Coalition or Opposition? The Behaviour of Regionalist Parties in Multi-level Competition: The Cases of the Basque Country and Scotland

This article compares the positioning behaviour of parties and reveals why one party joins a coalition, while others do not. The cases of two regionalist parties of the UK and Spain are depicted: the Scottish National Party (SNP) and Basque National Party (BNP). The study is a contribution to the problems of positioning behaviour of parties in decentralised systems with strong regionalist elements and incentives of regionalist parties (RPs) to enter a coalition with state-wide parties (SWPs). Based on process tracing and comparative analysis of the main regionalist parties, electoral data, public statements, and media statements we conclude that the ideological proxi mity on the economic dimension of the parties, and the territorial dimension are of key importance in evaluating the coalition potential between SWPs and the RPs. Moreover, competition at the regional level creates incentives for RPs to enter a coalition with SWPs and there is no evident support for the minority government to be disposed to enter a coalition rather than a majority government.


Introduction
Spain and the UK are decentralised political systems with strong regionalist elements. Bullmann [1996] notes that Spain can be classified as a regionalized unitary state and UK as a classic unitary state. However, since 1996, the domestic political situation in the UK has changed significantly. The UK can now be considered a regionalized unitary state.
In this research, we choose two regionalist parties (RPs) which possess power in their region. Regionalist parties can be conceptualized as a type of political party that exists only in a particular sub-territory of a state and set up protection of their region as their core political objective [Massetti, Schakel 2015]. In contrast, state-wide parties (SWPs) 'compete in all districts in all elections and exercise a crucial linkage function between levels of government' [Falco-Gimeo, Verge, 2013: 388].
The first RP is the Basque National Party (BNP) which can be considered as a leading regionalist party in the Basque Country. The second one is the leading regionalist party of Scotland -Scottish National Party (SNP). We are interested in reasoning of these parties to choose in favour of coalitions/coordination or against it. How do such parties make this choice?
In multi-level systems such as Spain and the UK, there are strong regional level or strong regionalist/subnational parties that are eager to enter a multi-level competition. Multi-level competition forms different combinations of relations among possible models, for example between SWPs and RPs. The different institutional settings of the countries let us consider the vertical form of a multi-level competition in the context of multi-level governance (MLG).
The factors affecting joining a coalition are ideological similarity and attitudes to territorial issues. In other words, the ideological (left-right) and territorial proximity dimensions are key criteria to create a coalition. The territorial dimension is the area of competition and bargaining between SWPs and RPs. The modality of the negotiation process depends on the degree of party radicalism in their political agenda.
Some authors claim that radical regionalist parties are less likely to get into coalition. Döring and Hellström [2013] pointed out that more radical parties are theoretically less likely to get into government. Parties positioned close to the mainstream party should be more likely to become government members by themselves or in a coalition.
This study compares the positioning behaviour of parties and reveals why one party joins the coalition, while others do not, according to the comparative analysis. The cases under investigation have been analysed using process tracing to demonstrate the evolution of the political landscape and party positioning within 6-7 years period. Process tracing allows tracking changes in regionalist and national-wide parties' behaviour given the changes in circumstances.
The scientific contribution of this research is manifold. Firstly, the recent political changes have not been yet considered in the scholarship like the Scottish Parliamentary elections 2021 and the coalition agreement of Scottish Greens and the SNP.
Secondly, the comparison of Basque and Scottish RP's behaviour is a relatively unused approach in the context of multi-level competition. Authors argue that the inclusion of foreign and security policy dimensions in the analysis should add a worthwhile research avenue because this dimension previously received little attention in the analysis of the relationships between RPs and SWPs.
Thirdly, the results of the analysis could be practically applied to other cases not only in the considered countries or in the EU, but to any systems with strong regionalist elements (India, Latin America). REGIONAL ETHNO-NATIONALISM: CONFLICT AND COALITION POTENTIAL Fourthly, the UK is a case in point in the context of MLG due to the strong vertical dimension and lack of institutionalised horizontal one. Traditionally, Spain is considered as an example of MLG [Leon, Jurado: 2020].
In this research, we use a wide range of sources: political programs of the main regionalist parties (the BNP in Spain and the SNP in Scotland), electoral data (based on El País in Spain and the Herald in Scotland), public statements, and media statements to trace the process of political changes and compare cases.

Analytical Background
The pivotal studies that provide some theoretical answers to conceptual clarifications of MLG are articles of Marks [1993] and Hooghe [1996]. They define the notion of MLG as a «system of continuous negotiations among nested governments at several territorial tiers -supranational, national, regional and local -as the result of a broad process of institutional creation and decisional reallocation that has pulled some previously centralized functions of the state up to supranational level and some down to local or regional level» [Marks 1993: 392]. Stephenson [2013: 821] points out that securing political representation and coordination for domestic structures includes «overcoming horizontal divisions and conflicts between/within regional politico-administrative bodies, and repairing distant and/or distrustful vertical relations between national and supranational levels». As an important prerequisite, the MLG system is more developed in Spain than in the UK.
At the same time, MLG in the UK is a special case. The regions in the UK do not interact with each other in the same way and regions there do not have institutions of devolution that could coordinate them together and make their «voice» common. Based on this, the development of MLG is beginning in the UK, but this MLG is not institutionalized. Despite Brexit, regions in the United Kingdom until recently were in full control of administering and implementing Cohesion Policy [Schakel 2020: 770]. Based on these considerations, the authors point out that their attention in the context of MLG is mainly on the vertical rather than horizontal interactions.
Decentralised systems complicate the strategic behaviour of political parties because they challenge these parties to coordinate party actions across levels of governance [Stefuriuc 2009]. The existence of a subnational (regional, secondary) level inevitably forces political parties to enter a multi-level game 'where parties trade coalition support at one level, for benefits at another level' [Tronconi 2015: 584]. The multi-level game makes possible different scenarios of cooperation or contestation between RPs and SWPs at both levels. Multi-level game is a part of a broader multilevel competition where at different levels political parties cooperate or contest strategically in order to get a political power.
Classical coalition theories consider the ideological proximity of the parties on the left-right dimension as the criteria to create a coalition. Filippova [2021] overviews these theories and concludes that the size of the coalition (fewer coalition members -more incentives to enter) and ideological distance play a role in coalition formation possibilities. However, the multi-level system adds one more dimension to party competition. RPs, by definition, represent the centre-periphery cleavage in decentralised countries [Alonso, 2012]. This is why the territorial dimension is the primary area of competition and bargaining between RPs and SWPs. Empirical analysis by Tronconi [2015] confirms that both policy dimensions matter: the less distance between RP and SWP on the centre-periphery and left-right axes, the more chances to make a coalition. Consequently, the reverse logic is valid as well: the more distance between the parties, the less likely the coalition agreement between them. One of the outcomes of this analysis is that RPs with radical separatist demands are unlikely to make a coalition deal with SWPs at a regional level.
The multi-level competition can end up with vertical congruence, meaning that 'sub-national government composition coincides with that of the national government' [Stefuriuc, 2009: 94]. Vertical congruence can be an optimal option for SWP because it seeks to facilitate intergovernmental relations in key policy areas, whereas RPs may seek congruence to guarantee the adequate representation of their region at the national level [Falco-Gimeno, Verge 2013].
The strategic potential of vertical congruence is conditioned by the institutional setting, type of party, and the respective bargaining weight at both levels of government [Stefuriuc, 2009]. Congruence plays a less important role in cases where the political strength of the RP is high. Grabevnik [2021] conceptualizes party strength in terms of electoral strength (share of votes in regional/national parliamentary elections) and institutional strength (share of seats in regional/national parliament and status in regional/national parliament).
The coalition potential depends on the type of government. If an RP enjoys the majority government at a regional level, it has less incentive to make a coalition agreement with SWPs. A minority government, on the contrary, increases the probability of coalition formation because the RP needs external support to realize its political agenda [Field 2016].
The blurring boundaries between domestic policies on the one hand and foreign and security policies on the other make the multi-level competition between SWPs and RPs even more contested. RPs give importance to the foreign and security policy in their electoral programs, including migration, trade, and military operations. The attention of RPs to foreign and security policy issues confirms the striving to be a mainstream party in their region. Mainstream parties are 'electorally dominant actors in the centre-left, centre, and centre-right blocs on the Left-Right political spectrum' [Field, Hamann 2015: 903]. This is a relative approach to the understanding of mainstream parties meaning that the dominant regionalist party could be a mainstream party at the regional level, however, at the national level, the party loses this status due to insufficient electoral power. Blarel and Willigen [2021] conclude that foreign policy is an area of bargaining between SWPs and RPs, which can be operationalized as coordination (policy convergence between SWPs and RPs on foreign policy matters), compromise (SWPs make concessions to the regional level), quid pro quo (SWPs push for their foreign policy agenda but make concessions to the regional level in other policy areas), and absent dialogue (SWPs ignores the regional level preferences). The foreign policy area accounts for both the centre-periphery and left-right dimensions. For example, one of the areas of contestation is migration policy. SWPs and RPs can have different stances on how to accommodate migrants or which migration policy (inclusive or restrictive) they should conduct.
To sum up, the relations between SWPs and RPs may have a character of coordination or contestation in multi-level systems. The likelihood of an SWP and RP coalition is conditioned by the proximity of stances on both ideology and territorial dimensions, the relative strength of parties, vertical congruence, type of governments, and foreign and security policy preferences.

Scotland
The SNP is the leading RP in Scotland. It dominates the regionalist agenda in Scottish politics. Grabevnik [2021] considers the SNP as having high institutional and electoral strength at the regional level. Another RP in Scotland is the Scottish Greens, which do not share seats in the national parliament and have only 8 seats in Scottish Parliament after the 2021 elections.
The SNP is considered to be a secessionist party on the territorial domain and mainstream left on the economic scale [Massetti 2009]. The SNP led the race for Scottish independence in the 2014 referendum, ended up with the defeat of the proindependence camp but with some concessions. A few days before the referendum date, UK party leaders D. Cameron (Conservative Party), E. Miliband (Labour Party), and N. Clegg (Liberal Democrats) made a 'vow' of a new devolution of powers for Scotland in the case of a No vote. The statement promised that 'because of the continuation of the Barnett allocation for resources, and the powers of the Scottish Parliament to raise revenue, we can state categorically that the final say on how much is spent on the NHS (National Health Service) will be a matter for the Scottish Parliament' 1 .
This 'vow' has been used by the SNP to foster negotiations on a new Scotland Bill with more enhanced devolved powers. The new fiscal framework devolved almost complete control over income tax and the £2.5 billion welfare budget. G. Osborne proclaimed that this deal fulfilled the referendum vow of the UK parties. N. Sturgeon noted that the more power Parliament has, the more it can deliver to the people of Scotland 2 . At the end of March 2016, the Scotland Act was enacted, and D. Mundell called it the beginning of a 'new exciting era in Scottish politics' 3 .
The SNP stands in strong opposition to the Conservative Party. At the SNP party conference before the 2015 General Elections, N. Sturgeon alluded to the deal with Labour by sending a message of 'solidarity and friendship' 4 . However, E. Miliband refused any agreement with the SNP because of their proindependence stance 5 . The 2015 elections became a Tory victory, which allowed D. Cameron to plan a Brexit referendum.
The 2016 Scottish Parliament Elections took place two months after the adoption of the Scotland Bill and one month before the Brexit referendum. The SNP eased their pro-independence stance but promised that they would call another independence referendum in case of 'significant and material change in the circumstances that prevailed in 2014' that references the unsatisfactory Brexit referendum vote [Scottish National Party 2021: 23]. In comparison to the 2011 Scotland Elections, the SNP lost 6 seats and its absolute majority in the Scottish Parliament. However, N. Sturgeon was able to form a SNP minority government. Scottish Greens co-convener P. Harvie ensured that his party would not block a bid for another independence referendum 6 .
Despite being on the same side in terms of Brexit, N. Sturgeon drastically refused to share a platform with D. Cameron and warned him not to campaign in Scotland to avoid harming the cause 7 . The results of the in/out referendum provoked a shock. N. Sturgeon proclaimed: 'Scotland faces being taken out of the EU against our will. I regard that as democratically unacceptable' 8 . These outcomes were considered by the SNP as a 'serious change in material circumstances', and they renewed its bid for an independence referendum by framing it as an attempt to secure the place of Scotland in the EU.
The outcomes of Brexit led to the new General Elections in 2017. The Conservative Party remained the largest single party in Parliament but they lost the absolute majority and formed the Tory-led minority government with a confidenceand-supply agreement with the Democratic Unionist Party of Northern Ireland (DUP) 9 . The SNP campaigned for the second independence referendum to take place before the end of Brexit negotiations with the EU. However, the 2019 General Elections returned Tories with an absolute majority of 365 seats 10 .
Apart from Brexit, the SNP highly criticized the UK foreign and security policy. The SNP provided a furious criticism of the authorization of airstrikes on Syria by Tory's absolute majority in Parliament. N. Sturgeon said that she is 'deeply troubled by the decision of the UK government to take the country into conflict' 11 . The SNP Manifesto for Scottish Parliament Elections in 2021 placed attention on migration policy. For instance, Manifesto 2021 states that 'We rely on migration to grow our population. … No-one should be made destitute as a consequence of the Tory policy of No Recourse to Public Funds (NRPF) which stops people getting access to basic essential services' [Scottish National Party 2021: 73].
The results of the 2021 Scottish Parliament Elections confirmed the dominance of the SNP with 64 seats lacking only 1 seat to get an absolute majority. N. Sturgeon formed a third minority government 12 .
To sum up, the Conservative Party maintained its majority at the national level, whereas the SNP secured its majority at the regional level. Both parties are in strong opposition to each other. Even in the case of such shared goals as keeping the UK in the EU, they found it necessary to conduct separate campaigns. The SNP champions the independence of Scotland, while the Tories strongly stand for the territorial integrity of the UK.

Basque Country
The BNP is the leading RP in the Basque Country. In the difficult political situation in Spain (political crisis of 2015-2016, referendum in Catalonia in 2017), the Basque Country has maintained greater stability due to the stable position of the BNP, which seems win every election [De la Granja, De Pablo Contreras, Pobes, Olábarri Gortázar, 2020, p. 305]. This party kept the power after May 2019 elections due to the pact with the Socialist Party of the Basque Country-Basque Country Left (PSE-EE) in the Basque Parliament and was the most voted party in the Basque Country in the general election held in 2019, after which it offered to guarantee the election of Sanchez.
According to Massetti [2009] the BNP can be considered as a mainstream right oriented party or a centre-right oriented party. The BNP is also a mainstream party at the regional level and the Bildu is their opponent party. After the last regional elections in 2020, the BNP formed a coalition with PSE-EE. The key reason was that there was a high share of votes for the Bildu in the regional elections of 2020. The BNP-PSE-EE coalition constituted the only possible strong alternative against Bildu and allows the BNP to maintain its ruling position. In this sense, the BNP and the Bildu both possess high electoral and institutional strength at the regional level (Tables 1, 2). In this research, we use a concept elaborated by Grabevnik [2021] that conceptualizes party strength in terms of electoral strength and institutional strength (Tables 1, 2).  The role of regionalist parties can also be highlighted at the national level. The leader of the Spanish Socialist Workers' Party (PSOE) Pedro Sanchez narrowly won the vote to govern in coalition at the 2019 elections in Spain. There were 167 deputies voted in favor (including BNP), 165 against and there were 18 abstentions. It meant that Pedro Sanchez heads the first coalition government since the days of the Second Republic (1931)(1932)(1933)(1934)(1935)(1936)(1937)(1938)(1939), i.e., for the first time in 80 years. The support of the BNP was of key importance there. In this sense, it could be pointed out that the BNP behaving in such a way helped Sanchez to be elected.
Regional parties also have their own position on foreign policy issues, for example, on migration policy. The BNP is in favour of pro-migration policies as do the socialists. In this case, the BNP has a relationship with socialist that can be characterized as coordination in a multi-level system. The electoral program of the BNP firstly adopted a European immigration and asylum policy as well as measures for an equitable distribution of unaccompanied migrant minors (MENA) (BNP Electoral Program, 2019-2023. Secondly, the BNP guarantees the rights of migrants that coincide with the ideas of the PSOE. Their program outlines that it is crucial to address the demographic challenge with family policies, co-responsibility, legal immigration, and the activation of employment and entrepreneurship, particularly of women. Last but not least migration policy is perceived as becoming a fundamental element for demographic balance and, consequently, for the development and consolidation of the labour market and for economic growth. There are two main factors contributing to the BNP strategic turn towards pragmatism [Mees 2015: 56]. The first factor is the strong hit to the economies of the autonomous community and the country as a whole, affected by the global financial and economic crisis. In this regard, the issue of unemployment and regional imbalances have been highlighted by the economic crisis. The second factor is the end of political violence and the self-dissolution of ETA (Euskadi Ta Askatasuna). Mees [2015] notes that parties representing left-wing nationalists were allowed to return to political activity (e.g., Sortu, Amaiur and Bildu).
This context has spurred the BNP to rethink its strategy and declare soft demands for more autonomy. In terms of Mees (2015), this phenomenon can be defined as a "silent strategic rethinking", that is, a turn towards pragmatism while elaborating a political strategy.
To sum up, the PSOE maintained its power in a coalition at the national level (the minority government of the PSOE), mainly due to the fact that the BNP supports it. The BNP continues to be the leading party in the Basque Country and maintains, albeit in a coalition, a ruling position.

Discussion and Conclusion
The BNP and the SNP are cases that demonstrate two strategies: the BNP has a coalition with SWP at the regional level, whereas the SNP is not in a coalition at the regional level and in strong opposition at the national level to the leading SWP.
To identify why some parties join a coalition, we should firstly consider the ideological proximity on the economic dimension of the parties. Surridge 13 analysed the party positioning in the UK before the 2019 general elections and concluded that the distance between the SNP (left-wing) and the Tories (right-wing) was the widest among all the participating parties. Stefuriuc [2009] states that the BNP and the PSOE are mainstream parties close to the centre in spite of being on different sides. Importantly, the ideological distance between the Bildu and the BNP is wider than that between the BNP and the PSOE [Stefuriuc, 2009, p. 107].
Secondly, the territorial dimension plays a crucial role in evaluating the coalition potential between SWPs and the RPs. The BNP eased their territorial demands from separation to more autonomy. The PSOE showed its disposal to extend the rights of autonomies and consider the federal way of territorial organization. Sanchez confirmed that 'the PSOE will work for federal Spain' 14 . On the contrary, the distance between the SNP and the Conservatives on the territorial domain is much greater. The renewed demands for independence put both parties on completely different ends of the spectrum. The territorial disparities between parties prevented the cooperation between pro-EU Conservatives and the SNP to share a platform on the Brexit referendum debates. Having the Tories as the government in Westminster allows the SNP to frame their political agenda as a contestation with the centre. The SNP tries to take advantage at the regional level by describing the UK policy as holistic against Scotland.
An example of this is the migration policy. The SNP associates the UK migration policy as led exclusively by Tories and contradictory to the interests of Scotland. The same tone can be found in relation to the UK foreign policy like Brexit or Syrian airstrikes. The SNP subsumes the international relations agenda in their core territorial claims and contestation with the central state. The BNP and the PSOE find themselves in coordination on the migration policy issue. Both parties consider an inclusive migration policy to maintain the demographic balance and economic development. The pro-migration policy stances of regionalist parties can be a feature of new nationalism and its shift from ethnic characteristics to civic [Jeram, 2014].
Thirdly, the level of competition on the regional level creates incentives for RPs to enter a coalition with SWPs. The BNP and the Bildu are strong RPs with institutional and electoral powers. These parties are far from each other on territorial and economic dimensions. The Basque Country regional level represents a segmented market of competition between two strong RPs to champion their region. Barrio and Rodriguez-Teruel [2017] consider it as an outbidding competition. In order to get a majority and maintain power, the BNP need to cooperate with the SWP to outbid their strong regionalist competitor the Bildu. The PSOE was the best choice because it is close in ideological terms and has state-wide power to potentially use vertical congruence. Vertical congruence is an institutionalized channel for the BNP to gain concessions for the regional level in exchange for their national-level support. The election of Sanchez demonstrates that not only did the BNP need the coalition with the PSOE to be in power in the Basque Parliament, but also the PSOE needs the BNP to ensure its governance in the Spanish Parliament. The Scottish regional level is dominated by the SNP. It has both high institutional and electoral strength. It has been in power in Scotland since 2007. Once the SNP was able to form a majority government after the 2011 Scottish Parliamentary Elections. Scottish Greens are another RP in Scotland but the relative strength of this party is incomparable with the SNP. However, the SNP and the Scottish Greens shared a platform in the 2014 independence referendum campaign. The Scottish Greens are close to the SNP on the economic and territorial dimensions. Having a dominant position at the regional level, the SNP has little incentive to cooperate with the SWPs. However, the proximity on economic and territorial dimension as well as similarities on other criteria like migration policy does not exclude the coalition potential with other regionalist parties. The example of coalition agreement between the SNP and Scottish Greens confirms that the vertical congruence for the SNP agenda has fewer advantages than the play on territorial contestation with the Tories. Anti-Tory rhetoric showed its efficiency to maintain the dominant position in the region.
Finally, the research did not find any support for the minority government more disposed to enter a coalition rather than a majority government. The SNP has been in a minority government since 2016 and their anti-centre and anti-Tory rhetoric has not changed much.
To sum up, the comparison of the SNP and the BNP behaviour in a multi-level competition allows us to test the hypothesis on incentives of RPs to enter the coalition with SWPs. Future research may include the Catalan RPs where both strong RPs decided to create a coalition and put aside the SWP, or the case of the support agreement between the Conservative party in the UK and the DUP.