. He was now confronted with a political opposition both at home and abroad who feared that Slovakia was pivoting away from democracy and the West. Using populist “up-down” framing, Mečiar excluded the political opposition from the authentic demos on the grounds that they had made accommodations to foreign businesses and powers at the expense of ordinary Slovaks. His speeches served to justify authoritarian policies aimed at his political enemies. Political opponents were interpellated as “elites” who were out to destroy his party—the true representative of the authentic Slovak people.

The second vignette is a famous speech by Moldovan President Vladimir Voronin in 2009. The speech was delivered in the context of an upsurge of political protests following his refusal to step down from the presidency after his election loss. The government responded with mass arrests as well as a media and internet blackout of police actions.¹² In his speech, Voronin offered a defense of these actions:

There are forces in the country that are capable of sacrificing democratic institutions, respect for the law, the European option of the Republic of Moldova and statehood for the sake of the ambitions to accede to power. These forces, in a few days, managed to throw the country back into an atmosphere of fear and mistrust of the early 90’s...I am extremely concerned that hundreds of young citizens are forced to be under arrest, to pay with their freedom for the mistakes that are on the conscience of politicians.

Using populist framing, Voronin thus redefined the demos more restrictively to exclude his political opponents, who were interpellated as illegitimate “forces” and “politicians” who had ginned up the protests for private gain. The resulting violence was thus “on the conscience” of the opposition politicians rather than his own regime. In this way, Voronin offered a post hoc justification for both the government crackdown and subsequent suppression of free media.

Nationalist Speech Vignettes

Nationalist rhetoric, on the other hand, calls for elevating or protecting the ethnos from enemy or rival nations or national “others.” In non-populist nationalism, there is little to no mention of evil or exploitative elites. To explore these expectations, we offer vignettes of speeches that scored at least 0.5 on nationalism but less than 0.5 on populism, making them “purely nationalist” speeches.

The first comes from Armenian President Serzh Sargsyan, who delivered a famous speech to “the People of Armenia and all Armenians.” The context of the speech was, oddly enough, the normalization of diplomatic relations between Armenia and Turkey, which had been frozen since the start of the Nagorno-Karabakh war in 1993. While agreeing to a process of protocols with Turkey, the president publicly aired his misgivings about his Turkish counterparts. Using nationalist framing, Sargsyan projected an idealized Armenian ethnos that must unite against an enemy nation. In his speech, Sargsyan declared that “any achievement of our people did not come easy. We have always succeeded due to our will and unity. This is the formula for our victories.” The problem, in his view, was that the nation was “trying to normalize relations with a country that had carried out policies of deportation and extermination of our people…during the Ottoman Empire.” In this nationalist “in–out” framing, all Armenians were united against all Turks.

The second nationalist vignette considered here is Canadian Prime Minister Stephen Harper’s ribbon-cutting speech, which he delivered in August 2012 on the site of Operation NANOOK 12—an annual operation in the Arctic aimed at “exercising Canadian sovereignty” in emergency response scenarios. In his speech, Harper explicitly linked Canadian sovereignty with the Arctic:

Our sovereignty, our presence and our ability to project that presence everywhere we place our flag, that is where you come in...You are part of the determined expansion that we are making to our sovereign presence in the Arctic, through expansion of the Canadian Rangers across the region… it has become Canada’s destiny to protect a large portion of our planet’s North….Yet, we also remain determined to assert our national interest and to protect our sovereignty in these lands.

Through nationalist framing, Harper projected an image of a unified Canadian nation determined to take a more assertive stance vis-à-vis rivals in the Arctic. This master frame aligned with a policy shift toward beefing up the Canadian military and aggressively asserting Canadian interests abroad since the Conservative Party took power in 2006. Although sovereignty concerns had also been present under the previous administration, Harper’s approach was widely “defined as unilateral, assertive and at times militaristic.”¹³

It is notable that neither of these nationalist governments adopted a transgressive approach toward international institutions, nor did they vilify their domestic political opponents. This is in line with the function of purely nationalist frames—they help to justify a leader’s defensive, even aggressive posture in the region, including militaristic stances toward “rival nations,” against which the entire nation must be unified.

Ethnopopulist Speech Vignettes

We hypothesized earlier that ethnopopulist frames configure “political elites” and national “others” as joint threats to the sovereignty of the people-nation, with the combination being especially pernicious to democratic institutions and ethnic minorities (Zellman 2019). To show what this looks like on the micro-level, a final pair of vignettes were drawn from leader terms scoring above 0.5 on both nationalism and populism rubrics—Hungary’s Viktor Orbán in his third and fourth terms and Turkey’s Recep Tayyip Erdoğan in his fourth (presidential) term.

The first ethnopopulist vignette comes from Orbán’s famous speech, which he delivered in his second term at a Hungarian language camp in 2014 in the Romanian town of Băile Tuşnad (Tusnádfürdő in Hungarian), an ethnically Hungarian town that had been part of pre-WWI Hungary. The speech was entitled “The Era of the Work-based State is Coming.” In it, Orbán intoned:

How beautiful it would be, how noble a form of revenge, if the political forces who voted against the re-engaging of the Hungarians living outside the borders of Hungary would be deservedly punished if a majority, or even a two-thirds majority, were gained by the votes of the Hungarians who live outside the borders of Hungary.

In this speech, Orbán not only equates the Hungarian demos with the ethnos, but also identifies the threat to the people-nation (“political forces”) as disloyal members of the Hungarian nation. Using ethnopopulist framing, Orbán thus interpellated his political opponents in Budapest and their “foreign backers” as an elite alliance against Fidesz supporters both at home and in the diaspora:

If we would like to reorganize our nation state instead of the liberal state, we should make it clear that these are not civilians coming against us, opposing us, but political activists attempting to promote foreign interests.

Orbán elaborated on the threat posed by “political activists” later in an international speech before the European parliament, in which he claimed that “George Soros and his NGOs want to transport one million migrants to the EU per year. He has personally publicly announced this program and provides a financial loan for it.” Orbán’s policies followed the prescriptive logic of ethnopopulist framing. To realign political and national borders, the government enacted dual citizenship and voting rights to the Hungarian diaspora soon after returning to office in 2010 (Pogonyi 2017). The government also dramatically increased funding for diaspora foundations, schools and businesses for Hungarians abroad to ensure that these new voters would be Fidesz voters. Meanwhile, in the name of fighting against liberal elites and their backers, the government placed restrictions and a new tax on foreign-funded NGOs in Hungary. In a final step, the government linked the threat of “invading” Syrian refugees with the threat of liberal elites, and in 2018 passed a law that effectively criminalized both the act of seeking asylum in Hungary as well as any act by individuals or organizations aimed at assisting asylum seekers (Waterbury 2020).

Our second ethnopopulist vignette comes from Turkish President and Justice and Development Party (AKP) leader Recep Tayyip Erdoğan, who delivered a famous speech at an “anti-coup rally” in August 2016, just weeks after the failed military coup by a disaffected faction of the Turkish Armed Forces. At a celebratory rally held in the Yenikapi meeting area, Erdoğan declared,

The enemies of our nation became exhausted in the morning of July 16^th seeing this unity [of people]. This view [of unity] is a sign that the ones who dare to capture a small piece of land from our country will pay a great cost. This nation does not accept enslavement.

In Erdoğan’s ethnopopulist framing, the “people-nation” was threatened from within by political opponents, who were backed by foreign elements, namely US-based cleric Fethullah Gülen and its political network:

We must analyze the July 15 coup attempt very well. We should not only evaluate those who have done this betrayal, but also the powers behind them and the motives that motivate them...Of course we have to expose [the Fethullah terror] organization to all its members and destroy it in the law. But if we only live with it, we will defend ourselves as a state and a nation against similar viruses and leave weak points in defence.

Erdoğan thus equated the ethnos with the demos, framing the political opposition as a threat to his ethnopolitical base. In this framing, the Turkish people-nation were beset by the hostile Gülenist political elites and a global network of foreign backers, including western powers. The “enemies” of the regime therefore had to be “cleansed.” Following the prescriptive logic of the frame, the government arrested and detained tens of thousands of civil servants, including university professors and journalists, on the basis of the thinnest of Gülenist associations. Media freedoms were severely curtailed, and dozens of magazines, radio stations and newspapers were shut down. The government later expanded its purge to the international level, using highly unconventional means (such as cancelling passports) to induce Gülenists living abroad to return home to face charges of terrorism (Öktem and Akkoyunlu 2016).

These vignettes suggest a core set of commonalities in ethnopopulist framing at the leadership level. In both the Hungarian and Turkish cases, the political opposition was interpellated or “hailed” as elites that served as a fifth column for hostile foreign interests. Ethnopopulist discourse appears to assist embattled leaders ensure their survival by justifying the otherwise extreme step of eliminating political rivals who are responsible for weakening the ethnonational character of the state.

Discussion and Conclusion

Several important findings emerged from this holistic study of leaders’ sovereign discursive frames. First, and most notably, there has been no aggregate secular increase in the prevalence of either nationalism or populism in the political communication of state executive speeches in our database. This is consistent with the findings of Bieber (2018a) on nationalism and the findings of Hawkins et al. (2019a, b) on populism. There has been a slight increase in aggregate nationalism since 2015 in Europe, possibly in response to the refugee crisis in Europe. However, our data suggest that the use of these frames in chief executive speeches does not appear to have increased significantly in the early 2000s, at least in government.

Second, there are distinctive sub-regional patterns to leader rhetoric in our study. Populism and nationalism are highly correlated in chief executive speeches in Eastern Europe, but not in Western Europe. It is also notable that very few Western leaders in our database scored above 0.5 on populism, perhaps reflecting the relatively high quality of democratic governance and socialization into national and EU institutions (although see post-Brexit UK leaders for exceptions). The strength of ethnopopulism in Eastern Europe may have something to do with the fragility of state institutions in post-communist Europe, along with widespread perceptions within Eastern societies that they are subjected to Western or EU dominance. While populism and nationalism are also highly correlated in the speeches of leaders of states in the former Soviet Union, the aggregate level of both populism and nationalism appears to have slightly decreased since 2000, which may be due to the declining utility of exclusionary mobilizational frames for authoritarian leaders who have come to depend on other sources of regime legitimation (Levitsky and Loxton 2013; Gerschewski 2013).

Third, our exploration of nationalist, populist and ethnopopulist case vignettes offers tentative confirmation that elites use populist frames to battle political opponents and/or foreign power-holders such as the EU, the UN or major powers, whereas purely nationalist frames are used only to unify the nation against the threat of rival nations. This nationalist framing is neither very revisionist nor partisan.

Ethnopopulist frames, by contrast, appear to be uniquely destructive to the body politic. By elevating public perceptions of threat posed by certain societal classes and legitimizing grand conspiracies between elites and national others, ethnopopulism lays the groundwork for policies of eliminationism with devastating effectiveness. This can be seen in Erdoğan’s campaign against “Gülenists,” the campaigns of Orbán and Gruevski to expel “Soros mercenaries” from their countries, and Lukashenko’s repression of “scumbags” and “roughnecks” whom he claimed enjoyed the backing of western intelligence organizations and other “foreign interests” aiming to subvert the will of the Belarussian people.

Our analysis suggests at least two avenues of future research. First, scholars should further test whether nationalist frames have the negative policy effects we have hypothesized, especially when articulated together with populism. We expect that nationalism is associated with bilateral conflict, the tightening of borders, and possibly higher military spending. It is also worth exploring whether the co-articulation of populism with nationalism is associated with eliminationist policies like expelling or discriminating against entire political or ethnic groups who are perceived as disloyal citizens.

Second, much work remains to be done on sub-regional and temporal trends in sovereign frames in the rhetoric of world leaders; text analysis might be used in combination with network analysis to map the pathways by which each frame spreads within particular regions. The diffusion of elite-level ethnopopulist framing should be of particular concern. To generate accurate early warning mechanisms, analysts might use machine learning to score leader speeches for ethnopopulism in order to forecast the adoption of exclusionary policies against both minorities and the political opposition.

Appendix

Table 3 Full list of leader terms and scores