Contribution
The first survey measure, by the paper’s account, of identification with a European community of football fans as distinct from club and national fan communities. It shows the three are related but that the European one separates from the other two in a factor analysis. It reports what the European identification goes with, among people already sampled as fans.
Key insight
“The fandom” has more than one level. Fans of rival clubs can still answer yes to “I consider myself a part of a bigger community of fans in Europe”, and that answer goes with trusting fans from elsewhere in Europe, not with liking the institutions that run European football. The community a fan feels part of and the governance a fan criticizes come apart.
Assumptions
- Social identity in Tajfel’s sense, with Ashmore et al.’s multidimensional components (p. 2), and the common in-group identity model (CIIM): contact under equal status, cooperation and common goals recategorizes subgroups into a superordinate group, possibly with a “dual identity” (Dovidio et al. 2003; pp. 2–4).
- The sample is fans. Of 7,040 respondents, those who self-categorized as fans, or scored football interest ≥ 7/10, were kept: 3,685 cases (p. 4). The panel is a commercial online panel with quotas on gender, age and education; the paper calls it representative.
- One item per level. The dependent variable is agreement (0–10) with “I consider myself a part of a bigger community of fans in Europe”, with parallel single items for club and nation (p. 4).
Key results
- Levels (Table 3, p. 6). Means are club 6.13 (SD 2.50), nation 5.57 (2.58) and Europe 5.29 (2.61). Correlations are nation–Europe .78, club–nation .69 and club–Europe .58. The text reports 1,962 respondents (49.3%) at ≥ 6 for Europe and 833 (20.93%) at ≥ 8. Those percentages imply a denominator of about 3,980, not the 3,685 analysed; the paper does not say which.
- Dimensionality (Table 4, p. 7). A one-factor solution explains 70% of shared variance, with loadings .72–.96. A two-factor solution explains 78%: club and nation load on one factor, Europe alone on the other (.931). The nation loading on factor 1 is 1.011 with uniqueness .012, a Heywood case the paper does not remark on. A three-factor solution “fails to provide additional explanatory value”.
- Correlates (Table 5, OLS, R² = .37, n = 3,685).
- Positive: trust in a European fan from another country (0.336, the largest); watching foreign leagues (occasionally 0.55, regularly 0.80, very often 1.19, against rarely); positive perceptions of fans (0.231); attachment to Europe (0.159); fan-interest representation (0.112); football interest (0.098); watching European competitions regardless of club (0.195); medium (0.37) and high (0.27) education against low; right-leaning self-placement (0.045); interpersonal trust (0.040).
- Negative: age (−0.020 a year); male (−0.197); Germany (−0.47) and Norway (−0.71) against Spain; cosmopolitan values (−0.056).
- Null: income; negative perceptions of fans; football universalism; critique of football governance.
- Interpretation (pp. 8–10).
- The European identification is “not simply an additive extension” of club or national attachments but “a related, but qualitatively distinct form of belonging”, consistent with CIIM and a “marble cake” of nested identities.
- The null for governance critique suggests “a conceptual separation between the community of fans and the institutions that regulate football”.
- The paper flags its own limits: no causal inference; four countries; no direct measure of cross-border contact.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | A substantial share of European football fans identify with a European community of fans (about half at ≥ 6/10) | moderate: a large quota sample, but one item, and an unexplained denominator for the percentages | Table 3, p. 6 |
| C2 | European fan-community identification is a distinct dimension from club and national fan identification | weak to moderate: rests on three single items; the two-factor solution has a Heywood case, and a factor defined by one item is not identified as a construct | Table 4, p. 7 |
| C3 | It is most strongly associated with particularized trust in European fans, and with cross-border viewing | moderate as association; direction not established (p. 10) | Table 5 |
| C4 | It is not associated with critique of football governance: fans separate their community from the institutions | weak: a null in a multivariate model with many correlated predictors | Table 5, p. 9 |
| C5 | The Europeanisation of football produced this identification (via contact, per CIIM) | weak: interpretive; no contact measure, cross-sectional | pp. 10–11 |
Concepts
- European fan community identification: the single item above.
- Europeanisation (of football): changes in football’s institutions and practices attributable to European integration. It runs top-down (downloading), bottom-up (uploading) and horizontally (crossloading) (p. 4).
- particularized trust: trust directed at a defined in-group, here “a hypothetical football fan that is a European citizen not from the respondent’s home country” (p. 5).
- dual identity / marble cake: holding superordinate and subgroup identities at once (Dovidio et al.; Risse 2015).
Connections
- Reimer et al. (NOTE-1251). Their chapter’s common in-group and dual identity models are the frame here. Their caution, that recategorization can subsume minorities and that in-group projection makes the majority’s attributes the default, is not taken up. The finding that identification is lower among fans of the dominant German league is suggestive and unexplained.
- Self-categorization theory (THEORY-1114). SCT predicts the European level becomes salient when the comparison frame includes non-Europeans or non-fans. The survey measures standing identification with no frame manipulated, so it cannot test that.
- Giulianotti (NOTE-1244). Cited (p. 3) for the varied modes of fandom. The heavy-foreign-viewing correlate fits Giulianotti’s “followers” and “flâneurs” more than his “supporters”; the paper does not draw this.
- Pettigrew (LIT-1408), unread, is the contact theory behind CIIM.
Bearing on the record
- Source of THEORY-1116 (Proposed): identification with a superordinate fan community, here European, coexists with club identification, goes with particularized trust and cross-border viewing, and is independent of attitudes to the governing institutions.
- CASE-104 and QUESTION-119. The governance null bears on the case (fans “hating” an owner) only obliquely. It suggests that judgment of a sport’s institutions and identification with fellow fans are separate attitudes, which is A2’s reason for not reading one off the other. That is a European-level finding and an inference to the club level.
- “The fandom” has no single boundary. It adds evidence for the exchange’s point that a fandom has nested scales.
- No ML instruction.
Limitations
- Single-item measures for all three identification levels. The “distinct dimension” claim is a factor analysis of three items, and the two-factor solution includes an out-of-bounds loading.
- Cross-sectional associations, as the paper says.
- The percentages do not match the analysed n.
- Fans only: identification relative to non-fans is not measured, and the inclusion rule mixes self-categorized fans with the merely interested.
- Four countries; country effects (Germany, Norway lower) are explained post hoc.
- The significance note in Table 5 marks p < .1 with one star while the text treats p = .031 as significant. This is a labelling slip, not a substantive one.
Open questions
- Does the European factor survive multi-item measurement, and does it shift when the comparison frame is manipulated (fans of European versus non-European football; fans versus non-fans)? That would connect the survey to SCT’s prediction.
- Does cross-border contact precede the identification? A panel design or a contact measure would settle direction.