When Yvan Colonna, a Corsican nationalist imprisoned for the 1998 murder of préfet Claude Érignac, was attacked by another inmate on 2 March 2022, mass demonstrations ensued on the island. Corsicans, especially the youth, protested against the treatment of Corsican political prisoners, and of the island more broadly, by the French state. The spark for these protests was a reminder of the not-so-distant history of political violence. Indeed, through the 2000s, Corsica along with the Basque Country were home to the only nationalist movements within liberal democracies to still use violence. This exceptionalism ended in the 2010s as the Front de libération nationale corse (FLNC) announced its ‘demilitarization’ in 2014 while Euskadi ta Askatasuna (Basque Homeland and Freedom—ETA) disbanded in 2018, seven years after having announced a permanent ceasefire. What have been the political consequences of the end of political violence in each case?

This article compares Corsica and the Basque Country to better understand how the end of violence impacts the politics of nationalist movements. The article shows that whereas in Corsica the post-violence context saw a (brief) alliance of autonomist and secessionist forces and a push for autonomy for the island, in the Basque Country no such alliance or momentum towards institutional change materialized. After tracing the process for the development of political violence in the two cases, the article provides a multi-factor explanation for the divergence in outcomes focusing on the constitutional status of the minority nation within the state, the position of nationalist parties in the regional party system, ideological polarization within the nationalist family of parties, and the level and nature of political violence when it existed.

This article is divided into five sections. The first section presents a theoretical framework centred on the permissive and productive conditions within critical junctures for helping to think about how the end of nationalist violence in Corsica and the Basque Country yielded different political outcomes. The second and third sections present the Corsican and Basque cases, respectively. The fourth section explains why the end of nationalist violence produced a nationalist alliance and new demands for autonomy in Corsica but not in the Basque Country. Finally, the conclusion sums up the findings and briefly discusses possibilities for change as sought by Corsican and Basque nationalism.

Permissive and productive conditions within critical junctures

Political opportunity structures highlight the importance of resources external to an actor that can act as ‘clues’ that either discourage or encourage particular behaviours (Tarrow 2011, 32). These can include the patterning impact on social and political processes of institutional arrangements and developments. Political opportunity structures can be considered those elements of the political environment that facilitate or impede particular actions, behaviours, or strategies (Tarrow 2005). They can include changing access to political institutions; shifts in potential allies, competitors, and opponents; and changes in the costs and benefits of particular behaviours (McAdam 1996). Political opportunity structures can be characterized as threats as much as opportunities (Goldstone and Tilly 2001).

A consideration of political opportunity structures allows for a focus on how a transformation in patterning circumstances can create opportunities for changes in the political behaviour or strategies of key actors—such as the way in which these actors engage in institutionalized politics, to what degree, with what goals, and using which strategies. Moments of significant shifts in such circumstances can be conceptualized as critical junctures.

These critical junctures function to facilitate the potential for change through the convergence of two sets of conditions: permissive and productive. Permissive conditions are those transformations to the underlying context that serve to undermine ‘mechanisms of reproduction’ that were established at previous critical junctures, and is so doing, set the stage for a new context to emerge where a change might take place (Soifer 2012, 1575). They do not in themselves create change—only facilitate the possibility. Within the context of this study, the cessation of violence serves as such a permissive condition.

Productive conditions are those conditions that interact with this new context to ‘shape the initial outcomes that emerge’ and may then in turn be ‘locked in’ (Soifer 2012, 1575). They are factors that impact what the permissive conditions will yield. These may include incentives for or against certain behaviours, changes in resources, or transformations in the perceptions and preferences of certain actors. These types of productive conditions speak to the possibility for the exercise of agency, without assuming a particular outcome. While the cessation of violence may open a possibility of increased cooperation between nationalist parties and the articulation of new, or renewed, self-determination claims, it alone does not guarantee such outcomes.

While they share an interest in nationalist politics, there are significant variations between nationalist parties in terms of preferences, resources, and strategies, which may all impact the ways in which they interact with one another in this new context. As such, these productive conditions are operationalized here in terms of (1) incentive structures; (2) institutional and constitutional status; (3) ideology; and (4) the legacy of past violence. Incentive structures involve electoral pressures as well as assessments of past and potential strategies for gaining and exercising political power. Institutional and constitutional status refers to the existing formal framework between the minority nation’s government and the state, which involves the presence or absence of political autonomy (and the degree and nature of autonomy when it exists). Ideology refers to the stated preferences of nationalist parties along the traditional left/right ideological axis. Finally, the legacy of violence concerns the extent to which violence was a feature of a conflict, and the degree of association of actors with violence and its aftermath. The nature of the interconnection between these conditions serves to pattern the behaviour of the actors in a way that either impedes or facilitates cooperation and favours or not new self-determination claims in the post-violence era.

The lack of significant change does not itself suggest the absence of a critical juncture. The presence of conditions that facilitate the exercise of agency and the possibility of change does not guarantee that such an opportunity will be seized, or that, if seized, relevant actors will be able to lock in that change: indeed, change is not a necessary element of a critical juncture (Capoccia and Kelemen 2007, 352). As such, it is important to consider the ‘near misses’ to generate a greater degree of accuracy as to the actors, events, and decisions featuring in a socio-economic outcome in order to guard against selection bias and develop more robust explanations (Capoccia and Kelemen 2007, 352). We chose cases that have in common the presence of a nationalist movement and a recent history of political violence but where there is a difference in outcomes (degree of cooperation between nationalist parties and nature of self-determination claims in the post-violence context), with the objective of pulling apart and assessing the nature of the productive conditions that facilitated the change in Corsica but not in the Basque Country. This, in turn, allows for more effective process tracing of the causal mechanisms that go beyond a narrative description of events to develop an understanding of the interaction between entities and activities that produces the observed result (Beach 2016). Hence, such process tracing involves considering the formative period for the violence since the consequences of the end of political violence were products of conditions that existed as the two nationalist movements developed a violent stream. Nationalism, like any form of politics, needs to be understood in time (Pierson 2004).

This article contributes to the literature on nationalism and on political violence. Nationalism can take different forms and pursue different objectives, and researchers have attempted to explain, for example, why some nationalist movements seek autonomy whereas others look for independence, or why some define the national community in ethnic terms whereas others are more ‘civic’ (Brown 1999; Lecours 2021). More specifically, scholars of nationalism have also long been puzzled by the ‘Janus-faced’ nature of the phenomenon, which includes both peaceful and violent manifestations (Nairn 1998). In this context, they have often looked to explain why some nationalist movements use violence whereas others do not (Medrano 1995; Conversi 1997). This article contributes to these efforts at understanding differences between nationalist movements by querying their post-violence evolution.

Even if violence is seen as a barrier to cooperation, its removal alone is not sufficient to engender cooperative behaviour. Permissive conditions alone are insufficient. The end of widespread political violence in Northern Ireland with the signing of the 1998 Good Friday Agreement (GFA) is the clearest case of a relatively successful de-escalation process in a liberal democracy after a high degree of political violence. While the institutionalized nature of consociationalism in Northern Ireland complicates comparisons, it nonetheless allows for generating some expectations on productive conditions for the development of cooperation between (moderate and radical) nationalist parties in the aftermath of violence.

The Northern Irish experience highlights the ways in which nationalist parties may be incentivized to eschew or pursue cooperation with one another post-violence. The end of violence may lead to parties previously associated with violent acts reaping electoral rewards. Indeed, in the aftermath of the GFA, the paramilitary-associated Sinn Féin (SF) quickly made electoral gains, and it soon overtook the moderate and constitutionalist nationalist Social Democratic and Labour Party (SDLP) (Mitchell et al 2018). Thus, there is reason to expect that parties associated with political violence will enjoy some degree of electoral gain following the end of violent activities, even without the development of a partnership with moderate parties.

In Northern Ireland, the potential gains from cooperation for SF and the SDLP, meaningful representation in the executive and legislative branches within the new institutional and constitutional architecture, represented a significant means of advancing the nationalist project as part of one of the two ‘national’ community blocks in the consociational structure. As the exercise of political power in the new structure required support from their ‘community block’, there were strong incentives for cooperation between the parties in the legislative and executive arenas.

The post-violence context in Northern Ireland appeared to favour the moderation of political parties, rather than their radicalization, potentially making cooperation more feasible. Indeed, despite early fears of ethnic outbidding and party radicalization in the post-violence context, there has been a notable moderation in identity and ethnicity-based rhetoric in Northern Irish parties (Mitchell et al 2018). Concurrently, there was some degree of policy convergence between the nationalist parties, both of whom fall to the left of the political spectrum. Voting preferences were nevertheless not determined by policy issues so much as by the degree to which voters felt the party represented the community identity (Tilley et al 2021). Therefore, where the nationalist parties are more ideologically compatible on non-nationalist issues, they will likely seek to compete for votes on the basis of their capacity to represent the community identity rather than on the basis of specific policy preferences. Where there exists a greater degree of ideological divergence, parties may have a more difficult time cooperating, as their understandings of how to represent the community differ.

Of course, the legacy of political violence impacts both the party associated with paramilitaries and the organizations of the broader nationalist movement on whose behalf militants claim to act. For instance, the SDLP was in many respects protected by the British and Irish governments from association with violence as a means of shoring up the party as a moderate alternative to the paramilitary associated SF (Mitchell 1995). SDLP party leader John Hume was able to use this protection to develop a series of initiatives of cooperation with SF in pursuit of a joint nationalist project (Mitchell 1995). When violence is seen by the moderate party as damaging either the nationalist project or their own party, cooperation is less likely, as the legacy of political legacy of violence lays heavy.

There are many incentives that could dissuade nationalist parties from working together in the post-violence context. These include low costs of prioritizing their own campaign, the potential rewards for ending violence, and being seen as the best representatives of the nationalist project. In this context, the parties once associated with paramilitaries have incentives to prioritize their own programme and to work with moderate nationalists only when and where this may help them be seen as more effective representatives of the national communities. Moreover, cooperation is expected to be more likely where there is a lesser degree of divergence across the left/right spectrum.

Moderate nationalist parties weigh the relative gains of cooperation. Where the exercise of political power is, or seems, feasible without cooperation with the radical nationalists, moderate nationalists are likely to not cooperate with the radical forces. This is especially the case if such cooperation carries costs to the moderate party, either in terms of its own electability or on potential cooperation with statewide parties. Conversely, when the exercise of political power appears to require a broader coalition, cooperation, at least temporary, is more likely.

The case of Northern Ireland, paradigmatic on post-violence politics in liberal democracies, suggests that the end of violence leads to cooperation between radical and moderate nationalist parties, as they look to exercise political power and achieve institutional and constitutional change. In other words, the expectation is that the end of violence leads to more numerous and significant alliances between nationalist actors. In the two cases of this study, Corsica, but not the Basque Country, conforms to these expectations.

Corsica: post-violence nationalist surge and the push for autonomy

The story of nationalism in Corsica is inseparable from the stranglehold the so-called clans exercised on politics on the island until very recently (Briquet 1997; Fazi 2009; Lenclud 2012; Sanguinetti 2014). Beginning in the late nineteenth century, two families of rural notables, the Rocca Serra and the Giacobbi families, developed, extended, and firmly established their political domination of politics in Corsica through clientelist networks that would be reproduced and prove extremely resilient through time (Briquet 1997, 143; de la Calle and Fazi, 2010). The development of nationalism in Corsica represents as much a reaction to the power of these ‘political dynasties’ (Fabiani 2018, 43) as it does a challenge to the French centralized unitary state and its concept of the nation as ‘one and indivisible’. In actuality, opposing the clans necessarily involved challenging the French state since the clans used (and distributed) the resources provided by that state to consolidate their power.Footnote 1 Hence, the families that dominated Corsican politics for over a century were always wedded to the centralized unitary model that consolidated their power, and therefore, they either opposed any form of autonomy outright or defined the concept in a very limited way so that its possible implementation would not threaten their position of influence and, indeed, dominance.

In the absence of a Corsican assembly and executive, claims for autonomy necessarily came from extra-institutional organizations (for example, the Front régionaliste corse, FRC and the Action régionaliste corse, ARC), which, in the 1960s, presented themselves as ‘regionalist’. A common thread among these organizations at the time was a reading of the relationship between mainland France and Corsica as colonial (Briquet 1997, 245). Part of the solution, therefore, was for Corsica to manage its own land and economic development. In this decade, the terminology of autonomy typically featured modest-sounding terms such as ‘autonomie de gestion’ rather than the more ambitious concept of ‘autonomie interne’, which was perceived on the mainland as something that led to independence (Fazi 2022a, b).

The development of political violence was the product of these claims for autonomy remaining unanswered, which involved the continuation of a clientelist system essentially closed to different political voices. In other words, the continued absence of formal Corsican political institutions where the variety of positions on the future of Corsica could be articulated and debated triggered a radicalization of methods, messages, and objectives of political actors who sought change in the relationship between Corsica and the French state. Such radicalization came primarily in the form of the creation of the Front de libération nationale corse (FLNC) in 1976. Over the following several decades, the FLNC and its numerous splinter organizations (Dominici 2004; Anderson 2019, 75) conducted thousands of attacks, mostly against property. It took a resolutely nationalist turn in comparison to most of the previous organizations that claimed to speak on behalf of Corsica, placing ‘the Corsican people’ at the forefront of its discourse.Footnote 2 In doing so, it targeted the French state, described as colonial, but also the ‘local notables’ (notables locaux) deemed to be ‘enemies of the nation’ (instruments traîtres à la nation) (Fabiani 2018, 56). Finally, it took a secessionist stance, framing political action in terms of a struggle for national emancipation (Dominici 2016, 105).

From that moment until 2014, nationalism in Corsica featured two streams: a first (often dubbed ‘legal’) ‘autonomist’, which sought legislative autonomy for the island and rejected violence (while at the same time showing concern for the plight of ‘political prisoners’); and a second, ‘indépendantiste’, which supported the use of violence for attaining outright independence (Dominici 2004, 101; Fabiani 2018, 58; Fazi 2022a, b). Although the autonomous stream exhibited greater coherence and continuity,Footnote 3 the radical, clandestine, secessionist stream dominated until the mid-2000s (Crettiez 2000, 59; Dominici 2004, 101) despite its own internal fragmentation.Footnote 4 The secessionist stream marginalized the whole nationalist political sphere because of its use of political violence. There were always tensions between the two streams resulting from political violence, but there were also periods of rapprochement between some of their respective parties and organizations (Dominici 2004, 106–107).

Nationalist political violence in Corsica was born and developed in the dual context of a clientelist clan system and the absence of political autonomy. Breaking the former and altering the latter (towards full independence) were the objectives pursued by the FLNC through extra-institutional means. French legislation on the territorial structuring of the state, or on Corsica more specifically, did little to bring about change on either front. The 1982 ‘decentralization laws’, which included legislation specifically for Corsica, established a Corsican Assembly (‘Assemblée de Corse’) whose members were directly elected by universal suffrage and provided a ‘distinct status’ (‘statut particulier’) for Corsica.Footnote 5 The Corsica legislation gave the Corsican Assembly some administrative leeway in formulating policy in fields such as environmental protection, culture, local transportation, and urban planning, but no legislative autonomy whatsoever. Although the clans and their parties opposed decentralization, the 1982 reforms consolidated their power (Briquet 2000) as they gained and maintained control of the new institutions. Hence, the Corsican political system remained closed to all nationalists, both autonomists and secessionists. The 1991 (so-called Joxe) law on Corsica established the Collectivité territoriale de Corse and, in the hope of breaking the clans’ clientelist stranglehold on politics, introduced responsible government to the Corsican institutions (a move once again opposed by the island’s dominant political class; Michalon 2000, 52).

The assassination in 1998 of préfet Claude Érignac by the FLNC and the botched immediate response of French law enforcementFootnote 6 triggered a crisis that led then Prime Minister Jospin to initiate a process (so-called Matignon) designed to address some of the root causes of the conflict on the island. The Matignon process was controversial not only because it involved discussions with elected officials of Corsica Nazione (who defended political violence) but also because the type of autonomy that could eventually have been the subject of negotiations (adapting French laws to the Corsican context) was viewed by some as contrary to the equality and unity principles of the Republic. Moreover, support for such autonomy in the Corsican Assembly was lukewarm.Footnote 7 The autonomy provision of the 2002 legislation on Corsica, which would have allowed the Corsican Assembly to experiment with legislative authority but under the full supervision and oversight of the French National Assembly, was struck down by the Conseil Constitutionnel. The prospect of discussions on the Corsican Assembly adapting French legislation to the island was effectively ended with the defeat of Lionel Jospin at the 2002 presidential election.

There was a significant decline in political violence starting in 2008–2009, as societal tolerance for attacks was seriously diminishing.Footnote 8 This decline boosted the electoral fortunes of nationalist parties at a time when the clans’ stranglehold on Corsican politics seemed to be vulnerable Indeed, the 2010 regional elections saw a first breakthrough for these parties, which altogether (autonomist and secessionist) garnered 30.1% of the vote in the first round of voting (more than doubling the 14.9% results of 2004) (Fourquet 2017, 32). The 2014 announcement by the FLNC of its ‘demilitarization’ allowed for something different at the 2015 regional elections: the autonomists (Gilles Simeoni’s Femu a Corsica) and the secessionists (Jean-Guy Talamoni’s Corsica Libera) joined forces in the second round of voting and won a plurality of seats, producing for the first time a nationalist executive.Footnote 9 The autonomist–secessionist partnership was formalized through the creation of the nationalist alliance Pè a Corsica,Footnote 10 which won a majority of seats at the 2017 regional elections (41 out of 63). The founding ‘strategic agreement’ of this alliance specified how the renouncement of violence was a ‘powerful political act that demands reciprocal initiatives in order to consolidate the peace dynamic’ (Pè a Corsica, n.d., 4).Footnote 11 Although the alliance crumbled before the 2021 regional elections, nationalist parties dominated these elections, with Femu a Corsica garnering on its own a majority of the seats and retaining control of the executive council. The end of political violence ultimately benefitted the autonomist nationalists most. In the 2010 regional elections, autonomist forces won 11 seats in the Corsican Assembly whereas in 2021 they won 39.

In turn, the surge of autonomist nationalists (and their alliance with secessionists from 2015 to 2021) has led to much stronger claims for legislative autonomy from Corsican politicians. While the last couple of decades saw the traditional political forces of the island (i.e. the clans) sometimes rhetorically endorse autonomy (or ‘decentralization’), they typically understood these terms as involving little opportunity for the Corsican Assembly to make its own political choices, potentially distinct from those of the French state.Footnote 12 Moreover, as noted previous reforms for Corsica were launched from Paris. In contrast, the nationalist forces who have formed the Corsican executive since 2015 have put legislative autonomy for the island at the top of their agenda and have sought to engage the French government in discussions over a significant autonomy statute.

The response to these claims for autonomy from the French political class has been ambiguous. When he was running for the presidency in 2017, Emmanuel Macron spoke several times, including once in Furiani where he referred specifically to Corsica, of a ‘Girondin pact’ (pacte Girondin), seemingly indicating a willingness to offer more autonomous powers to France’s regions, including Corsica (Mastor 2018). Once President, Macron showed no hint of any ‘Girondin’ perspective, except for mentioning that ‘Corsica’ could be written into the French Constitution (Mari 2018). In mid-March 2022, again just before a presidential election and in the aftermath of the attack on Yvan Colonna and the unrest it triggered on the island, Interior Minister Gérald Darmanin indicated that the ‘institutional question’ for Corsica would be reopened were Emmanuel Macron to be re-elected as President of the Republic and that Macron and his team were ready to go ‘all the way to autonomy’ (Antech 2022).Footnote 13 While the exact outcome of this apparent opening towards (legislative) autonomy for Corsica remains to be seen, there is no doubt that any process for addressing the claims of the Corsican executive will be bound by political values and constitutional norms of equality and indivisibility.Footnote 14

The Basque Country: post-violence continuity

The existence of multiple streams of modern Basque nationalism, both in the current post-violent stage and throughout decades of clandestine violence by ETA, is inextricably linked to the pressures of the Franco fascist dictatorship. From the late nineteenth century until the collapse of the Second Spanish republic following the Spanish Civil War and the establishment of Francisco Franco’s fascist dictatorship, the Basque Nationalist Party (Partido Nacionalista Vasco, PNV/Euzko Alderdi Jeltzalea—EA) served as the vanguard organization making claims on behalf of the Basque nation. The presence of a highly centralist nationalist regime that relied on repression to prevent and quell any calls for regional autonomy or dissent drove the PNV leadership into exile, which undermined the ability of the movement to pass its structure down to a new generation of activists (Conversi 1997).

In the 1950s, young nationalists grew increasingly disillusioned with what they saw as the passivity of the PNV (Zabalo and Iraola 2022) and drew inspiration from the social and political movements calling for decolonization and armed resistance globally to form ETA in 1959 (Zabalo and Saratxo 2015). While the repression perpetrated by the Franco regime initially afforded the armed organization a degree of legitimacy within its base and the broader Basque society (Ormazabal, 2018), as Spain underwent its transition to democracy in the late 1970s, ETA was faced with a significant narrative challenge. Not only had an insurrection not happened, but the need for such a process against a democratic state was much harder to sell than against a violent dictatorship. Accordingly, in 1978, ETA announced its desire to push for an independent and socialist state through negotiations with the soon to be democratic Spanish state (Zabalo and Saratxo 2015). Yet, ETA’s tactics did not match this stated commitment, as it undertook an intensification of its violence (Mees 2003). As the democratization process developed, ETA militants argued over the degree to which they should take advantage of the new political institutions. This resulted in a schism within ETA and the founding of two political party groupings with Popular Unity (Herri Batasuna—HB) founded in 1978 soon becoming the dominant party.

By the mid-1980s, there was once again only one ETA and the strategic aim of its violence shifted once more: the focus was now on using violence to force the Spanish state to the negotiating table. While the Spanish state did agree to negotiations in the late 1980s, talks broke down resulting from a lack of trust, mutual recrimination, and the state’s unwillingness to discuss much beyond the fate of political prisoners (Casanova 2007; Whitfield 2014).

Following the failure of the negotiations, ETA, sought to undertake a war of attrition against the state (Mahoney 2018), but found itself in an increasingly different, and organizationally difficult, situation. The end of the Cold War brought a number of dynamics that increasingly impeded the functioning of the organization, including a reduced domestic appetite for armed struggle (Tellidis 2020), renewed intensity in policing and in the legislative criminalization of the Basque nationalist left milieu, and the search for alternative pathways by the broader nationalist left for pursuing the movement’s goals (Zabalo and Saratxo 2015; Zulaika and Murua 2017). Concurrently, the success of the Good Friday Agreement in Northern Ireland provided renewed hope that negotiations could work (Conversi and Espiau 2019).

Throughout the democratic period, the PNV had maintained its position as the dominant political party of Basque nationalism, committed to the electoral process, and showing either little support, or often hostility, towards ETA’s violence. The PNV was consistently the best represented party in the Basque parliament in the post-transition period, and it had little difficultly in placing its candidate in the top executive position (Arévalo-Iglesias and Álvarez-Mozos 2020). The PNV fully supported and made use of the autonomous institutions generated by the new Spanish constitution and the corresponding Statue of Autonomy for the Autonomous Community. Indeed, the PNV sought to maximize the autonomy afforded to the Autonomous Community of the Basque Country through the Statute, and it opted to maintain a relatively ambiguous stance on independence in order to focus on the Statute and attract the broadest support base possible (Lecours 2007, 96).

This stance began to shift somewhat in the late 1990s, in part because many key competencies remained un-devolved (Lecours 2007). The PNV was not alone in this increasing disillusionment. Reinforcing this slow ‘sovereigntist turn’, the PNV was involved throughout the mid-1990s in talks, led by the peacebuilding organization Elkarri, with HB and the statewide party Izquierda Unida (United Left—IU) to discuss a resolution of the conflict through dialogue, modelled after the Northern Irish process (Casanova 2007, 383). These talks eventually led to the signing of the Lizarra-Garazi Treaty in September 1998 (Mansvelt- Beck 2005, 155). By its very nature, Lizarra-Garazi was a pact written by nationalists for nationalists, and so, it inherently bypassed Madrid and the existing institutions as a viable path forward (Lecours 2007, 102). Despite continued ideological differences, the PNV’s shift towards a more assertive nationalism (Mees 2015) increased its suitability as a potential ally for HB. This viability was augmented by the fact that, historically, the parties have largely drawn support from different demographics and, as such, are not in direct competition for votes (Irvin 1999, 125).

Yet, this attempt at a united nationalist front proved short-lived. The kidnapping and killing in 1997 by ETA of a young Popular Party (PP) politician sparked a significant backlash against the organization—including from its own base—and it was demonstrative of the ever-decreasing tolerance of Basque society for violence (Mees 2003). Although in 1998 ETA declared a ceasefire as part of the Lizarra-Garazi treaty, mutual recrimination about a failure to maintain commitments drove the pact to the edge (Whitfield 2014). As the pact collapsed, the PNV leadership announced that it would never again work with HB as long as it was associated with ETA and its continued violence (Zabalo and Saratxo 2015).

This statement did not signal the end of the PNV’s ‘sovereigntist’ turn, although a different approach would follow. Beginning in 2001, the PNV developed the so-called Ibarretxe plan, which called for a free association between the Basque Country and the Spanish state and was adopted in the Basque Parliament in 2004. However, in 2005, the ‘plan’ was not allowed to be presented for consideration in the Spanish Congress. In response to the proposal, the ‘PP government made it illegal to hold referendums that could compromise the political and territorial integrity of Spain’ (Lecours 2007, 1–2).

Meanwhile, on the nationalist left, the political parties faced increasing organizational challenges. In 2000, the PP won an absolute majority, which saw an increased criminalization of the Basque nationalist left. Throughout the early 2000s, the ‘everything is ETA’ (Whitfield 2014) approach put pressure on the broader Basque nationalist movement. ETA’s violence was increasingly seen by the nationalist left as an excuse for the Spanish state to avoid a political discussion instead of a factor forcing the state to consider the political conflict in the region (Vázquez Guevara 2018).

Despite a renewed negotiation process between 2004 and 2006, little progress was made (Khatami 2013, 59). In 2006, ETA attempted to pressure the state to increase its commitment to a negotiated settlement by bombing the Barajas airport in Madrid, but the attack instead served only to further undermine the talks and facilitate the state’s withdrawal. Throughout 2006 and 2007, activists within the nationalist left—often against the preferences of ETA—undertook a process of consultation on a broad change in strategy. In June 2010, an agreement was made public between ETA and the nationalist left that called for a three-pillar approach to the resolution of the conflict based on an accumulation of forces (essentially an increased focus on developing allies), democratic confrontation (greater use of institutional politics), and national consultation. In October, ETA announced a cessation of hostilities and called for international involvement and a political resolution of the conflict.

The response from the state was muted. There has been no legal framework at the level of the Spanish state to develop a response to the post-violent conflict context (Vázquez Guevara 2018). That has not meant that institutions have been absent from the attempt to develop a way forward—Basque regional political bodies are ‘legally and politically authorized under Spanish legislation, to undertake institutional actions that could give a political response’ to attempts at de-escalation and resolution; however, they do so ‘without legal form’ (Vázquez Guevara 2018, 1098). Questions, then, of collaboration between the political forces of Basque nationalism in the region take on a particular importance.

In the wake of the agreement between ETA and the nationalist left, a space was made for the formation of the new Bildu political coalition (Zabalo and Saratxo 2015), which was able to contest the 2011 regional elections and make significant gains. Following the October 2011 San Sebastian International Peace Conference (which largely took place without state involvement), ETA announced a definitive ceasefire (Zabalo and Saratxo 2015). In 2017, ETA formally disarmed under the scrutiny of an international team of observers, and in 2018, it announced its dissolution (Jones 2018).

The announcement of a permanent ceasefire in 2011 shifted the landscape for cooperation between nationalist forces in the Basque Country. The regional branches of the PP and Spanish Socialist Workers Party (PSOE) saw their seat share drop, while Bildu’s vote share surged to nearly 25% of the vote, well above HB’s habitual 10–15%. Though Bildu’s performance would take a hit in 2016 with the arrival of Podemos, it bounced back and has since maintained between 25 and 30% of the vote share in regional elections. Thus, between the PNV and Bildu—now Euskal Herria Bildu (EHB)—pro-nationalist parties hold a majority in the Basque parliament (Kerr 2019). The nationalist left was arguably significantly rewarded for the end of ETA’s violence.

However, the question of collaboration between the two nationalist parties remains complicated. Politics in the Basque Country is divided along two key axes—left vs right and centre vs periphery—the latter referring to a preference for increased autonomy or centralization. While both the PNV and EHB fall towards the periphery (EHB more so than the PNV), they remain in quite different places along the left right divide, with the PNV being a Christian democratic party and EHB falling significantly left of centre (Kerr 2019). Furthermore, the PNV has long been far more active at the level of the national parliament (Lecours 2007). The PNV has often made use of minority governments in the Spanish Parliament to trade its support for concessions on Basque autonomy. EHB and its predecessors were always largely absent from the Spanish Parliament, although in 2019 it managed to send four representatives (five if the representative from Navarre is included; to the PNV’s six: El País, 2019). Accordingly, the PNV has long had an important role in state budget debates and so has been also very calculated about when it chooses to leverage its support in order to extract the maximum possible concessions (Vázquez Guevara 2018).

In September of 2022, the PNV admitted that Prime Minister Sánchez is ‘not paying attention’ (‘no hace caso’) to Basque nationalist projects, and suggested that any such project (for example, a reform of the Basque Statute of Autonomy) be set aside (Muñoz 2022b). Instead, the party has spoken of the potential presence of a brief ‘window of opportunity’ where the central government—in need of legislative support—may be vulnerable to pressures for following through on outstanding commitments to devolve competencies while remaining unable to commit to broader structural reforms given the potential electoral costs such reforms might trigger elsewhere in Spain (Izarra 2022a, 2022b; Muñoz 2022a). Therefore, there exists a significant divergence in the PNV and EHB’s demands and strategies.

Assessing and explaining the consequences of the end of political violence in Corsica and the Basque Country

Until very recently, Corsica and the Basque Country were exceptions among nationalist movements within liberal democracies for their use of political violence. As we just saw, violence complicated the relationship between nationalist actors. It also served as justification for the French and Spanish states to not address issues of self-determination and even, especially in the Spanish case, to not negotiate autonomy. However, the end of violence has yielded fairly different political dynamics in Corsica and the Basque Country. Two differences are particularly noteworthy.

The first is that the relationship between secessionist and autonomist parties has remained considerably more strained in the Basque Country than in Corsica. In Corsica, a nationalist coalition, Pè a Corsica, was created after a founding ‘strategic agreement’ between secessionists and autonomists emphasized how the end of violence required such alliance between all Corsican nationalists. The alliance was short-lived, but it did not collapse because of residual tensions stemming from past history of political violence. Rather, it was the more ‘usual’ rivalries within a nationalist ‘family’ that led to the end of the coalition, together with the fact that the post-violence context seemed to have disproportionally profited the autonomists. In the Basque Country, no similar alliance between the PNV and EHB occurred. Their different past stances on political violence as well as distinct collective memories and approaches to reconciliation has led to significant roadblocks on attempted collaborations (Basabe et al. 2021). The question of political prisoners also proves detrimental to EHB’s attempts at full normalization. This being said, in 2018, EHB began working on a PNV-led multiparty project for the development of a proposal for a new statute of autonomy based on the right to decide (Zabalo and Iraola 2022), though the project was never finalized.

The second is that there has been a strong push for institutional change in Corsica after the end of political violence whereas a similar effort has not maintained the same momentum in the Basque Country. The Pè a Corsica strategic agreement stated as ‘a clear institutional objective’ (un objectif institutionnel clair) ‘a genuine autonomy with legislative, administrative, and fiscal powers’ (Pè a Corsica, n.d., 6).Footnote 15 This objective was noted in Paris, as President Macron and Interior Minister Gérald Darmanin, among other French politicians, acknowledged many times having heard Corsican claims for autonomy, although their willingness to meet the demands in full remains unclear. In the Basque Country, there appeared to be some momentum to push institutional change when an expert commission (where each main party appointed a member) was tasked to produce a draft text for a reformed Basque Statute of Autonomy. The commission’s report, however, was shelved by the PNV-led government.

What does these different post-violence political dynamics explain in Corsica and the Basque Country?

A first causal factor is the different incentive structures for autonomist nationalists. In Corsica, nationalist forces never controlled the island’s political institutions while the FLNC was active. Corsican politics, formal and informal, was dominated by the clans, which were the arch-enemies of nationalist forces because they were staunch defenders of the status quo. In this context where the first objective of nationalist politicians was to break the stranglehold of the clans on Corsican institutions, the alliance between autonomists and secessionists was a necessity. In contrast, the autonomist PNV has been the hegemonic party of the Basque Country in the democratic era, as it has led all governments since 1980 with the exception of the 2009–2011 period (Mees 2015). For the PNV, there was no incentive to work closely with EHB, especially since the Basque Socialists proved to be a willing coalition partner.

A second factor is the constitutional and institutional status of Corsica and the Basque Country within the French and Spanish states, respectively. Corsica has no legislative, fiscal, or even administrative autonomy. In addition, the traditional political forces of the island, the ‘clans’, by and large still oppose autonomy. In these circumstances, there is a strong incentive for nationalists to, first, unite to oust the traditional families from the executive power and, second, once in government, make forceful claims to the French state for Corsican autonomy. Not only is autonomy the fundamental objective of the dominant strand of Corsican nationalism, but any progress on that front would allow the nationalist parties to take credit. In contrast, the Basque Country enjoys virtually complete fiscal autonomy (the Basque provinces collect all taxes and then compensate the Spanish state for services it renders in the Autonomous Community) as well as considerable legislative and administrative autonomy. The priority of the PNV, especially in the recent turbulent context of Spanish politics where the COVID-19 and Catalan crises triggered centralization impulses, is to preserve the extensive autonomy enjoyed by the Basque Country rather than to seek change.Footnote 16

The third factor explaining why there was a (brief) alliance of nationalist forces in Corsica and a sustained push for autonomy that has not taken place in the Basque Country following the end of violence is the ideological make-up of the different nationalist parties. While Corsica Libera self-identifies as a left-wing party, Femu a Corsica’s ideology is ill-defined, although broadly liberal. More importantly, both parties subsume their left–right ideological profile within their nationalist agenda. This is not the case in the Basque Country. The PNV is Christian-Democratic while EHB holds on to the fairly far-left credentials of its predecessors. Hence, starting in 2016, the PNV made clear it was willing to work with EHB on nationalist projects, but suggested that it would need to make a choice between prioritizing its leftist or nationalist goals (Albin 2016; Kerr 2019). For EHB, the nationalist and socialist projects are intimately intertwined,Footnote 17so pushing for autonomy together with a conservative party does not elicit much support within its base.

Finally, the different level, even the different types, of violence that existed in Corsica and the Basque Country may have played a role in the relationship between nationalist forces and their position on autonomy. In Corsica, most attacks were against property. Clandestine organizations are responsible for approximately 60–80 deaths, with anywhere between 30 and 45% of victims being members killed by other members or (accidentally) by their own explosive devices (Fazi 2022a, b; Crettiez and Boirot 2021, 63). In the Basque Country, ETA is responsible for over 800 deaths.Footnote 18 Moreover, organizations representing ETA victims and their families were extremely visible and politically involved. In fact, the 1990s saw a growth in highly organized and mobilized civil society organizations focused on denouncing political violence, particularly that of ETA, through a focus on the rights of victims of ETA violence. For instance, organizations such as Basta Ya (Enough) were extremely vocal in their criticism of any expressed willingness by the state to meet either with representatives of ETA or parties of the Basque nationalist left, and it sought to mobilize civil society against such negotiations, framing them as an affront to the victims of ETA. Taking advantage of a growth of civil society movements focused on decrying ETA violence, state-led initiatives, as well as legislation which sought to codify protection of the victims, opened a space for the political instrumentalization of victims (Alonso 2017). This mobilization of victimhood within Spanish counterterrorism, Heath-Kelly and de Mosteyrín (2021) note, has served to deny the political roots of the conflict. Such mobilization of victimhood continues to have currency well after the dissolution of ETA. As late as August of 2022, the leader of the statewide PP, Pablo Casado, made reference to the victims of ETA violence to denounce those parties who would work with Bildu. In sum, it was easier, considering the extent of violence in each case, to politically leave it ‘behind’ in Corsica than in the Basque Country.

Conclusion

This article sought to understand how the end of political violence affects nationalist movements in liberal democracies through a comparison of Corsica and the Basque Country. In Corsica, the end of political violence led to a strengthening of the nationalist movement, in part through a punctual alliance between the nationalist parties, and a subsequent surge of claims for increased autonomy whereas in the Basque Country it did not. As Corsica and the Basque Country share the same permissive condition—the end of political violence—the question is why is there a difference in political outcome between the two cases?

The article suggests there are four reasons corresponding to the previously identified productive conditions. First, Corsican nationalists have an incentive to take advantage of the post-violence context to push ahead with their (autonomist) claims since Corsica has no autonomy within France whereas Basque nationalists have much less to gain considering that the Basque Country already has significant legislative and fiscal autonomy.Footnote 19 Second, Corsican nationalists have always been marginalized within Corsican politics, traditionally dominated by ‘clans’, which means that, free of political violence, they could for the first time strive to gain political power on the island. By contrast, the moderate Basque nationalists of the PNV have governed the Basque Country for most of the contemporary democratic period and had no clear incentive to adopt the agenda of, or forge an alliance with, more radical, but now peaceful, Basque nationalists. Third, the absence of significant (left–right) ideological cleavages within Corsican nationalism facilitated a common push on autonomist claims whereas the significant ideological gulf between (conservative) moderate and (far-left) radical Basque nationalists made such a common front more difficult. Finally, the much higher levels of past violence in the Basque Country than in Corsica made alliances between nationalist forces and a subsequent surge more realistic in the latter than in the former case.

Independence, the objective pursued by nationalist actors who used violence in Corsica and the Basque Country, remains as illusive as ever. In the Basque Country, there is some support for independence (Fullaondo and Zabalo 2020). A March 2022 survey by the Basque government found that 22% of respondents favoured Basque independence, with another 32% offered conditional support (Muñoz 2022a). In Corsica, support for independence is probably much lower but the question is hardly ever asked by pollsters. But, as the Catalan self-determination process has shown, a state, even a liberal democratic one, can prevent secession if it wants to, especially in constitutional contexts such as France’s and Spain’s.

Both the Corsican and Basque executives face central states that are reluctant to contemplate institutional change. France launched in the summer 2022 the so-called Darmanin process involving talks between the Corsican executive (as well as other political actors such as mayors) and the French Minister of Interior. However, from the beginning, Minister Darmanin stated that these talks would be bounded ‘by the idea of autonomy we developed in the Constitution of the French Republic’Footnote 20 (Corsi 2022). In Spain, the patience of the state for reforms in the Statutes of Autonomy was most likely exhausted by the Catalan self-determination process. There, as well, institutional change needs to be coherent with norms and values not dissimilar to those of the French Republic, as witnessed by the Spanish Constitutional Tribunal narrowly interpreting or striking down many of the provisions of the reformed Catalan Statute of Autonomy in 2010. Typical of past strategies, it seems like this Basque government will await an opportunity to trade its support in the Spanish Parliament against adjustments to Basque autonomy.