Abstract
This article presents a quantitative typology of football fans’ attitudes towards governance. Data collection is done through an online survey in six European countries: France, Germany, Poland, Spain, Turkey and the United Kingdom. Results reveal the existence of five types of supporters: club-militants, institutionalists, critics, moderns and globalists. The critics, moderns and globalists fans share a preoccupation for football governance problems but differ in the intensity of their views. At the same time, critics and globalists are heavy consumers of football games and merchandise. The results suggest that existing fan typologies that understand supporters in dichotomic terms of authenticity or consumerism fail to explain the complex reality of a game that has developed new structures over recent decades. Existing typologies need to be superseded in favour of a more multi-disciplinary approach that integrates a governance turn to inform a more nuanced and better understanding of football’s social reality.
Over the last 20 years European football (soccer) has undergone intense changes that have radically transformed the game. The de-regulation of audio-visual markets became the main propeller of a new football business, transforming clubs into prime-time televised content businesses (Boyle and Haynes, 2004). As a result, ‘fans’ have become ‘consumers’ (King, 2002: 203), and the main European leagues are prime examples of commercial development (Horne et al., 1999). The consequences of these transformations for fans have generated interest in the academic literature; whereas some authors consider fans to be victims (Sandvoss, 2003), others define them as rational actors who adjusted to the new reality (Crawford, 2004), or as activists trying to reclaim the soul of the game (Webber, 2015). The academic literature on football fans is indeed extremely varied, both thematically and methodologically. For the purpose of this article we focus on a particular body of work within this literature: the interest to explain the reality of football supporters through typologies to understand their behaviour and attitudes. Supporter typologies have firmly developed as one of the avenues which the sociology of sport has sought to explain identities and behaviours of football fans, and it is to this body of work that our article seeks to contribute.
In one of the first typologies, Clarke (1978) distinguishes between ‘genuine fans’ and ‘other’ fans. In the form of a dichotomy, this is the most basic classification of fans, but it has been strongly influential. It has been used and adapted by several authors. For example, Redhead (1993) distinguishes between ‘active/participatory’ and ‘passive’ supporters; Quick (2000) differentiates between ‘irrational’ and ‘rational’, and Boyle and Haynes (2009) speak of ‘traditionalists’ and ‘modernisers’. There are numerous other researchers who argue in the same vein. Usually, binary definitions can be narrowed down to one group of fans that is seen as traditional and passionate as opposed to another group that embody the so-called modern supporter. This kind of approach suggests that there are certain characteristics that define active, traditional or ‘authentic’ fans such as attending matches (Gibbons and Nuttall, 2016), having close ties to the local community (Brown, 2007) and immersion in local cultural practices (Williams, 2012). Of these approaches, Redhead’s (1993) categorisation is more nuanced. It shows the first signs of a complex distinction between different types of fans. Redhead’s classification is more three-dimensional, highlighting different stages in the behaviour of fans, rather than a simple dichotomy. In that respect, it has been lauded by several authors (Cleland, 2010; Cleland and Dixon, 2015; Millward, 2012). The evolution from Clarke to Redhead reveals a basic problem with this body of literature: those early categorisations were general and rigid because they only used a reduced number of categories to segment the fans along. Moreover, as Dixon (2013: 28) points out, these approaches work by categorising groups of fans ‘imposing a model from the outside’. There is a normative flavour that sees more desirable, passionate, traditional supporters with community values pitted against supposedly bad, new consumer-fans.
In response to the limitations of those dual approaches, Giulianotti (2002) presented a categorisation of fans along a horizontal axis between the ‘traditional’ fan and the ‘consumer’ and a vertical axis ranging from ‘cool’ to ‘hot’. This creates four quadrants that Giulianotti labels ‘supporters’, ‘followers’, ‘fans’ and ‘flâneurs’. Thus, this taxonomy considers levels of activity and consumer status among supporters. Although this model has been adapted and used widely, it has also been criticised as rather rigid. In that respect, we would echo Dixon’s (2013) argument that Giulianotti’s approach is still two-dimensional in nature because it is underpinned by the traditional vs. modern fan continuum. Davis (2015: 430) points out that all four types of the typology are essentially linked to ‘consumerist ideals’ and the commercial transformation of post-modern football, as compared to the (traditional) pre-commercial age fan. Finally, Giulianotti’s typology came at a time (early 2000s) in which his work cannot, naturally, include more recent developments linked to the transformation of football, such as the role of the internet, or the interest of fans in governance.
The work of Jamie Cleland (2010) is of relevance in this respect. In his study of fan-club relations in four English clubs, Cleland built on Redhead’s (1993) understanding of participatory and passive football fans to discuss whether they are included or excluded by the clubs’ structures. Although Cleland’s objective was not to develop a typology, he defined the terms ‘active’ and ‘passive’ football fans, which are two opposing points in a continuum considering whether fans are willing to engage in governance/management issues in their club or not. These are categories that he then develops further in his work (see Cleland, 2015; Cleland and Dixon, 2015). Cleland’s contribution basically proposes also a dual distinction of fans. In fairness, he suggests the use of a continuum rather than a dichotomy, as he points out that there will be fans in the midpoint. The value of this contribution needs to be acknowledged, though, first, because it is rooted in fan segmentation ideas, such us Redhead’s, and second, because Cleland clearly suggests the relevance of governance and the possibilities of categorising fans along these concepts. Similar to Cleland, Dubal (2010) also defines groups of fans in his comparative study of supporters’ reactions to the ‘neoliberalisation of football’ in Brazil and England. Dubal (2010: 138) differentiates between ‘Brazilian consumers’ and ‘English fans’. Dubal did not intend to produce a systematic typology of supporters, and the main division reminds us of some of the dichotomies mentioned above. However, his contribution clearly signals the academic interest in categorising supporters in relation to their governance opinions.
Whereas the literature on typologies of football supporters is substantial, it shares some limitations. First, models like those reviewed above tend to pay too much emphasis to locality by highlighting match attendance and close involvement with the club as an important factor in defining fandom. This is clearly not enough in view of the diversification of football clubs’ income in the last decade. Moreover, work has shown that increasing use of digital technologies by supporters can distort the boundaries between the local and the global (Millward, 2011). Second, whereas some of the most recent work points out the increasing salience of governance issues, they struggle to ‘capture the complexity of fan engagement in relation to governance’ (García and Welford, 2015: 527), because they also tend to address governance in a rather dichotomous approach. The rising importance of governance is suggested by some of the most recent work, but they do not present extensively elaborated definitions of governance, which then translates in obtaining relatively broad fan categorisations.
A final point of note is that these typologies are qualitative in nature. Most of the papers adopt a qualitative and inductive research strategy within a limited geographical scope, whose limitations need to be acknowledged. There is one exception, though, as Quick (2000) presents a quantitative segmentation of supporters in his marketing-based work. This typology was obtained in a different context, for Quick focuses his research on Aussie Rules football (also known as Australian football), the very popular Australian team sport that, despite its name similarity with soccer, is a different reality both in terms of play and organisation. Notwithstanding the different context, we take inspiration in Quick, for his work suggests that there is interest and academic credibility in adopting a quantitative approach to supporters’ typologies. Therefore, we argue that a quantitative approach could complement the existing qualitative focus and move this research agenda forward, especially in a pan-European context.
Although the geographical scope of the article is European, it is necessary to acknowledge, albeit briefly, the global nature of football. Perhaps one of the biggest recent expansions of football is in Asia, where the game developed commercially but, importantly, fans remain mostly oriented towards following European clubs (Cho, 2013). The marketing strategies of western media and football clubs have produced a form of fandom in Asia that focuses on the big European clubs and hampers involvement and interest with local competitions (Rowe and Gilmour, 2010). Before reaching the lucrative Asian market, football developed in Africa for quite some time, although in that continent there are not just commercial dynamics, but heavy historical, cultural and colonial factors (Darby, 2002; Martin, 1991). Fans of African football are known for their incredible passion and colourful displays. Although there are strong connections with local and national teams in African countries (see for example Chiweshe, 2011), the lure of European competitions is increasingly important for these supporters as well (Farred, 2002). This is different in the major Latin American countries. Before reaching other audiences, football developed strongly in Latin America. Fans there are not as globally minded as the Asians and their local competitions retain attention (Kittleson, 2014; Nadel, 2014). Given this devotion to local football, fans in Latin America have a history of protest against the establishment, be that clubs, federations or even government (Duke and Crolley, 2001; Moreira, 2017). Whereas there are very special local contexts, in that respect the Latin American fans (specially in Brazil and Argentina) share more similarities with the Europeans than the Asians do.
Against this backdrop, the paper’s main research aim is to design a quantitative governance-based typology of football supporters. Therefore, our first objective is finding out whether it is possible to obtain a statistically significant typology of supporters based on their governance preferences. Our second objective is to explore the socio-demographic and consumption patterns that characterise this typology. In completing these objectives, the article seeks to shed new light on the existing academic work studying the social dynamics of football fans.
The article proceeds now in five steps. First, we review existing academic literature to operationalise governance concepts. Second, we outline the methodology. Third, we present the results of our study. Fourth, we discuss the implications of our findings. Finally, the article concludes with a reflection on limitations and future research.
Systemic governance and good governance
Governance is defined as being concerned with ‘the patterns that emerge from the governing activities of social, political and administrative actors’ (Kooiman, 1993: 2). Van Kersbergen and Van Waarden (2004) identify up to nine different meanings of the term ‘governance’. The origin of this conceptual diversity lies in the fact that ‘definitions of governance depend largely on the respective research agendas of scholars or on the phenomenon that is being studied’ (Geeraert et al., 2014: 281).
Henry and Lee (2004) were among the first scholars to introduce the study of sports governance. They argue that sports governance can be encapsulated under two dimensions: systemic and good governance. The former refers to the structures that facilitate relations of stakeholders within the sports system. The latter refers to the normative principles that should define and govern those relationships within sport organisations. In this article we structure our investigation of supporters’ opinions on governance using these two pillars.
Systemic governance refers to a structure with a large number of stakeholders where power, authority and resources are diffused across the system (Henry and Lee, 2004). Football’s systemic governance resembles a network (Geeraert et al., 2013), with the interaction of governmental authorities and non-governmental organisations such as federations, leagues and clubs. European football’s network is composed by the European and international federations (UEFA and FIFA), the national federations, the national leagues and the clubs as primary stakeholders. This network is regulated by policies of national governments and the European Union (EU), which also need to be considered as stakeholders (García et al., 2011). European football is highly dependent on media companies’ investment, which are now also part of the network as secondary stakeholders (Chappelet and Kübler-Mabbot, 2008). Finally, supporters’ organisations are also incorporated as part of European football’s governance network.
Good governance is defined as involving the principles of effective, transparent and democratic management (Rhodes, 1997). The European Union Expert Group on Good Governance (2013: 5) defines good governance in sport as ‘the framework and culture within which a sports body sets policy, delivers its strategic objectives, engages with stakeholders, monitors performance, evaluates and manages risk, and reports to its constituents on its activities and progress’. Conceptually, good governance is formulated around normative principles that organisations are required to implement (Geeraert et al., 2014). Chappelet and Mrkonjic (2013), in their Basic Indicators for Better Governance in International Sport (BIBGIS), identified the principles of good governance as transparency, democracy, accountability, integrity and solidarity.
Accountability is defined as ‘a relationship between an actor and a forum, in which the actor has an obligation to explain and to justify his or her conduct’ (Bovens, 2007: 450). This is closely linked to the principle of transparency, defined as the ability of stakeholders to obtain information about the management and processes of the organisation (Henry and Lee, 2004). Democracy relates to the possibility of being represented and influencing those in office (March and Olsen, 1995). In the context of football clubs this is linked to the extent to which owners and directors consider the views of the fans. Finally, integrity and solidarity are defined as the structures that ensure a healthy development of an organisation and its environment (Chappelet and Mrkonjic, 2013).
Methodology
This article draws on data obtained through an online survey on football culture carried out as part of the FREE (Football Research in an Enlarged Europe) project. 1 The FREE Project research team initially targeted fans in nine European countries: Austria, Denmark, France, Germany, Italy, Poland, Spain, Turkey and the United Kingdom. These countries were selected to provide a broad range of regions and different football cultures, including north–south, east–west geographical scope as well as big and small leagues. The number of responses obtained in Denmark (226), Austria (145) and Italy (71) was deemed unsatisfactory for this article. Thus, the final sample used here includes France, Germany, Poland, Spain, Turkey and the United Kingdom, which still provide a north–south, east–west and small–big leagues coverage.
In order to distribute the survey as widely as possible, the FREE Project research consortium reached agreements with media outlets and fan groups in each of the countries. These project partners included a link to the survey in their websites and wrote posts explaining the objectives of the research. This was also disseminated through their social networks. Following a piloting phase, the survey was open to respondents from September 2013 to the end of February 2014. These dates were selected because they coincided with the initial stages of the football season. The selection of the dates might create a bias in the response that needs to be acknowledged, but we argue this is of a minor nature and, ultimately, data collection always needs to be done at a point in time.
The survey was part of FREE, a larger project investigating the sociocultural importance of football in Europe. Therefore, the questionnaire was not only focused on governance. It was structured around several sections, from identity to football consumption and family football traditions. It also included the main socio-demographic characteristics. The questionnaire included 68 questions, with two of them focused on governance (Q.64 and Q.65). These two questions comprised a total of 20 statements where participants had to reply using a five-point Likert scale (see Table 1). These two questions, with their 20 statements, provide the basis for this article.
Descriptive statistics.
Source: Own analysis with data from FREE Project online survey. SD = standard deviation.
Where 1= tend not to trust at all, 2=tend not to trust, 3=neither distrust nor trust, 4=tend to trust, and 5=tend to trust very much.
Where 1=strongly disagree, 2=disagree to an extent, 3=neither agree nor disagree, 4=agree to an extent, and 5=strongly agree.
When looking into supporters’ opinions of systemic governance, the survey builds on the notions of network governance, primary and secondary stakeholders. In line with the literature, it was considered that good governance structures are likely to yield trust and legitimacy. Therefore, the extent to which fans trust an organisation in the governance of football is used as a proxy for a positive or negative view of their structures. Hence, the survey asked participants about their level of trust in the stakeholders identified in European football’s governance network, including both primary and secondary stakeholders at national and supranational level. This is reflected in Q.65 (see Table 1). The other question relates to good governance. It intends to measure fans’ perceptions of the implementation of good governance principles at club level. The principles are operationalised through statements included in Q.64 (see Table 1). Democracy is encapsulated in statements 1 and 4; accountability is related to statements 2, 8 and 10; transparency is related to items 3 and 9; and, finally, items 5 and 6 relate to social responsibility.
The survey was administered online. This has several limitations that must be acknowledged due to the non-probabilistic nature of sampling, and thus not allowing to calculate sampling error. Another limitation was participant self-selection, as the research team did not have total control over who completed the questionnaire. To try and contain these limitations a large number of responses were collected, and the survey was widely distributed, as explained above. Another possible limitation is the fact that some members of the target population might not have access to the internet. However, given that the survey was administered in Europe, where internet coverage is wide, smartphone ownership is common, and there is also potential internet access in public places such as libraries or internet cafes, we feel this is an acceptable limitation.
Despite these limitations, the use of online surveys in the social sciences has grown due to their major advantages, and it is suggested that online surveys can offer the same level of rigour and quality than more traditional telephone or face-to-face surveys (Manfreda and Vehovar, 2008). Also, it is argued that the anonymity of online surveys provides ‘more frankness and honesty’ in the responses of the participants (Cashmore and Cleland, 2012: 376). With all this in mind, we would argue that the large number of responses provides a wide range of football fans’ opinions across the six countries, although we cannot claim it to be representative of all fans.
The dataset was subject to a process of data cleaning: responses that did not complete the socio-demographic variables were eliminated; similarly, responses that answered less than 80% of the questionnaire were also omitted. As a result of the data cleaning the number of valid questionnaires was 11,384 (3,490 in Poland; 3,120 in France; 1,804 in Turkey; 1,800 in Spain; 635 in UK; and 535 in Germany). However not all participants answered Q.64 and Q.65, the governance-related questions. Thus, the final sample used in this article is 7,360, distributed per country as follows: 1,699 in Poland; 2,234 in France; 1,274 in Turkey; 1,266 in Spain; 466 in UK; and 421 in Germany. The sample’s gender composition is 10.1% female, 89.9% male. The age distribution is: 27.1% below 21 years; 25.3% 21–25 years; 34.8% 26–40 years; 7.8% 41 to 50; and 5% over 50.
As with every online survey without an invited respondent list the research team had minimal control over the geographical location of the respondents. To ensure geographical coverage the questionnaire asked participants their place of habitual residence. This was used to ensure that the sample of each country was made up of people residing in that country. Due to the differences in sample sizes of the countries, the total data have been weighted considering the population over 15 years in each of them (Groves et al., 2009). This procedure adjusted the weight of each country in the overall survey scores according to the actual population of the country.
To fulfil the objectives of our research, principal component analysis (PCA) and cluster analysis were applied. The PCA analysis revealed the existence of a number of dimensions that simplify and structure fans’ opinions. Next, a typology of supporters was obtained by applying a cluster analysis to the principal components. Finally, we undertook bi-variate analysis of the clusters in relation to socio-demographic and consumption variables, complemented with chi-square test whose results, however, need to be considered indicative due to the non-probabilistic nature of the sample.
Results
Supporters’ perceptions of football governance are shown in Table 1, which presents the main descriptive statistics. The top half of the table shows the descriptors evaluating fans’ trust in stakeholders. The highest score is for fan groups/supporters’ organisations. A second group of (slightly less) trusted stakeholders includes international media, the EU and UEFA. Next, in order of trust are football leagues, national FAs, national media and FIFA. The lowest score in this question is for national governments.
The bottom half of Table 1 presents the descriptors used to evaluate fans’ perceptions of good governance at club level. Here there is much more difference between the highest and lowest scores, but of special importance is the fact that eight of the 10 descriptors reach values above 3.5. Given the way in which the questions were phrased, this indicates a negative view of governance standards.
Our statistical analysis strategy combines a PCA and cluster analysis. To implement the PCA, the extraction method was the analysis of principal components and the Varimax rotation using Kaiser’s method. The factor loading obtained in the rotated matrix allows us to establish the elements determining the structure of each component and their interpretation. We established a factor saturation criterion of scores above 0.45, which is the threshold accepted as statistically significant in the literature (Cea D’Ancona, 2002: 490). The analysis was significant (P < 0.001, Bartlett’s Test of Sphericity) with a sample fit of 0.793 (Kaiser-Meyer-Olkin Measure of Sampling Adequacy). This revealed the existence of six dimensions (with an eigenvalue higher than the unit), which explain 61.4% of the model’s variance, a percentage that can be considered satisfactory in social sciences (Hair et al., 2014: 107). Table 2 shows the saturation of the six dimensions resulting from our principal component analysis as well as the variance explained by each component.
Principal Component Analysis (PCA): Rotated Component Matrix.
Source: Own analysis with data from FREE Project online survey.
The first component resulting from the PCA relates to the degree of fans’ trust in the management of clubs. It is a bipolar component whose positive scores reflect a tendency to be critical towards the strategies of club owners. On the other hand, the negative scores reveal a tendency to trust and feel secure in relation to the decisions taken by owners/presidents. This component demonstrates the heterogeneous nature of supporters’ opinions. We would suggest that this component could be labelled distrust/trust in football clubs’ owners/presidents. The second component groups the level of fans’ trust in national football associations, football leagues and the national government. Considering the composition, we label this second PCA component trust in national sport governance bodies. The third component brings together the two international governing bodies introduced in our analysis: FIFA and UEFA. We label this component trust in international football bodies. The fourth component includes national and international media, as well as the EU. Given the fact that the EU role in sport governance has been mostly characterised as a ‘supervisor’ trying to control sport organisations (Geeraert et al., 2015), and the obvious role of the media in exposing football’s governance problems, we suggest that this component is best labelled trust in control and pressure agents. We continue now with the fifth PCA component. This moves to the area of good governance, for it brings together three items that were aimed at analysing fans’ perceptions of their own actions as active supporters. This could be conceptualised as the supporters’ feeling of their own efficacy. Since the three items in this component are phrased in a negative way, we consider that the best label for this component is supporters’ inefficacy. Finally, the sixth component brings together two items, one from each question, hence incorporating elements of both systemic and good governance. The two items describe, first, the level of trust in supporters’ organisations (item F.1) and, second, the feeling that supporters’ opinion should be considered by club managers (item F.2). This is a component that reflects the extent to which supporters consider themselves with a legitimate say on football. Thus, the most adequate label for this component is supporters’ voice.
Once we have identified the six PCA components, we move to obtain a typology through the cluster analysis technique. Hair et al. (2014: 446) recommend doing this in two stages: a first step to determine the number of clusters, and a second one to optimise the initial results. For the first stage we applied a hierarchical method (Ward’s method and square Euclidean distance) in order to determine the adequate number of clusters and their initial centres. In the second stage we applied the k-means optimisation method of the centres. Following these statistical techniques, we concluded that the most appropriate solution, considering the dendrogram and the sociological suitability of the resulting clusters, was a distribution in five groups.
We then applied a triple validation of the classification in five groups to ensure robustness. In the first validation test, the analysis of variance (ANOVA) revealed that variance between groups was higher than intra-group variance in the six PCA components introduced in the analysis (p < 0.001). Second, we carried a multiple discriminant analysis – using as dependent variable the five clusters obtained through the k-means method, and as independent variable the six PCA components. This enabled us to classify correctly 96.3% of the cases, which must be considered an extremely positive result. Our third and final validation consisted in examining the cluster differences considering the distance between the centroids. We used the Euclidean distance method, the most appropriate with metric variables (Hair et al., 2014: 431). Table 3 shows that the results are positive, as the clusters are sufficiently and clearly different from one another. This reinforces that a five-cluster division is appropriate.
Distances between final cluster centres.
Source: Own analysis based on data from FREE Project online survey.
Table 4 presents the final cluster centres and the percentage of cases in each cluster. The first cluster is characterised by a high level of trust in international football bodies and in national governance bodies. This level of confidence on the formal governance structures is combined with a neutral view on pressure agents and supporter groups and a positive, but rather modest, trust on club owners/presidents and supporters’ groups. Finally, this group of supporters is noticeable because it features the highest level of auto-inefficacy perception. These are supporters who believe the current governance structures, with the federations at the top, are best suited to ensure the future of the game while, at the same time, are of the opinion that as supporters they will have very little impact on how their clubs are managed. This cluster represents 20.3% of the supporters. They are confident in the system and could be labelled as institutionalists.
Final cluster centres.
Source: Own analysis with data from FREE Project online survey.
The second cluster presents contrasting characteristics. These fans feature the highest trust in control and pressure agents and high scores for trust in national governance bodies and clubs’ owners/presidents. However, they are extremely critical of the behaviour of FIFA and UEFA. When it comes to trust in supporters’ groups and perception of the supporters’ voice their scores lay in the middle range. This cluster is the largest, with 23.9% of the supporters; due to their high mistrust in international football bodies we have labelled it as moderns, for they present what we could conceptualise as a modern view in relation to the role of the public, the state and the media as facilitators of social progress.
The third cluster features the highest level of distrust in supporters’ groups and the second highest in football clubs’ owners/presidents. These fans also have a marked mistrust of international football governing bodies and of control and pressure agents. Moreover, they have a certain sense of auto-inefficacy as supporters, meaning they do not think supporters can make a real difference in the governance of football. In short, these fans present a high level of criticism towards every actor currently involved in the institutional governance of football in Europe. The only actors whom these fans do not percieve too negatively are national governing bodies. This cluster is the third largest, with 20% of the supporters; due to their low trust in stakeholders we have labelled it as critics.
The fourth cluster is mostly characterised by the highest trust in club owners and in supporters’ groups. These fans feature a marked mistrust of sport governance national bodies, international football governing bodies and control agents. These are, therefore, fans who are confident that owners would do the right thing for their club while, at the same time, considering that as fans they need to have a say on how the club is managed. When looking at wider governance structures, though, these supporters do not trust international football bodies and national governance bodies, although the scores are moderate in the former. On the other hand, the group features a moderate mistrust towards the control and pressure agents. These supporters seem to be able to discern between their outlook towards the club level and the wider governance structures of football. This cluster is the smallest, bringing together 17.4% of the supporters, and could be labelled club-militants, in the sense that they feel very much at ease with their club management and fellow supporters.
The fifth and final cluster is a stark contrast to the club-militants. These fans strongly believe their involvement can make a difference. Naturally, they tend to distrust club owners and sport governance national bodies (in both cases they feature the highest scores), while in contrast they show trust in international football bodies and control agents, although the latter is rather modest. This can be interpreted as a group of supporters who are very aware of their local/national context and hope that external stakeholders will help them in the search for change. These fans manifest a positive view of the supporters’ groups, but the level of confidence in them is neutral. This cluster represents 16.4% of the supporters and we have labelled them globalists because of their trust in the supranational organisations.
When analysing the differences amongst the clusters, we can see that the institutionalists have the highest level of trust in football stakeholders. Contrariwise, the critics, moderns and globalists present the most negative views of systemic governance. There are differences between these three, though. The moderns and globalists, while sharing an overall negative view of systemic governance, have different perspectives. The moderns tend to distrust more the international governing bodies, whereas the globalists focus their suspicions on the national level. The critics, however, have a negative view of both national and international governing bodies, although the scores for the national level are very close to being neutral.
Focusing on good governance, the institutionalists are the most sceptical cluster in relation to the efficacy of supporter participation in governance structures. In that respect they are in stark contrast with the globalists, who present the highest scores of the five clusters in support of the efficacy of fan involvement to improve good governance. Furthermore, the globalists and the critics share the most negative opinion of club owners. These fans have the worst perception of good governance in their football clubs, and they were also among the most critical in relation to systemic governance.
In general, the institutionalists and the critics could be considered as the most different groups in governance terms. Whereas the former trust firmly the current structures, the latter are very sceptical of all stakeholders, even of supporter participation. The globalists have a similar stance to the critics, but they differ in their relative trust of UEFA, FIFA and control agents. This support of international bodies and control agents needs to be seen as a cry for help in relation to governance at the national/club level, so overall the globalists also have a pessimistic view of good governance.
A socio-demographic analysis of the clusters is shown in Table 5. We can appreciate, first, that there are some gender balance differences between groups, with the highest level among the institutionalist and the lowest among the critics. We find marked differences in relation to age as well. The club-militants and globalists have a bigger presence among the youngest fans below 25, whereas the critics and the moderns have the highest presence in supporters over 50 years. The globalists are the biggest group among 21–25 years old supporters. We also find important differences in relation to education. The critics and the moderns have normally completed a higher education degree, whereas the institutionalists and the club-militants are more concentrated in secondary education. These results confirm that the most critical fans would be found among the most highly educated groups.
Socio-demographic profile of clusters.
Unit: Percentage. Percentages are rounded to just one decimal. Source: Own analysis with data from FREE Project online survey.
Besides the use of descriptive demographic variables, it is also possible to analyse the behaviour of the clusters in relation to football consumption and attendance to games. Table 6 presents the most relevant results in this area, all of them statistically significant according to the Chi-square analysis.
Consumption and participations patterns by cluster.
Unit: Percentage. Source: Own analysis with data from FREE Project online survey.
One of the main observations is that the club-militants and globalists share the highest consumption patterns, in some cases alongside the critics. Club-militants and globalists are the fans buying club merchandise more often. On the opposite side of the spectrum the institutionalists have a more reduced consumption. Club-militants and globalists are also the supporters who watch football on TV and discuss it on social media more often. When taken together, it is this triangle of the critics, club-militants and globalists who have a clear habit of being up to date with football news. This also suggests they are information-savvy and well-informed supporters.
Attending football games is traditionally seen as a sign of heavy engagement. In this area we find the club-militants, the moderns and the critics ahead of the other two clusters. The club-militants and the critics attend more home games and travel to follow their team more frequently, while the globalists buy more season tickets. Globalists also buy their club’s merchandise and collect club items more often, which is clearly associated with their higher socio-economic status, as we see in Table 5. Finally, the level of involvement of the globalists and club-militants is further reinforced by the fact that they are members of their club and members of a supporter’s club at a much higher rate than the other clusters.
Discussion
Our results reveal that it is possible to construct a statistically significant typology of football fans’ attitudes towards governance. We have uncovered the existence of five clusters: institutionalists, moderns, critics, club-militants and globalists; whereby the moderns, critics and globalists form a triangle with the most negative vision of football governance. The results reinforce that governance is now a salient topic for football supporters. This has been pointed out by the literature (Cleland, 2010; Numerato, 2018), but was not incorporated to earlier typologies. Our research can be seen as a quantitative advancement of earlier work, such as that of Giulianotti (2002), because it adds to the existing knowledge by detailing how different fans understand the football governance dimensions. This typology helps to appreciate the way in which the spaces of systemic and good governance are articulated among football fans. There are two groups, the critics and the globalists (closely followed by moderns) that demonstrate a clear preoccupation for governance issues. But, crucially, they do so in different ways. Whereas both critics and globalists share a negative view of club good governance (unlike the club-militants), they differ in their understanding of systemic governance. On the opposite side of the spectrum, the institutionalists are the only fan group with a firm positive view of both systemic and good governance.
These results confirm that existing supporter typologies need updating, for they do not capture properly the evolving dynamics of football fandom. Existing typologies need to be superseded in favour of more complex models with new dimensions. In line with García and Welford (2015), we argue that governance/activism can be one of these new dimensions in the study of football fans because there are statistically identifiable patterns that relate attitudes towards governance with more traditional understandings of fandom. Our approach offers new areas of interpretation, but also helps to quantify the reality of these groups, complementing the existing qualitative focus in this area. One of the advantages of this typology is that it offers a holistic vision of European football culture that is not anchored in the localism of existing work. Thus, the results are a better reflection of the heterogeneity and diversity of football culture in Europe, something that is not always possible to capture with a body of literature mostly based on qualitative research designs.
We have also seen that it is possible to analyse the clusters in relation to behavioural patterns. The critics, club-militants and globalists form a triangle of the most critical supporters, but they are also the fans who spend more time and money in engaging with their club. The critics are among those who attend home and away matches more often. The globalists are the cluster with more season ticket holders. Besides attending games, the globalist fans are the supporters who buy merchandise and related products more often, closely followed, again, by the critics. We can see, therefore, a clear pattern in which the most critical fans are, at the same time, the most involved with their club and, also, the supporters spending more money and consuming more football. This clearly problematises some of the existing typologies, which separated fans and consumers imposing normative models from the outside. Dubal (2010), for example, made a clear differentiation between engaged fans and consumers. Similarly, Giulianotti (2002: 26–31) suggested a tendency towards a ‘detached, cool, consumer-orientated [fan] identification’ with a thin solidarity, which he labelled as ‘flâneur’. Our typology suggests that it is more complex than that. Actually, the most involved and dedicated fans (the true fans, if we were to use the terminology of the existing typologies) are also those who consume more.
This demonstrates the complexity of football fandom. Although it might seem counterintuitive, the most loyal supporters (in terms of active engagement and consumption) are also the most critical ones. In that respect, our research advances that of Cleland (2010), who suggested the importance of governance for supporters, but did not analyse further their behaviour and engagement in other areas. Supporters are acutely aware of the governance problems of their clubs because they engage with them almost daily. This suggests that the loyalty bonds between fans and clubs are extremely hard to break. Paraphrasing Hirschman (1970), this would suggest that negative opinions on governance might translate into protest (voice), but will not necessarily force active disengagement (exit).
These results are also relevant for club owners and wider discussions on football governance. Clubs should realise the need to build bridges with these critical groups, so they can rely on their continued business and support. The more owners can ensure critics and globalists do not slide into the institutionalist category in terms of engagement and consumption, the better. Should owners approach constructively these groups, they would minimise risks that their critical views could end up alienating them.
Conclusion
This article explores supporters’ attitudes to governance through the design of a quantitative pan-European typology. The picture revealed here is complex, as our results expose crudely the dilemmas of football supporters. This leads us to agree with calls to question understanding supporters only in terms of authenticity and modernity/consumerism (Millward, 2012). The article suggests that work on fan typologies needs to be open to different considerations that have gained salience among fans over the last decade. Governance is one of them. The significance of this research lies in opening new avenues for that body of academic work, while at the same time also articulating a more complex understanding of the different dimensions of governance from the fans’ point of view.
We would argue that our quantitative pan-European approach provides new insights, if only to quantify what other authors had suggested conceptually. We advocate for collaborative and complementary approaches in this area of academic work. We suggest that multi-disciplinary enquiry that departs from normative views of football fandom will enhance understanding of this complex reality. This ‘governance turn’, we argue, has the potential to further our understanding of the complex reality of football supporters, and we invite academic colleagues to be aware of it.
Finally, it is necessary to reiterate some of the limitations of this article before concluding. The non-probabilistic nature of the sample due to the use of an online survey needs to be acknowledged. As pointed out in the methodology section, online surveys are increasingly used in the social sciences, including many recent articles published in this journal. The advantages of this method do not eliminate biases in terms of coverage and the impossibility of calculating sample error. However, we feel that the limitations have been duly mitigated and the results are relevant and help us to move our understanding of football fans forward.
Footnotes
Acknowledgements
The authors would like to acknowledge the three anonymous reviewers and the journal editor for their useful feedback comments on previous versions of this article. The authors also would like to acknowledge the contribution of Anni Peckie to this research.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the European Union’s 7th Framework Programme for research, technological development & demonstration (FP 7) [Grant Number 290805].
