Contribution
A single structural model that holds rivalry’s identity-threatening and identity-enhancing sides together, where earlier work had studied the threat (blasting, schadenfreude, derogation). It introduces perceived reciprocity of the rivalry, whether “the own team serves as the ‘rival for the rival’”, as a mediator, and finds it carries nearly all of the positive effects.
Key insight
A rival costs fans the respect of one out-group and buys them distinctiveness, cohesion and standing with everyone else. What makes the purchase pay is not hating the rival more but believing the rival hates you back. An unrequited rivalry, where the other side shrugs, is the one that does not help.
Assumptions
- Social identity theory (Tajfel & Turner 1979): a salient out-group is needed for a positive social identity. Optimal distinctiveness (Pickett & Brewer 2001). Disidentification (Elsbach & Bhattacharya 2001). Public collective self-esteem (Luhtanen & Crocker 1992), split here by reference group: the rival’s fans (PCSE-R) versus other clubs’ fans (PCSE-NR).
- Rivalry is subjective: “a subjective competitive relationship that is independent of the objective characteristics of the situation” (Kilduff et al. 2010; p. 14).
- The causal order is the model’s, not the data’s. Rivalry intensity → mediators → self-concept, with team identification as a control.
Key results
- Sample (pp. 13–15).
- December 2014, 25 interviewers at home games of eight clubs. Four had famous rivalries (Dortmund, Schalke, Cologne, Mönchengladbach) and four did not (Paderborn, Hoffenheim, Freiburg, Wolfsburg).
- N = 748 intercepted (mean age 36.1, 78.6% male). 164 named no archrival, 147 of them from the four no-famous-rivalry clubs, and were dropped. 13 more were dropped for missing data, giving N = 571.
- Measures (pp. 15–16).
- Rivalry intensity is one item on a 1–10 scale (M = 8.50, SD 1.82).
- Multi-item 7-point scales: identification (α .73), disidentification (.85), reciprocity (.81, self-developed), distinctiveness (.85), cohesion (.88), PCSE-R (M 2.27, α .82) and PCSE-NR (M 4.75, α .93).
- Identification’s AVE is .44, below the .50 threshold.
- Fit. Measurement model: χ²/df 2.46, RMSEA .051, CFI .96. Structural model: χ²/df 3.30, RMSEA .063, CFI .93, TLI .91, which the authors call “acceptable”. A latent method factor took 9.6% of variance and was retained.
- Effects (Table 2, standardized).
- Totals: PCSE-R −.31***; PCSE-NR .09*; distinctiveness .27***; cohesion .20***; disidentification .18***; reciprocity .49***.
- Direct (all ns): PCSE-NR −.08; distinctiveness .04; cohesion .01.
- Indirect via reciprocity: PCSE-NR .18***; distinctiveness .20***; cohesion .18***.
- Indirect via disidentification: .00 (ns); .03*; .01 (ns).
- R² ranges .10–.31.
- Reciprocity’s indirect effects exceed disidentification’s for PCSE-NR and cohesion (p < .01), not for distinctiveness.
- Alternative models (e.g. disidentification as outcome) fit worse; they are not reported in detail (p. 26).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Fans who perceive a more intense rivalry feel less respected by the rival’s fans | moderate: a large effect in one cross-sectional sample; consistent with prior work | Table 2 |
| C2 | Fans who perceive a more intense rivalry perceive their fan community as more distinct and cohesive, and as more respected by other clubs’ fans | moderate for distinctiveness and cohesion; weak for respect from others (total .09, p = .033) | Table 2 |
| C3 | These positive associations are carried by perceived reciprocity of the rivalry, not by disidentification | moderate as a statistical mediation in one model; causal order assumed; reciprocity is a self-developed scale | Table 2, pp. 19–20 |
| C4 | Disidentification with the rival contributes little to fans’ self-concept beyond reciprocity | weak: small indirect effects; the authors’ explanation (ceiling from high identification) is speculative | pp. 22–23 |
Concepts
- perceived rivalry intensity: one item, how intense the rivalry with the named archrival is (1–10).
- perceived reciprocity of rivalry: the extent to which fans perceive the rivalry as mutual, the own team as “rival for the rival” (p. 12).
- disidentification: cognitive separation of one’s identity from an organization’s (“the archrival’s failures are my successes”).
- PCSE-R / PCSE-NR: public collective self-esteem, how one believes the rival’s fans, or other clubs’ fans, regard one’s fan group.
Connections
- Hornsey 2008 (NOTE-1253) and Reimer et al. (NOTE-1251). The paper is SIT-framed: positive distinctiveness and collective self-esteem. Hornsey notes that self-esteem hypotheses became a contested strand. Here “self-esteem” is public collective self-esteem, others’ perceived regard, not the personal self-esteem those hypotheses measured.
- Self-categorization theory (THEORY-1114). The PCSE split (respected by the rival, not; by others, yes) fits SCT’s claim that the meaning of an in-group depends on the comparison: the same fan group evaluates itself differently against different out-groups. The paper does not invoke SCT.
- Talarico & Moore (LIT-1125) and Qi et al. (LIT-1131) were filed for the same case and are not read.
Bearing on the record
- Source of THEORY-1113 (Proposed): rivalry raises fans’ in-group distinctiveness and cohesion through perceived reciprocity.
- CASE-098. The case says this paper is “evidence about identity, not about the rivalry as a unit”. Read, it says a little more. The benefit to fans depends on perceived mutuality, a relational property, more than on fans’ own antagonism (disidentification). That is still evidence about individual perception, not about the rivalry as a collective unit. But it shows fans’ identity is tied to how they think the other side sees the relation, which is the two-sidedness the case’s “institutionalized rivalry” needs. A note is added to the case.
- The 164 excluded fans (no archrival, mostly at clubs without famous rivalries) are a finding the paper does not use: about 22% of intercepted fans named no rival at all.
- No ML instruction. The managerial advice (promote perceived reciprocity, co-market rivalries) is sport marketing.
Limitations
- Cross-sectional and correlational, as the authors say; mediation is modelled, not identified.
- A single-item predictor with a ceiling (mean 8.5 of 10).
- Selection: fans with no rival are excluded, so rivalry intensity is only varied among fans who have one.
- One league, one month, stadium attendees.
- The reciprocity scale is new and validated only here.
- Structural fit is borderline (CFI .93, TLI .91).
Open questions
- Does manipulating perceived reciprocity (for instance, exposure to the rival’s fans dismissing or acknowledging the rivalry) change distinctiveness and cohesion? An experiment would test the mediation causally.
- Is reciprocity perceived accurately? Pairing surveys of both clubs’ fans would show whether perceived mutuality tracks actual mutuality, which is the relational fact CASE-098 needs.