National and European Identification in Italy: a case
study of regionalist Lega supporters
Author
Federico Bucchi, 12758132 Supervisor and First Reader
Dr. T. Kuhn Second Reader
Dr. B. Burgoon June 2020
Master Thesis Political Science, European Politics and External Relations Final word count: 12016
2
Abstract
The aim of this paper is to demonstrate that although ‘Lega Nord’ left its regional confines to operate at a national level as ‘Lega, Salvini premier’, it never lost its strong regional ideology in favour of a national and European one. It does so by studying the relationship between Lega political partisanship for Lega and individuals national and European identification through the use of a binary logistical regression. This is the best approach to analyse a dichotomous variable. The theory proposes that Lega has not moved away from its original values and its ideology maintains a strong regionalist stance. Therefore, this paper suggests that the non-northerners that have voted for Lega also strongly identify themselves in their own region at the expense of other larger identities. The analysis is done using the 2017 European Values Study data set as it is the most recent, relevant survey available that was run close to the 2018 general elections. The results show no statistical significance between identification and League political partisanship. This disproves the hypotheses and support the existing literature of nested-identity and multi-levelled identity. Comparisons with older data sets should be done in order to verify if there is a change in the degree of identification in national and European identity of Lega Nord/ Lega supporters.
3 Index
Abstract ... 2
1 Introduction ... 4
2 History of Italian National Identity and its effect on the rise of Lega Nord ... 7
3 Theoretical framework ... 9
3.1 National Identity ... 9
3.2 European identity ... 13
3.3 Populism ... 15
4 Methodology ... 18
4.1 Dependent Variable: League Supporters ... 20
4.2 Control Variable: Tolerance of immigration ... 21
4.3 Model 1: ‘Do League supporters reject belonging to a national identity’? ... 23
4.4 Model 2: ‘Do League supporters reject belonging to a European identity’? ... 25
5 Results ... 28
5.1 National Identity Dimension results ... 30
5.2 European Identity Dimension results ... 32
6 Discussion ... 34
7 Conclusion ... 37
Bibliography... 40
Appendix A: Party Appeal and League Supporters ... 44
Appendix B: Tolerance of Immigration of League supporters ... 46
Appendix C: National Identity of League supporters ... 51
4
1 Introduction
Since the post-war period Italian identity politics have been dominated by parties that supported the idea of a unified Italian identity and had favourable views of European unification, integration, and identity. The Christian Democracy Party led Italy through the cold war and was responsible for the country’s affiliation to the European project. Following its collapse, the leadership of Italian politics was alternated between centre-right and centre-left parties that carried on Christian Democracy identity politics. However, this was true until the 2018 general elections when populist and Eurosceptic parties won the majority of the votes and formed Italy’s first populist government. Nevertheless, what is most surprising is not the victory of a coalition of populist parties but rather seeing ‘Lega, Salvini Premier’ gaining so many votes nationwide and being included in the ruling coalition.
Lega Nord, as described by full name Lega Nord per l’Indipendenza della Padania (Northern League for the independence of Padania) is a regionalist autonomist/independentist party, which, has for decades campaigned for Northern federalism and autonomy as a mean to protect its culture and economy, opposing any kind of ‘foreign’ identity that did not conform to that of Northern Italy. This ideology has been widely accepted by many disgruntled northerners who felt disenfranchised by the Italian government. However, since the takeover of Matteo Salvini as leader of the party, Lega Nord started to operate outside the ‘borders of Padania’ including central and southern Italy, regions considered to be incompatible with Northern culture, ultimately culminating in the creation of a parallel party to Lega Nord: Lega con Salvini Premier (Lega with Salvini prime minister) or Lega for short. Whilst Salvini might have brought Lega Nord to the national level by introducing policies that would affect all Italians, the core values of regional autonomy and federalism, dear to the party, have not been abandoned. Moreover, these elements are reflected in Lega’s sceptic approach to European Union and European identity. Lega does not believe in a centralised union of Europeans but rather in a confederation of European people where its members do not lose their sovereignty and diversity is maintained. With Lega gaining an increasing number of votes certain questions arise: What does this mean for Italian national identity? Has the expansion of Lega Nord to the rest of the peninsula weakened national identity in favour of a regional one? How does supporting Lega affect Italians’ affiliation in European identity?
If one looked at the case of Lega supporters through the concept of nested identity proposed by Brewer and Gardner (1996), and the literature that supports it, one would expect to see a semblance of this matryoshka like identification, where a regional one can be compatible with a national one or
5 an European one. However, this approach may not work at all. According to Eva Garau (2015) Italy lacks a unitary and centralised sense of national identity, or Italianness if you will, in which all Italians can identify themselves in. According to Garau this is where Lega Nord’s regionalist ideology thrives. This thesis proposes that the idea of a multi-level nested identity (regional to national to European) cannot work in an Italian context, especially when looking at those individuals who support regionalist parties. Therefore, the objective of this paper is to understand whether those who support Lega and its regionalist ideology effectively have no sense of belonging to both national and European identity. This will be done by trying to answer the question “Can a Lega supporter refuse to recognise oneself both in a national and European identity?”.
In order to better tackle this research question, this paper will operate under two sub-questions, whose objective is to narrow down the topic even further to make the formulation of strategies for data collection, case strategy and theoretical application possible. The first sub-question operates under the premise that Lega Nord has been campaigning against a unitary Italian identity and against the central Italian government since 1991. In recent national elections the regionalist and autonomist party has been collecting an increasing amount of the electoral share. This raises a question, is the rise of Lega’s popularity a sign that Italians have become disfranchised with the Italian identity, therefore, “Do Lega supporters reject belonging to a national identity?”. The second sub-question instead tackles with the compatibility of Lega endorsed identity with that promoted by the European Union. The stance of the Lega Nord towards the European Union is a complicated one as the party has shifted its position continuously over the years going from supportive to fully Eurosceptic, and this continuous change has affected its view of European identity. Lega Nord has also been a strong supporter of a confederal idea for the people of Europe (Lega Nord 2004) hinting towards some form of appreciation for European identity. As of 2017 is this still the case? Afterall, Salvini’s Lega has shifted towards a strongly Eurosceptic stance. “Do Lega supporters reject the idea of a European identity?”
This paper hypothesises that individuals who support Lega will have a stronger sense of regional belonging compared to a national and, or a European one. Furthermore, this association will be scrutinised through the use of control variables to eliminate spurious cases and check for a significant relationship between partisanship and identification. This paper will achieve these results by analysing the data provided by the final version of the European Values Study, 2017, published in May 2018. This thesis will be using the questions and variables provided by the questionnaire for the data analysis. This data set and master questionnaire were chosen as the survey is the one completed
6 the closest to the 2018 Italian general elections. It also provides the most relevant questions concerning both national and European identity.
This paper will provide an important contribution to the scholarship on national and European identification as the analysis of the supporters of Lega and how they identify themselves will provide a crucial insight on whether the rise of regionalist right-wing populism is affecting how Italians understand their own national identity and what impact it has on their identification as Europeans. Furthermore, this analysis will prove to be an interesting insight in Italian identity politics as, along with Lega, the other two parties that attracted the majority of the votes during the 2018 general elections were the Five Star Movement, a big tent (open to both right, centre and left in the political spectrum) populist movement that does not question either Italian nor European identity and is only moderately Eurosceptic, and the Democratic Party, a centre-left party with a strong pro-European stance. At first glance it seems a polarisation of identity is taking place between Italo-Europeans and regionalist Italians.
This thesis will be divided in six sections: a brief history of Italian identity politics, a definition of concepts and theory, a methodology section, data analysis and results, discussion of the results and limitations and the conclusion. The first section a brief introduction to the history of Italian national identity will be provided as it is crucial in understanding why the Lega Nord proposes the independence and/or autonomy of the Italian regions. Secondly, a thorough explanation of the theory behind the thesis will be provided. Thirdly, a section will be dedicated to providing a detailed description of how the data will be analysed, justifying why the specific dataset and questionnaire were picked, explaining the case selection, the variables, and the method. This section will present technical mathematical language. The fourth section will present the results obtained through data analysis according to the methodology, this will be done in the form of tables with written explanations of the results along with a descriptive summary of the tables. In the fifth and sixth section a discussion of the results will be done along with the conclusion to this paper. Here the results will be discussed and compared with the theoretical framework and the major literature. Furthermore, limitations of the process will be argued to identify what issues there are with the paper and what improvements can be done, if any are required.
7
2 History of Italian National Identity and its effect on the rise of
Lega Nord
The basis of regional identity and regional federalism advocated by Lega Nord is not farfetched and has solid foundation in Italian unitary history. The history of Italian identity is a complicated one and it is important to understand it in order to understand Lega’s autonomist ideology.
The most significant period concerning the Italian modern identity debate has its roots in the ‘Risorgimento’ when Italy went from being seen as a geographical expression, as the Austrian chancellor Metternich would simply put it, (Rossi, 1973, p.159) to a unified state.
In twenty years (from 1850 to 1870) Italy went from being a peninsula politically split into small nations states, each with its own cultural identity, laws and language to a unified nation state with its own national language, political institutions, laws and identity. It has to be noted that there is a debate over the actual timeline of the risorgimento, mostly concerning the annexation of the Alto Adige – Süd Tyrol region in 1918. Certain scholars consider this as the completion of territorial unification rather than a cultural.
Nevertheless, with unification the concept of Italian identity was not universally accepted and by many considered to be an imposed construct. This was true especially in the South and certain areas of the North where the Kingdom of Italy was seen as just another foreign occupier. Massimo d’Azeglio, a senator of the neo-born Italian government commented on the government’s lack of commitment to endorse the formation of unified national identity that would tie its citizens together. In his memoires d’Azeglio states: ‘Unfortunately Italy has been made, but Italians are not being made’ underlining poor governmental efforts towards the creation of an Italian identity and the strong cultural divisions within the country (D'Azeglio, 1899, p.5). In fact, this lack of commitment caused significant problems in the first decade of post-unitary Italy. Most notably, a guerrilla civil war, which at the time was branded as brigandage to maintain unified appearances, had erupted in the south where bands of Bourbon legitimists and criminals rebelled against the government for socio-cultural, economic, and political reasons. These actions were the first indications of a profound difference between the North and South, a gap that will later be better known as the ‘questione meridionale’, the southern question, that would become an inherent aspect of Lega Nord ideology.
Throughout most of the 20th century the concept of regional identity declined, and a steady
8 of Italian identity in the form of national struggle and shock; men hailing from all over the country fought together to expel the last remnants of foreign rule in Italy and achieve territorial completion. The regrouping of the defensive line behind the Piave River following the Caporetto disaster in 1917 is to this day seen as symbol of the struggle for independence for Italy and its cultural identity (Catalano Gaeta and Vajro, 2006, p.49). During the fascist period, Italian identity was glorified, and regional identities were crushed. Towards the end of the war the struggle over identity became political, between Fascism and Resistance. This was continued in the Cold War between Communism and Christian Democracy. Whilst the concept of identity was split into two political camps, efforts were still made to build on an Italian identity. The development of a national television and radio was one of the most notorious examples that encouraged the propagation of Italian identity through the use of cultural programmes, sporting events, and politics, essentially opening a window on the entire country (Casamassima, 2015).
Regardless of these efforts, by the 1980s regionalist movements emerged in Italian politics, especially in the North. These parties campaigned for the defence of regional values and independence from Italy. These parties were small and politically insignificant even after their confederation into Lega Nord and remained in the shadow of Italian politics (Garau, 2015, p.103). This all changed with the collapse of the Communist Party and the Christian Democratic Party at the end of the Cold War. These were the traditional mass parties that dominated Italian politics throughout the post-war and Cold War period and their collapse not only left a large vacuum in Italian politics, but also caused a feeling of bewilderment and a sense of loss of identity (Garau, p.105). According to scholars Lega Nord greatly benefitted from this vacuum of power as Italians were attracted to parties that promoted traditional values and possessed a strong ideology (Garau ibid.; Zaslove, 2011, p.73). Furthermore, the collapse of the mass parties brought back to the surface the ‘questione meridionale’. Despite the many attempts made by the government to industrialise the South, the whole area remained heavily economically dependent on the North, angering many northern taxpayers who felt exploited by “lazy southerners”. Lega Nord took advantage of the fear of economic exploitation to promote its regionalist agenda by promising fiscal and cultural protection from Italy in the form of a federal independent North (Diamanti, 1993, p.14).
9
3 Theoretical framework
Under the leadership of Umberto Bossi (1989-2012) the objective of Lega Nord was to promote its regionalist-federalist agenda with a strong anti-immigration and ethnonationalist-cultural protectionist stance in northern Italy. Since Matteo Salvini took over the role of party secretary in 2013, Lega Nord took a turn in its approach to identity and developed a more inclusive agenda that would branch out to the central and southern regions of Italy. However, the expansion of Lega Nord into a more inclusive Lega does not mean that the regionalist-autonomist ideology has changed. In fact, Lega is only considered to be a parallel party of Lega Nord and still preserves its core ideology of regional-federalism (D’Alimonte, Zeitlin and SPUI25, 2019; Gazzetta Ufficiale, 2017, pp.64-65) and anti-immigration stance. Since the objective of this thesis is to understand whether supporters of Lega identify themselves in a national and European identity, the objective of this chapter is to explore the debates and the research that has been done over the years concerning the perception of identity and why individuals choose to be represented by it. It will be necessary to analyse the behaviour and ideology of the ‘Salvinian’ Lega through identity lenses provided by the theory, in order to relate the research with the hypothesis presented in the introduction and the variables in the methodology.
3.1 National Identity
In order to understand the relationship between Lega partisanship and Italian national identity we have to understand what national identity means first.
National identity is viewed by many scholars as a multi-layered and constructed phenomenon. It cannot be reduced to a single element as it draws upon the elements of myriad other collective identities and memberships in which individuals associate themselves into (Greenfeld and Eastwood, 2009, p.257; Kuhn and Nicoli, 2020, p.5-6; Smith, 1991, p.15). It is possible to approach the concept of a multi-layered national identity by splitting it into two elements: a political identity and a central/cultural identity (Greenfeld & Eastwood, ibid.; Smith, p.8; Wodak, 2009, p.16).
Political identity can be understood as the concept that unites people under common rules and in a defined territory. According to Anthony D. Smith (p.9), the role of political identity in the larger concept of national identity is rather tenuous, however, it is also significant as political communities are a distinctive element that constitute Western national identity. These communities have common laws and institutions, they share close ties to well-defined territories and represent the relationship
10 between the identity of individuals and the distribution of power in a given community (Greenfeld & Eastwood, p.256; Wodak, p.21). In fact, the nation links its population under the same legal rights and duties. Furthermore, the concept of territory plays an important role in both political and Western national identity. According to authors such as Smith and Wodak nations consist of a population sharing a historic territory, which is bounded by borders, and serves as a depositary for the collective memory and the common myths of a population (Smith, pp. 9-10; Wodak, ibid.).
Whilst political identity is linked to national identity through territory and the relationship between the individual and core power, a central identity, also known as cultural identity or collective identity, is the criteria that individuals use to distinguish themselves as part of a larger group and separate from ‘others’ (Kuhn & Nicoli, p.5; Wright, 2011, p.838). This makes the concept of cultural identity inseparably bound to that of national identity, as it allows single individuals to know ‘who they are’ by creating the idea that all co-nationals are in a way equals by defining and locating individuals through personality and distinctive culture (Greenfeld & Eastwood, p.258; Smith, p.17; Wodak, p. 22). Cultural identity can be understood as the ethnic basis for national identity. It is the collection of all the social and cultural bonds that tie people together under a collective proper name, such as homogenous historical memories, traditions and symbols, shared values, aspirations and sentiments, the appreciation of differentiating elements within the common culture and a feeling of solidarity towards different sections (social classes, socio-political movements) of the population (Hall 1996, p.615; Smith pp.11, 21; Wodak p.21). Moreover, central identity can be linked with political identity as the cultural values it represents are tied to a concept of a geographical homeland.
How does Lega relate to national identity? According to authors like Kuhn and Nicoli (p.4), a major cause for identity-based conflict between the European Union and a member state nation is the integration of core state powers, as this action challenges the concept and understanding of national sovereignty. This paper believes that this disaccord is not limited at a supranational-national level but can also be projected at a national-regional level. This can be seen in the long relationship between Lega Nord and the Italian State. In fact, obtaining regional economic and administrative autonomy is one of the primary policies pursued Lega Nord (Garau 2015, p.116; Graziano, 2010, pp.44-45). Lega Nord has called for a return to pre-unitary state of affairs where the regions of the north could utilise their economic power as leverage to overcome their lack of political power and negotiate with occupying powers to ensure their autonomy (Graziano, p.47). Regionalist politicians use this notion to call for the return of particular prerogatives to northern municipalities against the believed ‘methodical plunder’ of produced wealth by the Italian state (Graziano, ibid.). The defence of autonomy is further enhanced through the advancement of a northern regional identity which satisfies
11 the national identity criteria set above. Lega Nord, claiming that an Italian identity is only a construct, has encouraged northerners to rediscover their identity by exhuming elements of past symbology (such as the Camunian Rose, the Sun of the Alps and the warrior of Legnano) related to periods when the north was independent or fought for its independence. Moreover, the concept of Padania has been promoted as the territorial homeland of all northerners, that has acted through time as the depositary of northern memory and culture, and that ultimately ties all northerners together (Garau, ibid.).
There have been previous studies that looked at the effects of political partisanship on national and regional identity. Some relevant examples can be found in the studies conducted by Cameron Anderson and Michael McGregor, and Rune Dahl Fitjar.
The Anderson and McGregor paper looks at the link between political partisanship and national identification by studying the 2015 Canadian general election (Anderson and McGregor, 2019, pp.1-2). Their claim is that election outcome has a say on how close do citizens feel to a Canadian identity (ibid., p.2). According to the authors, generally individuals who saw their party lose at the national elections had a lowered perception of their national identity (ibid. p.15). The sole exception to this finding is the Bloc Quebecois, a regionalist-independentist party, whose supporters regardless of election outcome will be anti-Canadian. This study could be used as a guideline in predicting the attitude towards Italian identity of Lega supporters following elections. For example, it is possible to predict the attitude to national identity of Lega supporters after national elections especially when looking at policies concerning immigration and foreign policy; one would expect a stronger attachment to national identity if Lega wins the elections as that would mean that Italians’ rights are being protected, whilst losing in the elections might lead to a distancing from national identity and a closer approach to regional autonomism. However, the reliability of Anderson and McGregor’s approach is limited when analysing the relationship between Lega supporters and national identity. Lega supporters will always have a lesser sense of belonging towards national identity whether the party wins at the elections or not as the party supports regional autonomy and federalism whether it wins or not. Furthermore, Lega and Bloc Quebecois cannot relate as the latter is confined to the French-Canadian population whilst Lega targets all regionalists in Italy. Moreover, Lega operates at a national level indicating a general dissatisfaction with Italian identity for political, cultural, and economic reasons.
The Fitjar paper on the other hand concentrates on the development of sub-state regional identities by seeking to identify those regions that have a strong regional identity (Fitjar, 2010, p.522). The paper attempts to understand the common factors that mobilise such identity and what kind of effects it has on the voting behaviour (ibid.). The author explores what kind of characteristics affect the
12 development of strong regional identities in certain areas. Aspects such regional language, geographic location, level of economic development and immigration tolerance are taken into consideration in this paper. According to Fitjar, regions with distinctive localist parties, their own local language and links to European integration develop stronger regional identities (ibid. p.538), whilst economic development does not affect the evolution of regional identity but rather affects the relationship between region and central government where economically prosperous areas ask for increased decentralisation as they do not wish to subsidise poorer regions (ibid., pp538, 540-541). This study is useful as it supports the reasons behind the rise of Lega Nord and its conflict with the Italian government. Nevertheless, this study is quite limited as its approach helps to explain the rise of localised regionalism. It cannot explain however, the rise in popularity of parties like Salvini’s Lega, which is now active on a national level and reaches out to differing regional cultures.
With the rise of Salvini’s national Lega, the concept of identity has become more difficult to understand. The topic of identity is an ongoing key element of Lega’s public discourse, especially when emphasising its rediscovery and the dangers that threaten it. Salvini has clearly steered Lega identity rhetoric towards a national stance, but at the same time maintains the regionalist values and keeps on fighting for regional federalism (as seen in the autonomy referendum of Lombardy and Veneto in 2017) (Regione Lombardia). Before Salvini’s take over the enemy of northern identity were Southern Italians, considered to be a dead weight on northern economy and culturally incompatible with the north. Now there is a new enemy, the non-European migrant that threatens both northern and Italian identity. Some may argue that Salvini’s decision to turn the regionalist party in a national one is only a ploy to attract voters following the vacuum left by Forza Italia caused by the fall of its leader, Silvio Berlusconi, which in part is true. Nevertheless, Lega is not the only party in the right-wing spectrum, there are plenty of other parties to which voters could have flocked to. Therefore, one asks the question, why do Italians from the centre and south vote for Lega, the only major regionalist party in Italy? Why not the other parties in the right-winged spectrum that have similar nationalist policies? Does this mean that Italians hold a stronger regional identity than a national one? This may be the case. One does not decide to vote for a party that has relentlessly attacked non-northerners for decades, just for its populist and anti-migration policies. Especially when there are other parties that push for the same policies. The rise of Lega must have awakened dormant regionalist feelings in all Italians.
13 3.2 European identity
European identity is one of the most complicated concepts to understand as there are many diverging opinions concerning it. On one side there are scholars and academics who theorise that European identity is linked to European politics and integration but plays a very small role, causing it to fracture and multiply rather than consolidate (Checkel and Katzenstein, 2009, pp.5-6); others deny the existence of a European identity, claiming that such a concept cannot exist as there is no sense of collective identity amongst the European people (Risse-Kappen, 2010, p.38). Furthermore, the supporters of this strand of thought claim that the lack of a common European language prevents the formation of a European public community. On the other side of the debate, scholars have argued that multi-diversity is not a hindrance to European identity but in fact one of its core elements (Bardeli, 2016; p.144 Martinelli, 2017, pp.7-8). For instance, the argument that a community must share a single language in order to be part of a common identity provides very little ground to build upon. In fact, there are many European nations that are multilingual or that recognise regional and local languages as official languages (Risse, p.39). Furthermore, the argument of no common European language is unfounded. The English language has emerged as a common mean of international communication (Risse, ibid.). Moreover, the lack or rigid borders amongst European nations have allowed people to become more aware of neighbouring cultures. Although, it has to be said that those who generally identify themselves as European are those who received higher forms of education and can speak English fluently.
A study proposed by Brewer and Gardner (1996, pp.86-87) argues that individuals can describe themselves as being part of multiple social identities. Meaning that one may have a sense of belonging to Europe, to their nation-state, region and so forth. This is supported by other studies concerning multi-levelled identity such as the one done by Medrano and Gutiérrez (2001) on nested identities in Spain. The authors hypothesise that Spaniards will have closer identification to their region and nation rather than Europe as the latter is perceived to be very distant and has a weaker influence over Spaniards’ feel of belonging (Medrano & Gutiérrez, p.761). However, their study discovers that those Spaniards who have a stronger identification in either nation or region also show a stronger identification with Europe, proving that individuals are capable of holding multiple identities (ibid., pp.770-771). Risse also conducted similar studies, although the author covered all European nations, reaching to similar conclusions as the Medrano-Gutiérrez study, also adding that over time there is an increase in European self-identification and that nations within Europe hold different degrees of
14 European identification (Risse, pp.40, 41, 44). Although one may understand the concept of multi-levelled identities as a Russian matryoshka doll, where the local identity is the core and the European identity is the outer layer, the reality is very different. Multi-levelled identification is not a fixed model but is heavily dependent on the individual’s perception of what they believe is their core identity (Risse, p.45). Individuals may be fully nationalists or regionalists refusing any other identities, they may be inclusive nationalists where nation comes first and Europe second, and finally, certain individuals place Europe in a higher position (ibid.).
Is there a difference between European Union (EU) identity and European identity? On the theoretical plain the two concepts are indeed very different. On one side, EU identity has no resemblance to traditional national identity and is intrinsically linked to the political nature of the European project, where identity attachment is associated with the core values of the European Union such as peace, democracy, enlightenment, human rights, cultural diversity, tolerance, and freedom of travel (Castiglione, 2009, p.39; Kaelble, 2009, pp.203-204; Risse, pp.50-52). Adding to this, individuals who do recognise themselves in this identity are not always aware of doing so as the concept has been internationalised and became part of the daily life (Kaelble, p.204). European identity, on the other hand, is culturally based. Here individuals identify in Europe through common historical and religious heritage, shared cultural traditions and cultural diversity.
It has to be mentioned that in recent times, with the increase of immigration and the rise of populism, a small group of Europeans have started to identify themselves in a chauvinist version of European identity. Huddy and Del Ponte have described national chauvinism as a sense of national superiority and dominance, xenophobic tending stance, leading to attachment of certain cultural values and the rejection of other nations and cultures (Huddy and Del Ponte, 2019, p.44). I believe that national chauvinism can also be used to describe Europeanist populism as the rise of anti-immigrant sentiments and islamophobia has risen over the years in European discourse, and there have been cases where Union applicants have been rejected as they do not share similar core cultural values as other European nations (i.e.. The case of Turkey’s application to EU membership, Bilefsky, 2009). Nevertheless, strong European attachment generally correlates with a positive attitude towards immigrants as those who identify as European are usually found within the central area of the political spectrum.
Do Lega supporters identify themselves in a European identity? In its first years of action, Lega Nord proposed itself in favour of the European Union and the identity attached to it, as it saw the supranational organisation as a protector of regional, ethnical and cultural values (Garau, p.138). Furthermore, true to its populist nature, Lega Nord propaganda encouraged Italians to identify in
15 Europe through a chauvinist approach. A uniquely Italian derogative term, ‘extracomunitario’, was developed to classify those who are not European, branding them as outsiders and enemies of common European values (Garau, p.111).
However, throughout the years the party shifted towards a more Eurosceptic and sovereigntist stance supporting the idea of a federal Europe of peoples and regions rather than a supranational centralised state. The core values of Padanian federalism and regionalism were embroidered in Lega Nord’s approach to European identity and integration, supporting a Europe built on cultural, ethnic, and economic diversity, close to regional and local realities, in essence a Europe of peoples and regions (Lega Nord, 2004). When Salvini took over he continued the anti-Europe stance and joined the coalition of far-right populist parties in the European parliament. Nevertheless, in recent times Salvini has changed Lega’s approach towards European identity. He abandoned the Eurosceptic and sovereigntist stance and returned to promoting the concept of Europe of peoples and regions.
Lega’s identity policy is heavily linked to economic and centralisation issues. It has been contesting increased centralisation policies, lack of appropriate reaction to immigration and economic crisis the same way it does in the national parliament, all elements that endanger the north, and to larger extent, Italy. In a similar way to its approach to national identity one should expect supporters of Lega to refute the European identity in favour of their regional one. Although the recent changes in party policy towards Europe, and the encouragement of a federal European Union, Lega supporters are encouraged to defend their regional and national values over European one. These individuals are expected to be inclusive regionalists and maybe accept European values but only to a scant extent and in tertiary position (region first, nation second, Europe third).
3.3 Populism
Although its position on the political spectrum is not fully clear, scholars have defined both Lega Nord and the current Lega as a populist parties (D’Alimonte et al. 2019; Zaslove, 2011, p.4). Understanding what populism means will help us better understand Lega’s approach to identity. Populism can be defined as that ideology that separates society into two distinct groups: the corrupt élite and the pure people. Furthermore, this ideology argues that politics should be the expression of the general will of the people and not the domain of well-dressed politicians (Canovan, 2004, pp. 247-248; Mudde, 2004, p.543; Zaslove, p.9). Populism diversifies itself on both sides of the political spectrum. Right-winged populism is characterised by a tendency towards authoritarianism and
16 nativism, term that can be explained as the defence of the homeland and of the national identity from dangerous, invasive foreigners (Mudde, 2007, p.18). Policy proposals from a right-wing populist party generally tend to call on the limitation and increased control over immigration (Hooghe and Marks, 2017, pp. 113-114). Social populist, or left-wing populist, parties on the other hand are more concerned with tackling economic, social, and political inequalities (emphasis on egalitarianism). Social populist agenda seeks to improve constituents economic and social rights, and unlike the right-wing branch it is not concerned in nativist ideology (Mudde, 2007, p.30). Certain scholars also look at the importance of economic factors in determining populist voters. For example, the concept of ‘losers of globalisation’ is determined by the effects caused by economic globalisation. This process has brought the automatization of industry and cheap workforces that can be found abroad, causing a gradual erosion of local workforces and damaging the socioeconomic position of the low and middle classes (Hooghe & Marks, p.114-115; Mudde and Rovira Kaltwasser, 2018, pp. 1670, 1674). Now living in economic insecurity, many of the members of these classes have turned to populist parties that promise to protect them.
It can be argued that both Lega Nord and the Salvinian Lega are right-winged populist parties. Lega Nord’s regionalist rhetoric is strongly interlinked with its populist ideology. The prime element of populist ideology is the separation of the pure people from the corrupt élite, and this can be clearly seen from the dawn of Lega Nord. Many of the slogans and speeches calling for regional autonomy and federalism are accompanied by disparaging accusations against the government (Zaslove, p.87) where ‘Rome’ is accused of thieving from northerners through taxation and Lega Nord is described their protector that will not these actions go further. Similar actions were taken by the Salvinian Lega against the government. With the migration crisis Salvini started to step Lega towards a national level viciously attacking the government for its decisions concerning migration and the establishment of a ministry of identity and integration, accusing it of damaging regional, and at large national identity (Garau, p.192).
The attachment to regional identity by Lega supporters can be described as chauvinistic as there have been clear elements of xenophobia and a sense of cultural superiority, first towards southerners and then towards non-European migrants, emphasising the strong nativist public discourse of the party (Garau, pp.109-110; Huddy & Del Ponte, p.45). In fact, Lega regionalist values have been backed by strong nativist discourse against non-northerners, accusing southerners of degrading northern values and culture. This rhetoric was then moved to accuse migrants of the same ‘crime’ during the recent migrant crisis, where Matteo Salvini has called for the closing of seaports and stronger control
17 policies over migrants. It has to be said that the nativist-chauvinist discourse, is often used by Lega most extreme supporters and not openly shared by all its leaders and members.
However, there is one element of right-winged populism that does not fit with Lega ideology: authoritarianism. As mentioned, Lega Nord and its Salvinian parallel have maintained a constant federalist, regionalist, and autonomist policy. Federalism is generally associated with decentralisation, local participation in decision making, whilst populism demands strong state centralisation. Nevertheless, Lega’s bureaucracy, although almost inexistent, hierarchy and decision-making processes are heavily centralised and dependent on the charismatic leader and their restricted council of loyal members (Zaslove, p.75). This was done to prevent party infighting, ideological divisions, therefore, in a way overcoming the lack of the authoritarian element of the party’s ideology. Due to the populist element of Lega, the element of identity belonging is blurred. Under Bossi’s Lega Nord, supporters would be strongly regionalist and nativist as southerners would be considered to be the enemies. However, under Salvini, the ‘enemy’ arrives from outside the country, making it a threat for all Italians, therefore, the nativist stance has changed. Nevertheless, one has to consider the federalist element of Lega. In my opinion, the populist aspect of Lega alters Lega supporters’ perception of identity, but only when it comes to external events that affect Italy. If not, their approach remains regionalist.
18
4 Methodology
The methodology section seeks to answer the research question through two sub questions. The first one attempts to find out whether Lega supporters reject Italian national identity and the second assesses whether Lega supporters reject European identity.
The general hypothesis of this paper (Ha) is that attachment to both national and European identity will have an effect on Lega political partisanship. The lower the attachment of individuals to both dimensions the higher the probability they will vote for the League. Its null Hypothesis (Ho) is that the attachment to both national and European identity will have no effect on Lega political partisanship.
Data
For this thesis I used the second release of the 2017 European Values Study (EVS) data set. I chose to use this dataset as the EVS questionnaire provides questions that are closely linked to both national and European identity and can be explained by the theoretical framework.
This dataset was chosen as it is the most recent (at time of research) publicly available survey close to the 2018 general election in Italy.
During the data analysis process both original (in English) and country specific (in Italian) questionnaires were used. Both have the same questions that are asked to the interviewees.
In all countries, fieldwork was conducted on the basis of detailed and uniform instructions prepared by the EVS advisory groups. The main mode in EVS 2017 is the interviewer-administered method (face to face). An alternative method is also employed by the EVS, this is a self-administered form done in a web/mail format where individual answers the questions (2nd Pre-Release EVS). Only a handful of countries chose the alternative method, Italy was not one of them. All the interviews were done face to face.
The survey conducted in Italy in 2017 had a response of 2277 respondents. Sampling strategy is provided by target country that belongs to probabilistic samples. All respondents are aged 18 or over with no roof age limit, who possess a residence address in the target country. Survey teams are required to avoid non-covering territorial, ethnic, social groups, or people with special legal status
19 (non-citizens) (EVS Methodological Guidelines, 2017). Since the European Values Study utilises SPSS for its Data Sets, I will also use said data analysis programme to obtain my results.
CaseSelection
This applies for all the following models, dependent variable, and control variable. Since I am using the EVS 2017 Dataset for Italy I will not need to do a case selection as it was already done by the survey teams according to EVS regulation and Italian directions.
Validity
This applies for all the following models, dependent variable, and control variable.
Internal Validity: The research can be considered internally valid since the data collection method for this sub-question provides information relevant to the research topic. The use of statistical analysis allows me to analyse a large amount of data without having to interview all the samples myself, providing me with a clear insight on how one’s national attachment determines their political preference. As explained in the literature and theoretical framework in the case of the League.
External Validity: The research is externally valid as EVS data collection and investigation will look at a random sample of Italian citizens. This should allow a generalisation of the sub-question to a larger population.
20 4.1 Dependent Variable: League Supporters
Operationalisation
The dependent variable that will be used in this data analysis is League partisanship. This variable is coded as a dummy variable where: Supporters of League as ‘1’ and Not Supporter of League as ‘0’. In the EVS 2017 dataset there is no variable that is so specifically orientated, thus, I had to calculate this dummy variable from an already existing one. In the questionnaire there is a question concerning political partisanship: Question 49, variable 174 (v174), ‘Which political appeals to you most?’
I re-coded this variable in a dummy variable by putting those who demonstrated support for the League as value ‘1’, and all the supporters of other parties and those who did not demonstrate any partisanship where coded as value ‘0’. (See appendix A: App-A 1 through 3)
Missing answers were removed from the analysis.
Party Support N %
Supporters of League 266 14.1
Not Supporters of League 1623 85.9
21
4.2 Control Variable: Tolerance of immigration1
League ideology is strongly anti-immigrant. The degree to which one is tolerant of immigrants will provide a strong control variable when analysing the relationship between national identification and Lega partisanship. I expect League supporters to have a higher degree of intolerance compared to partisans of other parties.
Operationalisation
A Tolerance of Immigrants variable does not exist on the EVS questionnaire, therefore, I had to create a dimension by using the following variables selected from the EVS questionnaire according to the guideline set by the theoretical framework.
- ‘How would you evaluate the impact of these people on the development of Italy?’ (v184) - ‘Do you think immigrants take jobs away?’ (v185)
- ‘Do you think immigrants exert a strain on the welfare system?’ (v187)
- ‘Do you think immigrants should maintain their distinct customs and traditions?’ (v188) - ‘To what extent do you feel concerned about the living conditions of Immigrants living in
your country?’ (v219) These are all Likert Scale variables.
Not all variables share the same scale. Variables 184 and 219 use a five-level scale and variables 185 to 188 use a ten-level scale. Since it is not possible to recode v184 to a ten-level scale, all other variables have been recoded to a five-level scale. Variables 188 and 219 have been recoded as their scales were inverted. (See appendix B).
1 Note: Tolerance of Immigration is not a result of ‘inherent racism’ in a respondent but a result of government actions
concerning immigration that leads individuals to be tolerant/intolerant.
Scale Table for Tolerance of Immigration Dimension
Variable Scale
V184 1 = Very bad, 5 = Very good
V185 (recoded) 1 = Take jobs away, 5 = Do not take jobs away
V187 (recoded) 1 = Are a strain, 5 = Are not a strain
V188 (recoded) 1 = Better if they did not maintain, 5 = Better
if they maintain
V219 (recoded) 1 = Not at all, 5 = Very much
Dimension Scale
22 To examine the internal consistency of multiple Likert Scale questions I used a Cronbach’s Alpha (CA) Reliability Test. This allows me to control for latent variables that might influence an individual’s response. The Cronbach’s Alpha test tells me if the measuring is accurate or not. The CA for these variables is above 0.700, therefore, scale well.
The dimension of Tolerance of Immigrants has been computed from these variables with the following equation: 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 = 𝑆𝑢𝑚 𝑜𝑓 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠 or 𝑇𝑜𝑙𝑒𝑟𝑎𝑛𝑐𝑒 𝑜𝑓 𝐼𝑚𝑚𝑖𝑔𝑟𝑎𝑛𝑡𝑠 =(𝑣184 + 𝑣185 + 𝑣187 + 𝑣188 + 𝑣219) 5
A correlation was run between degree of tolerance of immigrants and political partisanship to see if it affected a respondent’s political partisanship. The significance level is below 0.05, thus a significant association, demonstrating a correlation between the two.
Age and Sex (v225) where picked to be control variables but the correlation significance was above 0.05 and therefore, discarded.
23 4.3 Model 1: ‘Do League supporters reject belonging to a national identity’?
Concept and operationalisation
To understand whether Italians are experiencing national disfranchisement one has to look at their closeness to Italy. Regional attachment is also been included as it is an inherent element of Italian identity.
To create a National Identity dimension the following variables have been selected from the EVS questionnaire according to the guideline set by the theoretical framework.
- ‘How close do you feel to your region?’ (v165) - ‘How close do you feel to Italy?’ (v166)
- ‘How proud are you of being an Italian citizen?’ (v170) - ‘How important is to be born in Italy?’ (v189)
- ‘How important do you think is to respect Italy’s political institutions and laws?’ (v190) - ‘How important do you think is to have Italian ancestry?’ (v191)
- ‘How important do you think is to speak Italian?’ (v192)
- ‘How important do you think is to share Italian culture?’ (v193)
The dimension of National Identity has been computed from these variables with the following equation:
𝑁𝑎𝑡𝑖𝑜𝑛𝑎𝑙 𝐼𝑑𝑒𝑛𝑡𝑖𝑡𝑦 =(𝑣165 + 𝑣166 + 𝑣170 + 𝑣189 + 𝑣190 + 𝑣191 + 𝑣192 + 𝑣193) 8
All of the variables are considered to be four-levelled Likert Scales where 1 = Very close/
important/ proud, and 4 = Not at all close/ important/ proud. The dimension being computed from the variables maintains the same scale, where 1 = Very close and 4 = Not close at all. (See
24 Before the data analysis started the variables used to create the National Identity dimension were different. I ran a Cronbach’s Alpha test on the initial variables and the resulting CA was below 0.700, thus, not scaling well. I had to use a new set of variables (the ones shown) that obtained a CA above 0.700, therefore, scale well.
Methods
Taking into consideration that the dependent variable is dichotomous, using a binary logistic regression to analyse the relation between political partisanship and national identity is the best course of action. This method allows me to analyse the relationship between x (individual) and the independent variable calculating the probability of x falling in a group rather than the other (dependent variable). A control variable will be used in this analysis to make sure that no spurious elements are present.
Crosstabulations between partisanship and national identity and its elements will be done to observe the trend of the relationship. These, however, will not determine the results but simply provide an insight and visual aid on the direction of the relation.
3 Note: Essentially the scale measures lack of national identity as 1 is very close and 4 is not close at all. This was done
to maintain uniformity with all the dimensions as the scale direction of the large majority of the variables is the same.
Variable Scale
V165 1 = Very close, 4 = Not close at all
V166 1 = Very close, 4 = Not close at all
V170 1 = Very proud, 4 = Not at all proud
V189 1 = Very important, 4 = Not at all important
V190 1 = Very important, 4 = Not at all important
V191 1 = Very important, 4 = Not at all important
V192 1 = Very important, 4 = Not at all important
V193 1 = Very important, 4 = Not at all important
Dimension Scale
25 Hypothesis
My Hypothesis (Ha1) is that attachment to national identity will have an effect on political partisanship. The lower the attachment of an individual to Italian national identity, the higher the probability they will vote for the League.4
My Null Hypothesis (Ho1) is that the attachment to national identity will have no effect on political partisanship.
4.4 Model 2: ‘Do League supporters reject belonging to a European identity’?
Concept and operationalisation
To understand whether Italians are experiencing European disfranchisement one has to look at their closeness to Europe. In order to create the European Identity dimension, the following variables have been selected from the 2017 EVS Master Questionnaire (2018) according to the guideline set by the theoretical framework.
- ‘How important is it to be born in Europe?’ (v194) - ‘How important is it to have European ancestry?’ (v195) - ‘How important is it to share European culture?’ (v197)
The dimension of European Identity has been computed from these variables with the following equation:
𝐸𝑢𝑟𝑜𝑝𝑒𝑎𝑛 𝐼𝑑𝑒𝑛𝑡𝑖𝑡𝑦 =(𝑣194 + 𝑣195 + 𝑣197) 3
All of the variables are considered to be four-levelled Likert Scales where 1=Very important and 4=Not at all important. The dimension being computed from the variables maintains the same scale, where 1 = Very close and 4 = Not close at all. (See Appendix D)
4 Note: when looking at tables and graphs a low national attachment will fall in the ‘3’ and ‘4’ area, not in the ‘1’ and
26 Before the data analysis started the variables used to create the National Identity dimension were different. I ran a Cronbach’s Alpha test on the initial variables and the resulting CA was below 0.7, thus, not scaling well. I had to use a new set of variables (the ones shown) that obtained a CA of 0.695. Although this is still below 0.700 the reliability has been deemed acceptable as it is not possible to reduce or increase the variables and the difference from 0.700 is close to insignificant.
Methods
Taking into consideration that the dependent variable is dichotomous, using a binary logistic regression to analyse the relation between political partisanship and European identity is the best course of action. This method allows me to analyse the relationship between x (individual) and the independent variable calculating the probability of x falling in a group rather than the other (dependent variable). A control variable will be used in this analysis to make sure that no spurious elements are present.
Crosstabulations between partisanship and national identity and its elements will be done to observe the trend of the relationship. These, however, will not determine the results but simply provide an insight and visual aid on the direction of the relation.
Hypothesis
My Hypothesis (Ha2) is that attachment to European identity will have an effect on political partisanship. The lower the attachment of an individual to European identity, the higher the probability they will vote for the League. 6
5 Note: Essentially the scale measures lack of national identity as 1 is very close and 4 is not close at all. This was done
to maintain uniformity with all the dimensions as the scale direction of the large majority of the variables is the same.
6 Note: when looking at tables and graphs a low European attachment will fall in the ‘3’ and ‘4’ area, not in the ‘1’ and
‘2’.
Variable Scale
V194 1 = Very important, 4 = Not at all important
V195 1 = Very important, 4 = Not at all important
V197 1 = Very important, 4 = Not at all important
Dimension Scale
27 My Null Hypothesis (Ho2) is that the attachment to European identity will have no effect on political partisanship.
28
5 Results
The Results section only shows the necessary tables to tackle the Hypothesis for each question. All tables can be found in the relevant Appendix and Syntax file.
Please be aware that in the case of the independent variables the scale goes from ‘1’ very close to ‘4’ not at all close and in the control variable the scale goes from ‘1’ very intolerant to ‘5’ very tolerant. Therefore, when I mention increments in the level of the independent variable, I am talking about moving along the scale away from the ‘1’ value. I have kept the scales like this for the sake of consistency as the variables on the master questionnaire are coded like this.
How to interpret binary logistic regression output.
The Beginning Block can be ignored as it does not include any predictors (independent variables) thus, is not useful.
Block one includes the predictors.
1. Omnibus Test of Model Coefficients.
a. Chi-Square Test. All Chi-Squares will be the same as the predictors have been imputed at the same time.
b. Tests if the model is a significant improvement over the null model (Beginning Block). If the test is statistically significant, (P-value < 0,05) it indicates that the full model is an improvement in fit over the null model.
2. Model Summary.
a. This can be ignored as it does not show the proportion of variation in the dependent variable accounted by the predictor.
3. Hosmer and Lemeshow Test.
a. Also, Chi-Square test. If there is no significance it demonstrates that the full model fits better than the null model. The P-value must be greater than 0,05 to show a good fit of the model to the data.
4. Classification Data.
a. Shows the observed data compared to the predicted data. The correct percentage demonstrates the accuracy of the sample.
29 a. Shows the fit in the model in terms of the individual predictors. Which is the one that concerns us the most. In this table there are three elements that we have to take into consideration when analysing the output for the independent variables: B, Sig., and Exp(B).
b. B is the regression coefficient that predicts the amount of change on the dependent variable in raw score units for each one unit increase on the predictor variable. Since this is a binary logistic regression, where the dependent variable is dichotomous, we cannot talk about change in raw score units, here we want the probability of falling into a target group (element ‘1’ of the dependent variable). We cannot look at the relationship between the independent (predictor) and the probability of falling into the target group as this is a non-linear relationship. We want to study the relationship between the predictors and the dependent variable using a linear equation. We can do this by converting the probabilities to log-odds. Odds are a ratio of probability of falling in the target group over the probability of falling in the non-target group. Log-odds is taking the natural log of the Log-odds, ln(Log-odds). Therefore, we talk about the predicted change in log-odds every time there is an increase of one unit on our predictor variable.
c. Exp(B) is the odd ratio. It represents the probability of an individual falling in either group of the dependent variable according to the change in the independent (predictor) variable. The odds ratio reflects the change in odds for every unit increment on the predictor variables. That is, the odds at a given level on a predictor are multiplied by the odd ratio as we increase to another level. The value of Exp(B) is interpreted as following:
i. If Exp(B) is greater than ‘1’ then the probability that an individual will fall in the target group (‘1’ in the dependent variable) is higher than falling in the non-target group (‘0’ in the dependent variable).
ii. If Exp(B) is equal to ‘1’ then the probability of falling in the target group is same as the probability of falling in the non-target group.
iii. If Exp(B) is less than ‘1’ then the probability that an individual will fall in the target group is lower than falling in the non-target group.
d. Sig. is the statistical significance of the relationship between dependent variable and predictor. If the P-value is above 0,05 there is no statistical significance.
30 5.1 National Identity Dimension results
Hypothesis reminder.
• Hypothesis (Ha1): The lower the national attachment of an individual to Italian national identity, the higher the probability they will vote for the League.
• Null Hypothesis (Ho1): The attachment to national identity will have no effect on political partisanship.
The P-value output in the Omnibus Test is 0,00, thus statistically significant demonstrating that the model is a significant improvement of the null model. The P-value output in the Hosmere and Lemeshow test is 0,188, thus not statistically significant, therefore, confirming a better fit of the full model compared to the null model. (See Appendix C App-C 12)
The Classification Table shows that none of the observed League supporters are predicted to fall in the target group, whilst all those who are observed to not be supporters of the League fell in their predicted target group. This means I have a terrible prediction for those who support the League when it comes to national identity, but an excellent prediction for those who support other parties.
The Variables in the Equation table shows that the regression coefficient B is negative, telling us that individuals with a stronger attachment to national identity are more likely to support the League. The value of the Exp(B) is 0,805, therefore every time we increase the predictor variable by one (mind that this means moving away from national identity attachment) the probability of falling in the target group decreases. Finally, the P-value for this variable is 0,208, thus, above 0,05, meaning that there is no statistical significance. For these reasons I have to reject Ha1 as when one moves away from an attachment to national identity the probability that said individual supports the League decreases rather than increases. Furthermore, the lack of statistical significance means that in this sample, national identity has no effect on political partisanship.
The results for the control variable Tolerance of Immigration shows a negative regression coefficient, indicating that intolerant individuals are more likely to support the League. This is further enhanced by the Exp(B) which is less than one, meaning that every time we go up one level on the scale (moving towards stronger tolerance) the likelihood of an individual falling in the League supporters group decreases every time. The P-value is 0,00 showing that the relationship with League support is statistically significant.
31 Classification Table for National Identity
Observed Predicted Supporters of League (DV) Percentage Correct Not Supporter of League (0) Supporter of League (1) Step 1 Supporters of League (Dependent Variable) Not Supporter of League (0) 1423 0 100,0 Supporter of League (1) 248 0 ,0 Overall Percentage 85,2
Variables in the Equation for National Identity
B
Standard
Error Wald df Sign. Exp(B)
95% C.I. for EXP(B)
Lower Upper
Step 1 National_Identity -,217 ,172 1,587 1 ,208 ,805 ,574 1,128
Immigration_tolerance -,583 ,088 44,016 1 ,000 ,558 ,470 ,663
32 5.2 European Identity Dimension results
Hypothesis reminder.
• Hypothesis (Ha2): The lower the attachment of an individual to European identity, the higher the probability they will vote for the League.
• Null Hypothesis (Ho2): The attachment to European identity will have no effect on political partisanship.
The P-value output in the Omnibus Test is 0,00, thus statistically significant demonstrating that the model is a significant improvement of the null model. The P-value output in the Hosmere and Lemeshow test is 0,395, thus not statistically significant, therefore, confirming a better fit of the full model compared to the null model.
The Classification Table shows that none of the observed League supporters are predicted to fall in the target group, whilst all those who are observed to not be supporters of the League fell in their predicted target group. This means I have a terrible prediction for those who support the League when it comes to European identity, but an excellent prediction for those who support other parties.
The Variables in the Equation table shows that the regression coefficient B is positive, telling us that individuals with a weaker attachment to European identity are more likely to support the League. The value of the Exp(B) is 1,178, therefore every time we increase the predictor variable by one (mind that this means moving away from European identity attachment) the probability of falling in the target group increases. Finally, the P-value for this variable is 0,145, thus, above 0,05, meaning that there is no statistical significance. Although, when one moves away from an attachment to European identity the probability that said individual supports the League increases, the lack of statistical significance means that in this sample, European identity has no effect on political partisanship. For this reason, I have to reject Ha2. (See Appendix D App-D 9)
The results for the control variable Tolerance of Immigration shows a negative regression coefficient, indicating that intolerant individuals are more likely to support the League. This is further enhanced by the Exp(B) which is less than one, meaning that every time we go up one level on the scale (moving towards stronger tolerance) the likelihood of an individual falling in the League supporters group decreases every time. The P-value is 0,00 showing that the relationship with League support is statistically significant.
33 Classification Table for European Identity
Observed Predicted Supporters of League (DV) Percentage Correct Not Supporter of League (0) Supporter of League (1) Step 1 Supporters of League
(Dependent Variable)
Not Supporter of League (0)
1432 0 100,0
Supporter of League (1) 252 0 ,0
Overall Percentage 85,0
Variables in the Equation for European Identity
B S.E. Wald df Sign. Exp(B)
95% C.I. for EXP(B)
Lower Upper
Step 1 European_Identity ,164 ,113 2,121 1 ,145 1,178 ,945 1,470
Immigration_tolerance -,625 ,087 51,065 1 ,000 ,535 ,451 ,635
34
6 Discussion
The two sets of dimensions gave somewhat similar results: in both analyses the predictor variables have no statistical significance with the dependent variable, therefore, refuting this paper’s hypotheses. Had there been statistical significance in the relationship with Lega partisanship the results would have been completely different. Having to accept the null hypotheses as true means that belonging to both identity groups, one of the two or none at all, has no influence over an individual’s partisanship decision. In fact, none of those who were observed to be Lega supporters were predicted to be so. Therefore, the dimensions of national and European identity are both bad predictors of Lega political partisanship. This is unexpected as the hypotheses were based on the Lega’s regionalist ideology. The only good predictor was the control variable that showed a stronger attachment to the Lega when there is an intolerant opinion of immigrants. This was expected as it is in line with Lega ideology and right-wing populist theory.
The lack of statistical significance suggests that both national identity and European identity do not affect Lega’s political partisanship. Therefore, I have to reject the hypotheses for both sub-questions and that of this paper. This means that the answer to the question ‘Can Lega supporters refute to recognise in a national and European identity’, is no; Lega supporters do not reject other forms of identification outside their regional one. National and European identities may be integral parts of who they are. These results are unexpected and contradicts this paper’s hypothesis that both national and European identity have an effect on Lega partisanship. The results are surprising, this paper predicted a negative relationship, as a strong identification in either Europe or Italy should lead away from Lega regionalist values. These results undermine and reject the theoretical approach proposed by this paper, where Lega supporters will not identify in any other group other than a regional one, and support the literature on multi-levelled identity, where one can belong to regionalist group but also in a national one and European one.
There is no doubt that Lega supporters have strong regionalist sentiments, this element appears both in the theory, as it is a core ideology of the party, and in the data processing to achieve the overall results (See Appendix C, App-C 2, 4, 5). However, the lack of statistical significance in the binary logistic regression between political partisanship and national/European attachment demonstrates that there are Lega supporters that hold different forms of identification.
The concept of nested identity proposed by Brewer and Gardner (1996), supported by the studies made by Medrano and Gutiérrez (2001) and Risse-Kappen (2010), comes back to mind when looking at the results of the data analysis. Since both hypotheses have been disproved it is obvious that Lega
