The Use of Sovereignist Claims in Election Campaigns in France and Italy: Different Twins?

In this article we examine the phenomenon of the use of sovereignist claims, not only by nationalist or populist leaders but also by actors who would not normally fall into these categories. We zoom in on two different cases: France and Italy. Through an analysis of Twitter we examine the discourse of the political leadership in election campaigns. We document some interesting commonalities, as well as some differences, concerning the emphasis on sovereignist claims. We produce an account of the patterns of use of sovereignist issues, we identify which parties/leaders have been the main promoters of sovereignist claims and how their competitors have responded to this challenge. Finally, we analyse the main drivers of sovereignist party discourses. Through regression analysis we show how, both in France and in Italy, the sovereignist supply has been influenced by ideology and citizens ’ demands.

In this article we examine the phenomenon of the use of sovereignist claims (De Spiegeleire et al. 2017), not only by nationalist or populist leaders but also by actors who would not normally fall into these categories. Although this phenomenon remains under-investigated, not least because theory and concept-building are still lacking, it can be easily delineated as a recent and distinct occurrence that has emerged in the very context of globalization and Europeanization, presumably as an (over)reaction to the perceived negative consequences of both. Christopher Bickerton et al. (2022) maintain that the concept of sovereignty has become detached from the framework of the national state in post-Maastricht Europe with a dispersal of political power across a variety of actors and institutional settings and an increased focus on the EU. According to the same authors, the discourse and practice of shared sovereignty are at the heart of the conflicts of sovereignty that we observe nowadays in Europe. As a reaction, sovereignism reflects the wish of politicians, as well as citizens, to 'take back control' with renewed emphasis on the state level, along with a call to recover, at this precise level, power that has slipped away to more distant layers of governance (Kallis 2018).
Stephan De Spiegeleire et al. (2017) maintain that whereas nationalism implies the idea of making the state congruent with a nation, sovereignism primarily asserts that no international/supranational power should limit the authority of the state in its jurisdiction. Its focus on the restoration of sovereignty explains why sovereignism has a peculiar meaning and should not be considered a simple proxy for muchstudied nationalism (Basile and Mazzoleni 2020). The two concepts may overlap in some cases, but they are also distinctive and may materialize in different combinations or be separate from each other. In any case, it is commonly recognized that, under the pressure of global forces and as a consequence of increased transnational integration, sovereignism has been on the rise in recent times (Sondel-Cedarmas and Berti 2022).
In this article we make use of a multidimensional notion of sovereignism encompassing political, economic and cultural aspects. Through our empirical analysis, we focus on two major political challenges introduced by sovereignism: (1) an impact on parties and political leadership and (2) an impact of demand on supply. In more specific terms, our first research interest is to clarify to which parties/leaders sovereignist strategies and preferences can be attributed. Second, our interest is to track with accuracy the main determinants behind the use of sovereignist claims in the political system, paying specific attention to the impact of demand. For this purpose, we zoom in on two different cases: France and Italy. These countries epitomize opposite patterns in the process of state-and nation-building, with 'defence of state sovereignty' being entrenched in the French culture, but not so much in the Italian one. Hence, our case selection allows us to control for the mediating power of different domestic cultures. We test whether, despite moving from different preconditions, these two countries share some similarities and have in common the same determinants of use of sovereignist claims within the respective party systems.
For this purpose, through an analysis of Twitter, a platform used by political leaders to communicate their desired messages to the public like an un-mediated press release, we examine the discourse of the political leadership in election campaigns. We document some interesting commonalities concerning the adoption of sovereignist claims in the analysed countries. More precisely, we produce an account of which parties/leaders have been the main promoters of sovereignism (and how their competitors have responded to this challenge). Additionally, we analyse the main drivers of sovereignist party discourse. Through regression analysis we show how, both in France and Italy, sovereignist supply has been influenced by party ideology and voters' demands, a finding that may have important implications for the patterns of issue competition in these two countries.
The article is structured as follows. In the next section, we present our framework for analysis, including our working hypotheses. In the subsequent section, we put the issue of sovereignism into the context of our two selected country cases. Then, we discuss the method and the data used to test the hypotheses. The following section discusses the results of our empirical analysis. Some concluding remarks summarize the main findings of our work with specific attention devoted to the theoretical implications of our results.
The importance of sovereignism for political competition The de-territorialization of the centres of power is a phenomenon that has certainly weakened the decision-making authority of the state, creating, as a result, concern about its capacity to see public demands addressed with effective policies (Benhabib 2002;Stefanova 2018). In citizens' growing lack of trust towards supranational institutions and their policies, it is certainly possible to trace a sign of their distance from those institutions (Dahlberg and Linde 2016). In this context, a sovereignist upsurge advocates a return to an international order where states take back control over laws, policy and international interactions and focus on the protection of the self-identified interests of the native population (De Spiegeleire et al. 2017). In this perspective, supranational institutions and global market forces have become the main target of a sovereignist course of action, as have immigrants who, in the same perspective, are seen as 'outsiders' to the native population and its interests (Conti et al. 2018). If sovereignty refers to the political authority of governing bodies within a bounded territory, sovereignism specifically claims political control within the boundaries of the traditional nation state. Politically, sovereignism projects a 'reinvention of the border' of the state as 'the marker of redeemed sovereignty' (Kallis 2018: 298) with a return to the traditional understanding of sovereignty based on the idea of mutually exclusive territories defined by the traditional form of the state (Basile and Mazzoleni 2020: 154). The increasing centrality of sovereignist claims in political campaigns has nourished the notion of the primacy of a transnational cleavagea new focal point in party competition Marks 2009, 2018; Jackson and Jolly 2021)within contemporary political systems. This cleavage has at its core a reaction against open borders, pointing to a large cultural and economic conflict triggered by transnational integration (Kriesi 1998).
Sovereignism is also a complex phenomenon that applies to different fields. In particular, we considered sovereignism across different dimensions pertaining to the political, economic and cultural spheres. On the political level, because of the primacy of EU integration in the European continent, we considered sovereignism as a process of getting back control of borders, flow of people, decision-making processes and policy competence with respect to the EU (the main supranational body to which the European countries delegate shares of state sovereignty). On the economic level, we considered the call to decrease global and intra-EU economic interaction and interdependence, and to increase a protectionist approach and the state's reappropriation of independent action in areas such as monetary policy and trade. On the cultural level, we considered the construct of friend/foe and ingroup/outgroup schemes with respect to immigrants. This multidimensional inspection allowed us to include in our analysis different (political, economic and cultural) nuances of sovereignism and to capture the multifaceted nature of this phenomenon. As we explain in greater detail below, this goal has been achieved through the use of a compound measurement of the dependent variable aggregating different indicators.
In a context of increasingly sizable support for sovereignist ideas and mobilization (Basile and Mazzoleni 2020; Sondel-Cedarmas and Berti 2022) asserting national authority/autonomy over certain values and policy areas (Heinisch et al. 2020), our analysis is guided by an attempt to address the following research questions. First: Which parties/leaders have been the main entrepreneurs of sovereignist issues and how have their competitors responded to this challenge? This question aims to deliver an accurate description of the phenomenon of sovereignism at the party level (the supply side). Second: What are the main drivers of sovereignist entrepreneurship and what is the specific role played by public attitudes in this respect? Parties compete by assigning salience to different matters through a process of selective issue emphasis (Bélanger and Meguid 2008;Budge and Farlie 1983). The concept of salience in party competition hinges on the notion of issue ownership or credibilitythat is, the degree to which a party has developed a good public image in handling a certain policy issue (Petrocik 1996;Van der Brug et al. 2007). Parties choose to dismiss certain conflicts that could potentially jeopardize their electoral stability or, vice versa, they assign salience to conflicts that could constitute an electoral asset. Political actors tend to mobilize voters by prioritizing the issue they own and shelving those issues that could provide electoral advantages for their opponents (Bélanger 2003;Bélanger and Meguid 2008). Thus, by emphasizing it, parties have the ability to transform sovereignism into one of their signature issues for political contestation. As issue salience is a fundamental element of party competition, our approach aims to investigate the main determinants of sovereignist supply, with a focus on the structural properties of parties and on the 'demand' side of the political system.

The role of party agency
Part of the case made here is that party agency does matter by virtue of the relevance of party-level characteristics for party reaction to social change and public demands. Our most basic assumption is that the ideology of a party will influence the strategic choice of where, across the issue agenda, to place more or less salience. This assumption follows Jan Rovny and Stephen Whitefield's argument (2019) about the role of political agency in stimulating the discussion and relevance of different issues in their appeals to voters. We expect party ideology to be a fundamental factor in this game, as different party families normally exhibit distinct patterns in the evolution of their issue profiles. Ideology influences which issues the party really 'owns' (Petrocik 1996); that is, which issues it has a tradition of emphasizing as key elements of its identity, and where it can project itself as being more competent than others to manage them policy-wise.
In particular, parties on the (radical) right will put more emphasis on the concept that sovereignty belongs to the territorially bound state and it can neither be shared nor relinquished (Deimantaitė 2021). These parties will normally condense different aspects of sovereignism, from protectionism and defence of national sovereignty in policy to ethno-nationalist xenophobia (Eger and Valdez 2019;Rooduijn 2015). Thus, it will be interesting to document how the emphasis on sovereignism varies as we move from the centre to the extremes of the political spectrum and if reference to sovereignty issues has become a constituting part of this ideological divide. Similarly, parties on the TAN pole of the GAL-TAN dimension are more likely to be in favour of traditional societal structures and practices, more opposed to social change induced by liberal values and open borders, and to be advocates of deference to the state (Jackson and Jolly 2021). On the contrary, on the GAL side, parties are more likely to support global progressive and environmental policies, the defence of civil rights worldwide and international cooperation. Finally, Eurosceptic parties normally oppose the deepening of authority transfer to the EU level (Basile and Mazzoleni 2020). By grounding their Euroscepticism in a mix of motivations such as a defence of national identity, the economic protection of native populations (De Vries and Edwards 2009) and other stances, these parties usually pledge to restore state sovereignty and to halt the increase of EU prerogatives.
We thus aim to test the following hypotheses pertaining to the role played by party ideology: Hypothesis 1a: Right-wing parties are more likely than the other parties to emphasize sovereignism, with radical right parties being the most likely of all.
Hypothesis 1b: TAN parties are more likely than GAL parties to emphasize sovereignism.
Hypothesis 1c: Eurosceptic parties are more likely than pro-EU parties to emphasize sovereignism.

The role of public demand
We are also interested in understanding the role played by public demand on the sovereignist supply. The Responsible Party Model considers parties as the core mechanism in the process of democratic representation (Dalton 1985: 270-271): on the one hand parties should be responsive to voters' preferences, while on the other hand voters should vote for the party with views closest to their own. In general, voters hold parties and governments accountable and reward or punish them electorally based on the degree to which they are seen to have governed in line with voters' interests (Mair 2011). In this article, we consider as evidence of the alignment between citizens and parties: (1) the public assessment of party credibility on a given issue; and (2) party/citizen congruence on issue prioritization (i.e. salience-based congruence, on this point see also Giger and Lefkofridi 2014). For the assessment of this alignment, we therefore considered three different indicators: • Public credibility of parties concerning sovereignist issues; • Constituency support for sovereignist issues; • Public prioritization of sovereignist issues.
We consider this breakdown inclusive of the preferences and views of both the broad public (public credibility and issue prioritization) and of party voters (constituency issue support); thus we have a comprehensive and diversified set of measures of public demands with respect to sovereignism.
Political parties hold a reputation in terms of credibility or competence on a given issue, depending on the positions they have adopted and the image they have developed in handling the same issue over time (Bèlanger and Meguid 2008;Petrocik 1996). Specifically, the concept of credibility offers us a proxy of the notion of 'competence', not necessarily confined to technical expertise (D'Alimonte et al. 2020). Credibility allows us to analyse party-voter affinity in terms of the party's authority to achieve some issue goals based on issue differentiations between parties rather than on technical expertise (Carrieri and Angelucci 2022). According to Lorenzo De Sio and Till Weber (2020), in order to win more votes at the elections, parties tend to make more salient those issue goals that yield a high level of credibility. Thus, we advance the following hypothesis: Hypothesis 2a: Parties with stronger public credibility on the relevant issues are more likely to emphasize sovereignism (Party credibility).
Even if parties are willing to adopt catch-all strategies to defeat their competitors electorally, they first have to secure the widespread support of their own constituents to avoid perilous electoral losses. Consequently, parties should have, first of all, the incentive to emphasize those issues more largely supported by their electoral base to minimize the risks of internal divisions within the party, regardless of the level of total issue support among the electorate at large (De Sio and Weber 2014). It is worth noting that many parties have indeed enjoyed high levels of constituency support for sovereignist issues in recent times (Carrieri 2020;Conti et al. 2018;De Sio and Weber 2020). Hence, we hypothesize that: Hypothesis 2b: Parties with higher levels of constituency support for the relevant issues are more likely to emphasize sovereignism (Intra-party support).
Second, parties may shape their salience strategies based on the issue priorities expressed by the public at large in order to maximize votes beyond their own constituents. Indeed, the general public holds its own set of priorities concerning the issues on the agenda, with parties' ultimate electoral success depending on their capacity to show responsiveness and attention towards voters' main concerns (Wagner and Meyer 2014). To win more votes, beyond their usual supporters, (especially catch-all) parties should tend to emphasize those issues prioritized by voters, 'riding the wave' of people's core concerns during electoral campaigns (De Sio and Weber 2020; Spoon and Klüver 2014). Thus, we expect the following: Hypothesis 2c: When citizens prioritize sovereignism, parties are more likely to give emphasis to it (Public priority).
In the following section, we place the phenomenon of sovereignism within the context of the two analysed countries, before moving on to the subsequent sections where we test the above hypotheses empirically.

The historical role of national sovereignty in France and Italy
The structural differences between France and Italy make these two cases particularly interesting for comparison, as they allow us to control for the influence of domestic culture on the politicization of sovereignism on the domestic scene. Considering the different traditions that, presumably, should create different predispositions in these two countries about principles of national autonomy over values and policies, an identical causal mechanism behind the party emphasis on sovereignist claims would prove particularly relevant from a theoretical point of view.
Traditionally, national sovereignty has been a key word in the vocabulary of French political thought and discourse for centuries (Hoffmann 2019). The political and territorial unity of Franceresulting in a profound state-centric conception of the nation stateis the product of a historical process of construction in the face of external threats and internal divisions linked with feudalism and later with the wars of religion (Joxe 2008). Having achieved unity, the French have proved to be particularly uncomfortable with the notion of forces operating in competition with the state that are not subject to the state's control. In this respect, according to scholarship, globalization and the transnationalization of politics raise challenges for the state that history has not prepared the French to take on (Joxe 2008: 161).
It is worth recalling that one of the most important messages the French received from the first president of the Fifth Republic, De Gaulle, was that defence of French sovereignty and interests should be an absolute priority even in the context of international cooperation (Sutton 2011). Indeed, in France, the establishment of a statecentric conception has been accompanied by an assertiveness abroad, constantly seeking to remind othersas well as itselfthat the country remains a major international independent force. Despite the fact that France has become greatly downscaled during the course of history in comparison with the Gaullist dream, defending state sovereignty has always been an important motto of French politics. In this respect, although pursuing European integration convincingly, at least since the presidencies of Giscard d'Estaing and Mitterrand and further under Chiracwho definitively broke with the old Gaullist notion of a 'Europe of nation states' -French elites have always made efforts to find the winning combination that could legitimize transnational integration while preserving state sovereignty. The right balance has been sought with a discourse underplaying the loss of state sovereignty while emphasizing the gains to interests and identity achieved through French leadership in Europe (Schmidt 2007).
But as the French 'no' vote on the EU Constitutional Treaty in 2005 showed, this kind of discourse has gradually lost much of its effectiveness. France's leadership role has been slipping, its identity has come into question and there have been growing questions about whether the EU actually serves the French national interests. According to some scholars, the 2005 'no' vote marked the termination of trust in a 'social welfare and economic growth' confidence pact dear to the French and that a social Europe should have protected (Ivaldi 2006). Since then, the domestic debate on French sovereignty has become even more compelling. This is shown, for example, by the EU question playing as a major theme in the elections, with Marine Le Pen's promise for a UK-style in-out referendum on EU membership if she were to win the 2017 presidential election, challenged by Emmanuel Macron's opposite injection of the concept of 'European sovereignty' into public discourse. Although exiting the EU was no longer a Le Pen proposal in the 2022 French elections, establishing the superiority of French laws over EU laws still was.
In contrast to France, where a stable unified state structure preceded the modern national community and was established earlier in history, the Italian state only emerged at the end of the 19th century mainly as the idealized construct of a small group of modernizing elites (certainly, not of a nation in search of a state). Here, state-building proved to be a much later process, one that produced a different outcome compared to France. In the first half of the following century, this process was still in its infancy and had little to do with the concrete experiences of the population (Cardoza 2010). Indeed, the primacy of non-national affiliations based on class, religion, monarchy and regionalism among others (so-called identity containers; see on this point Caruso and Späth 2020) persisted well after the formation of the Italian state and contributed from the beginning to its fragmentation. Before the Italian nation state could reach consolidation, the defeat of Italy in World War II brought about a widespread negative perception of nationalism that fascism had attempted to introduce by forcing a representation of a nation unmarred by internal divisions (on this point, see Ben-Ghiat 1997).
Italian identity has certainly developed as a weaker national identity than the French one (although some studies on immigration confirm an invigoration of this identity when confronted with immigrants: on this point see Garau 2014). At the same time, post-war Italy has gained a reputation for having a weak state interventionist yet administratively weakwith fragile government coalitions and few legislative instruments available to the executive; a state easily penetrated by societal and political interests, with deep territorial divides dictated by strong regional and local identities/traditions and autonomous regional government structures (Putnam et al. 1994). Furthermore, the establishment of asymmetrical external relations under the American sphere of influence, from the early post-war period to the end of the Cold War era, has greatly contributed to limit the sovereignty of the Italian state. As to its relationship with the EU, Italian domestic elites long shared a belief in the need for externally imposed constraints and discipline to overcome the problems posed by domestic corruption and the chronic inefficiency of the state at home (Dyson and Featherstone 1996). It is clear, at this point, that the issue of (defence of) state sovereignty has been much less central to Italian contemporary politics compared to what happened in France. However, since the turn of the 21st century, sovereignist and identitarian groups have become more visible, popular and politically influential in this country too, creating, as a result, a collision with the established course of action in Italian politics that was epitomized by the advent of a right-wing self-defined sovereignist government in 2022. 1 In the following sections we test whether, despite coming from different backgrounds and different predispositions towards sovereignist claims, France and Italy have become closer in respect of the use of sovereignism on the domestic scene.

Data and method
The contemporary literature recognizes the importance of social media communication to capture party strategies in politicizing different issues (D'Alimonte et al. 2020). In this respect, a so-called 'press release assumption' has been proposed (De Sio et al. 2018), suggesting that Twitter cueing does not solely affect social media users, it also holds an indirect impact outside the Twittersphere with the traditional media covering and conveying parties' tweets to a broader audience (Broersma and Graham 2013;Von Nordheim et al. 2018). Indeed, Twitter has been considered by some scholars as very much 'media oriented' and an ideal tool to influence media coverage (McCombs et al. 2014). In this perspective, Twitter can be considered a platform performing an agenda-setting function in politics, with news media often treating Twitter feeds as news. Confirmations of this phenomenon spreading across Europe have been brought up by several scholars (Paulussen and Harder 2014;Von Nordheim et al. 2018) who demonstrated how social media have become a 'convenient and cheap beat for political journalism' Graham 2012, 2013). Given its increasing importance, in this work we decided to observe salience at party level precisely through Twitter communication. We therefore specified our dependent variable as the Twitter salience of sovereignism among parties and party leaders. This research strategy provides us with a direct, unmediated measurement of party efforts to mobilize the issues that are most relevant to our analysis.
To test our hypotheses, we rely on the Issue Competition Comparative Project (ICCP; De Sio et al. 2019), which provides a mix of individual-level and party-level data. Firstly, ICCP surveys collected information on citizens' attitudes, covering among others a differentiated set of issues related to sovereignism. These computerassisted web interviews (CAWI) were designed by national experts and carried out in the month preceding the 2017 presidential elections in France and the 2018 parliamentary elections in Italy, the two main general elections in these two countries. Secondly, the ICCP collected (in a different data set) information on the campaign activity of political parties/leaders on Twitter during the four weeks preceding the same elections (during the analysed period, the relevant Tweets were coded every day). ICCP coders assigned each issue-related tweet to one specific issue statement. 2 Thus, our dependent variable is the Twitter salience gauged by calculating the proportion (resulting in a continuous variable varying from 0 to 1) of party/frontrunner tweets dedicated to each sovereignist goal (listed below) over the total of issue-related tweets in their platforms. 3 Since we do not use data for the most recent elections in these two countries, later in the article we discuss developments since the data were collected and the relative applicability of the studied period to the current time.
The ICCP Twitter and individual-level data sets treat each issue in terms of 'issue goals' with a choice between two rival goals (i.e. a sovereignist and a cosmopolitan goal, such as 'Reduce the number of refugees' vs. 'Increase the number of refugees'). This measurement is consistent with the directional theory of voting behaviour, which assumes that voters evaluate parties on two opposite policy goals (Rabinowitz and Macdonald 1989). 4 Thus, our Twitter salience variable does not solely capture the emphasis on sovereignism; it also captures the opposite policy objectives accounting for cosmopolitanism. This gave us the opportunity to also estimate how cosmopolitans responded to sovereignism. Specifically, our dependent variable includes six different issues, each compounded by two alternative and opposite policy goals pointing precisely to the sovereignist-cosmopolitan divide: • Staying versus leaving the EU; • Staying versus leaving the euro; • Maintaining versus reducing access to welfare benefits for immigrants; • Accepting more versus limiting the number of refugees; • Encouraging versus limiting economic globalization; • Keeping current asylum rules versus making asylum rules more restrictive. 5 In order to track the specific party positions for every issue, coders associated each party with one particular policy objective (based on the electoral potential of each objective, i.e. the issue yield). 6 This enabled us to calculate the salience of each of these policy objectives at party level, clearly separating between sovereignist and cosmopolitan claims. Table 1 reports the objectives associated with each party/ presidential candidate based on the Twitter data set. On the one hand, we see that the League, the Brothers of Italy (FdI), the Five Star Movement (M5S), 7 Marine Le Pen and Nicolas Dupont-Aignan adopted a coherent (although differently nuanced) sovereignist policy package that qualifies them as Eurosceptic, 8 antiimmigration and opponents of economic globalization (fully sovereignist profiles). On the other hand, the Democratic Party (PD), Free and Equal (LeU), Emmanuel Macron and Benoît Hamon profiled themselves as cosmopolitan actors, endorsing the process of EU integration, the free circulation of migrants and transnational economic interdependence (fully cosmopolitan). Instead, the two centre-right party/candidates -Berlusconi's Go Italy (FI) and François Fillondeveloped Make asylum rules more restrictive Sovereignist mixed stances, colluding with the radical right on migration issues but not on Euroscepticism while supporting, at the same time, economic globalization. The radical left candidate, Jean-Luc Mélenchon, opposed economic globalization, adopting a sovereignist position on this issue, while expressing more cosmopolitan orientations on the other aspects. 9 We have developed a sovereignist index by assigning a score of 1 to each sovereignist policy objective pursued by a party and 0 to each cosmopolitan one. The index varies from 0 (fully cosmopolitan) to 6 (fully sovereignist). 10

Sovereignism as a function of ideology
The ICCP Twitter data set has been structured so as to create a party-issue dyad (parties × issues) as unit of analysis. This means that each party appears as many times as the number of issues under study. For our analysis, we selected those parties/presidential candidates achieving at least 3% of valid votes at the elections in order to include only those actors with some voice in their respective party systems and to exclude the so-called 'witness' parties/candidates (see Sartori 2005). 11 The first step of the empirical analysis aims to estimate the Twitter salience of sovereignism controlling for left-right (H1a), GAL-TAN (H1b) and pro-/ anti-EU (H3c) party ideologies. To assess the party ideology, we relied on the 2017 Chapel Hill Expert Survey data (CHES; Polk et al. 2017), based on experts' evaluations of party stances on several issues. CHES provides party positions on the general left-right dimension, locating parties along a 10-point scale ranging from 0 (extreme left) to 10 (extreme right). Similarly, experts ordered political parties along a 10-point scale, varying from 0 (GAL) to 10 (TAN), summarizing overall party orientations towards democratic freedoms and rights. On the one hand, those parties in favour of expanding personal freedoms (e.g. access to abortion, active euthanasia, same-sex marriage or greater democratic participation) were labelled as green, alternative, libertarian (GAL). On the other hand, those parties in favour of order, tradition and stability and the government's firm role in guaranteeing moral authority on social and cultural issues were labelled as traditional, authoritarian, nationalist (TAN). Finally, CHES experts ordered parties along a seven-point scale according to their positions on European integration, ranging from 1 (strongly opposed to European integration) to 7 (strongly in favour of European integration). 12 Thus, to test H1a, H1b and H1c we included in the analysis the left-right, GAL-TAN and pro-/anti-EU ideological variables, measured in terms of party positions. To avoid collinearity problems, the effects of these ideological variables on the dependent variable (Twitter salience) have been analysed in different models. In order to confirm H1a and H2b, the coefficients related to the left-right and GAL-TAN factors should be significant and positive. To support H3c, the coefficient related to pro-/anti-EU ideology should be significant and negative.

Sovereignism as a function of public demand
The second step of our empirical analysis seeks to estimate the impact of individual-level predictors on the Twitter salience of parties/leaders. Firstly, we have hypothesized that parties with strong public credibility on sovereignist issues are more likely to emphasize the relevant policy goals (H2a). Therefore, we included in the models the party credibility (as assessed by citizens) variable for the different issue goals. In the ICCP individual-level surveys, respondents were asked, first to select their favoured goal related to the specific issue and then to rate each party's credibility in the selected goal, with the scores ranging from 0 (not credible) to 1 (credible). To include such a variable within our models, we merged the Twitter data set with the individual-level surveys. At this stage, we reshaped the new merged data set, multiplying each individual respondent by each issue and each party under study (individuals × parties × issues) to create individual-issue-party triads to test H2a. This is because the individual rating of party credibility varies not only by party, but also by issue. Thus, our individualissue-party triads cover all the existing combinations in the data set with respect to parties, individuals and issues and allow us to assess whether those actors holding public credibility on different policy goals are also more likely to emphasize the same goals.
Secondly, more specifically, we expect that those parties with a high level of constituency support for sovereignist issues are more likely to emphasize the relevant policy goal (H2b). The level of constituency support for the different issues is measured as the proportion of party voters (i.e. those who declared their intentions to vote for a given party in the ICCP surveys) supporting the issue goal at stake. This continuous variable, varying from 0 to 1, allows us to assess whether the party's strategic efforts are consistent with the preferences of their own voters. Here, the unit of analysis is the party-issue dyad, a measure of overall constituency support over each issue, and one that varies by party and by issue (not by individual).
Finally, we posit that when the general public prioritizes sovereignist issues, parties are also more likely to emphasize the relevant issues in order to maximize their votes (H2c). To test this hypothesis, we developed a public priority variable, controlling the level of importance assigned by each voter to each issue (a citizenissue dyad). Each respondent assessed their level of priority for each issue using a three-point scale (1 for 'low priority', 2 for 'average' priority and 3 for 'high priority'). This variable allows us to verify if party emphases are congruent with public priorities. 13 As is customary in this type of analysis, purely for control purposes, we included in the models some sociodemographic variables: age, years of education and religiosity. We operationalized these as the average sociodemographic characteristics of the internal party constituency (those claiming to vote for a given party).

Analysis and discussion
To carry out our comparison between France and Italy, we developed different sets of linear regression models for each country (clustering standard errors at the individual level). In the first set of models (1.1 and 2.1), sociodemographic controls and the individual-level predictorscredibility, priority and constituency supportwere included. In the second set of models (1.2 and 2.2) we plugged in left-right party positioning; in the third set of models (1.3 and 2.3) we inserted the GAL-TAN party ideology; and in the fourth set (1.4 and 2.4) we included the pro-/ anti-EU party positions.
In Models 1.1 and 2.1, we find a first confirmation of salience-based congruence between demand and supply: the party-based salience on the sovereignist issue dimension is indeed influenced by public demand. Indeed, public credibility shows high statistical significance in France and Italy, positively affecting the party salience of sovereignism (and, conversely, of cosmopolitanism among those parties with greater public credibility as cosmopolitans). By emphasizing the issues on which they hold greater credibility, political actors may aim strategically to maximize their electoral potential (De Sio and Weber 2020). Intra-party issue support is also positive and displays a high level of statistical significance; like party credibility, it thus influenced the Twitter salience displayed in both countries. So far, a positive and significant relationship could be confirmed with party credibility (H2a) and intra-party support (H2b) in both France and Italy. This result suggests that parties tend to emphasize those claims on which they hold a reputation for being more credible as well as those issues with greater support from their own constituents. These results suggest that sovereignist issues have become an important part of citizen/ party mobilization (Tables 2 and 3).
In addition, in Italy, the priorities expressed by the public at large have also significantly affected party salience. In general, Italian parties prove to be responsive to the general public mood and sensitive to 'riding the wave' of public opinion's demands. On the contrary, in France, the salience political leaders assign to sovereignist (and, in parallel, to cosmopolitan) issues, in general, is not related to the preferences of the general public. On these issues, French political leaders tend to align with (and to mobilize) their own voters rather than the public at large. As a result, the French party system appears more divided and confrontational on defence of state sovereignty than the Italian one, where parties appear more inclined to follow the general trend of public opinion.
Turning to ideology, when we include the left-right, GAL-TAN and pro-/ anti-EU positions (in Models 1.2/2.2, 1.3/2.3 and 1.4/2.4, respectively), we find the expected results. Salience is influenced by party ideology: in particular right-wing parties are more likely to emphasize this issue dimension (H1a confirmed). The same is true for TAN (H1b confirmed) and Eurosceptic parties (H1c confirmed). These findings hold both for France and Italy. We take this as evidence of the fact that (defence of state) sovereignty has become the signature issue of those actors located to the right of the political spectrum, on the TAN and anti-EU pole, who proved to oppose transnational integration and open borders the most.
In order to better differentiate between sovereignists and their competitors, we performed another regression model including party/candidate dummies (see Appendix 3 in the Supplementary Material) and plotted the impact of these dummies on issue salience (see Figures 1 and 2, which also report the scores obtained by the different parties in the sovereignist index in parentheses). Using as reference categories FI and Fillon (those party actors who display middle-of-the-road sovereignist scores), we can see that the coefficients of the radical right parties/candidates -League, FdI, Le Pen and Dupont-Aignansignificantly outweigh those of the other parties. The only outlier is the M5S, a party with an underlying but fading     192 30,192 30,192 30,192 R 2 0.258 0.312 0.310 0.302 Notes: Standard errors in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.
sovereignist stance (see Table 1), which in its Twitter campaign preferred to focus more on other issue dimensions. The analysis also demonstrates that the sovereignist actors have been more likely to prime this issue dimension than their (cosmopolitan) counterparts. Indeed, Figures 1 and 2 show that cosmopolitans (those with lower scores in the sovereignist index) are significantly less active in this dimension compared to their sovereignist counterpart (those with higher scores in the same index), with Macron, Hamon, PD and LeU manifestly downplaying the relevant issues on Twitter. Thus, from the analysis, it has emerged that, both in France and in Italy, the Twitter salience of sovereignist issues is a function of ideology, with (radical) right, TAN and Eurosceptic parties being their chief users while, comparatively, the opposite side of the political spectrum has given less emphasis to those cosmopolitan issues on which it should exert its ownership.
In the same 1.2/2.2 and 1.3/2.3 models, we find that the relationship between party credibility (H2a) and intra-party support (H2b) on the one hand and Twitter salience on the other can broadly be confirmed. The only exception to this is represented by party credibility in Model 1.2 where, when including the left-right ideological variable, the effect of party credibility on salience becomes null. Maybe because in France contestation of national sovereignty has always been an overcrowded, transversal field of political competition, when introducing the left-right variable it becomes more difficult to recognize who is more credible in relation to the subject matter. is not updated to the most recent elections, it is evident that in both countries several parties and leaders have made efforts to emphasize sovereignist issues in their Twitter communication. At this point, it seems relevant to consider party contestation of this dimension, as our results appear to vindicate the argument of several scholars that transnational issues are becoming more of a focal point in party competition Marks 2009, 2018;Jackson and Jolly 2021). The main promoters of sovereignist issues are (radical) right, TAN and Eurosceptic parties, who also emphasized this dimension much more than the other parties, priming arguments of 'taking back control' with a special emphasis on the state level. On the other hand, the left-oriented, GAL and pro-EU parties have not responded with the same emphasis to the sovereignist challenge (our analysis shows that they actually de-emphasized those issues), since, we imagine, they are keener to compete on other dimensions. This result appears to contradict the argument of Daniel Jackson and Seth Jolly (2021) that the transnational-nationalist dimension is orthogonal to left-right. From our in-depth analysis of two countries, the relationship with ideology appears instead to be confirmed and it is something that could help explain the extraordinary electoral achievements of the radical right in France and Italy, an attainment which, in our view, should be considered carefully in light of sovereignist claims that have become one of the (demand-driven) signature issues of this political camp. Despite coming from different backgrounds, France and Italy appear close to each other with respect to mobilization on sovereignist issues. Although the two countries remain different in many respects, they are similar as concerns party competition, where a demarcationist camprooted in claims for the re-appropriation of state sovereignty and border controlis opposed to the rest, marking a fundamental pattern of political contest. Future work could investigate whether the sovereignist challenge will push the cosmopolitans to react more convincingly, thus reinforcing the politicization of this ideological divide in the longer term.
Although it has been shown that congruence between parties and their voters varies depending on different dimensions (it has been found to be considerably higher on the left-right dimension than on other dimensions, see Costello et al. 2012), in general voters hold parties and governments accountable and reward or punish them electorally based on the degree to which they are seen to behave in line with voters' interests (Mair 2011). Through our analysis, we found that parties and their frontrunners appear harmonized with their voters on issues about sovereignism (thus creating a profitable electoral market and an important component of party support over this dimension). In both countries, parties/frontrunners emphasize sovereignism (and, with less emphasis, cosmopolitanism) to keep their constituents united, minimizing the risks of internal divisions. Interestingly, in Italy parties were also closer to the issue priorities of the public at large than their French counterparts, maybe because state sovereignty is a more confrontational matter in France than in Italy, or because Italian parties have a more catch-all strategy. More research would be needed to give a definite answer. Certainly, in both countries parties have emphasized those sovereignist issues on which they were supposed to be more credible, probably as a means to amplify their perceived competence. It appears from our analysis that public opinion incentives matter for the party supply on sovereignism. Although our work does not deal directly with cleavage theory, it is certainly worth noting that the ascendancy of sovereignist claims in the political systemand their capacity to mobilize citizens and political representativesis what theory on the transnational cleavage predicted and that, in many respects, has been vindicated by our analysis.
We are aware that our results based on Twitter may not hold for other media platforms. Twitter has been defined as an intra-elite network mainly affecting the debate among journalists (Dagoula 2019). For example, Luca Carrieri et al. (2023) showed that, in several Western European countries, parties' campaigns via Twitter have a weak impact on electoral behaviour compared to other sources of political communication. Although we cannot draw any definitive conclusion on the external effects of this social media platform outside the Twittersphere, our findings suggest that Twitter has emerged as a platform where parties/frontrunners respond to the demands of their constituents, emphasizing those (sovereignist) issues on which they are regarded as being more credible and for which there is a clear public demand. In our view, this is important information for research on Twitter as a tool of political and electoral activity.
Interestingly, our results are rather homogenous for France and Italy, but we should be cautious about their generalization. Indeed, these findings may well depend on the particular contextual factors in place in these two countries that may have sparked off the (demand-driven) ascendancy of sovereignist claims and the relative decline of liberal values in society. To expand the generalizability of these results, considering the difficulties of running a large-scale analysis of party campaigns on Twitter, the same test should be extended to include at least a more diversified set of countries. For the moment, by documenting a similar pattern of sovereignist party supply in two countries, we hope we have been able to offer a robust and enduring empirical reference for other works in the field, possibly with a view to reiterating our method and expanding the comparative scope of the empirical analysis using our work as a benchmark.
Finally, having studied one point in time in each country, we cannot predict if the situation that we were able to depict is bound to remain stable in the future. In this respect, we can, however, point to the ever-expanding and renovating nature of the sovereignist camp that, in recent times, has seen the emergence in France (Eric Zemmour) and the ascendance in Italy (Giorgia Meloni) of new political leadership, while the more established sovereignist leadership has also confirmed its popularity (Le Pen) and influence (Matteo Salvini). It is useful to recall that in Italy, in 2020, before becoming prime minister, Meloni petitioned with her party for a national referendum to establish the primacy of Italian law over EU laws. In France, the 2022 presidential elections saw a surge in support for far-right candidate Le Pen. In this article, we have been able to show that the sovereignist supply in these two countries was largely demand-driven. In light of our work, we interpret this phenomenon as an effect of a sovereignist trend in public opinion starting years before, of which current developments are just a reflection.
5 In Italy, the question about asylum rules was phrased in a different, but compatible, way: making citizenship easier for legal immigrants' children versus maintaining the current legislation on immigrants' citizenship. 6 The ICCP project is grounded in the issue yield (IY) theory developed to study issue competition. YI summarizes the extent to which a goal provides a party with both public support at large and internal party support (De Sio and Weber 2014). Thus, based on a careful calculation that considers both public and constituency support, coders systematically assigned to each party the goal with higher YI (rather than the rival goal with lower YI). Here we provide some examples of tweets from the ICCP raw data set (translation by the authors): 'If you elect me, I will not sign any free trade treaty with anyone, because it would end up destroying everything.' (Jean-Luc Melénchon's Twitter account, tweet coded as Limit economic globalization) 'I will make sure that illegal immigrants present within our territory are never going to be regularized!' (Marine Le Pen's Twitter account, tweet coded as Make immigration rules more restrictive) 'Healthcare for immigrants, including the illegal ones, costs us 11 billion. What about the pensions for immigrants who have never paid a penny? 2 billion.' (Matteo Salvini's Twitter account, tweet coded as Restrict access to welfare benefits for immigrants) 7 Despite the M5S changing its position a great deal after its first participation in government, it has been shown that in 2018 a large share of its voters held a 'demarcationist' profile influenced by anti-EU, antiglobal and anti-immigration attitudes (Emanuele et al. 2022). 8 The M5S expressed anti-euro positions, but without endorsing Italy's withdrawal from the EU. On the other hand, Dupont-Aignan supported so-called Frexit while remaining a pro-currency party. 9 Although the literature has often defined Mélenchon as Eurosceptic (Ivaldi 2018), his constituents proved, instead, to be predominantly pro-European (determining his issue yield on the EU). 10 Appendix 2 in the Supplementary Material summarizes the scores of the different parties. 11 In Italy, we considered the following parties and frontrunners contesting the 2018 general elections; Go Italy (Forza Italia, FI)/Berlusconi; League (Lega)/Salvini; Five Star Movement (Movimento Cinque Stelle, M5S)/Di Maio; Democratic Party (Partito Democratico, PD)/Renzi; Brothers of Italy (Fratelli d'Italia, FdI)/Meloni; Free and Equal (Liberi e Uguali, LeU)/Grasso. As to the French case, we selected the following candidates for the presidential elections: Macron; Hamon; Mélenchon; Fillon; Le Pen; Dupont-Aignan. In France, data were collected from the Twitter accounts of the presidential candidates, not of their parties, given the monocratic nature of these elections. 12 Appendix 1 in the Supplementary Material reports the parties/candidates' positions. 13 Appendix 4 in the Supplementary Material shows the structure of the stacked data matrix and explains in greater detail the method followed to build it.