Analysing match attendance in the European Rugby Cup: Does uncertainty of outcome matter in a multinational tournament?

ABSTRACT

Research question: The paper analyses the determinants of attendances in rugby union’s European Rugby Cup (ERC) and, in particular, whether uncertainty of outcome had a significant effect on spectator demand.

Research methods: We use data on attendance from 1226 ERC group stage matches played between 1995/1996 and 2013/2014 to estimate a model of spectator demand for ERC matches.

Results and findings: We find that short-run (match) uncertainty had little effect on attendances, whereas the strength of the home team had a significant effect on attendances. Medium-term uncertainty, that is, the possibility of the home team reaching the knock-out stages, had a significant impact on attendances.

Implications: The findings are significant because the ERC was replaced with a new European Rugby Champions Cup (ERCC) from the beginning of the 2014/2015 season. Changes in the distribution of revenue between participating teams combined with increases in English and French teams’ domestic league broadcast revenues suggest that the ERCC may be more unbalanced than the ERC. This has implications for the business models of clubs from smaller countries and their ability to compete and may lead to the emergence of an Anglo-French duopoly in the new competition.

KEYWORDS:

• Professional team sports
• competitive balance
• uncertainty of outcome
• league structure
• consumer demand

Acknowledgements

We are grateful for helpful comments provided by two anonymous referees. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The issue of professionalism first surfaced in rugby in the late nineteenth century and ultimately led to a group of clubs mainly based in the industrialised North of England breaking away and establishing the sport of Rugby League. Dobson et al. (2001) reported that prior to 1995 all but one of the 32 teams in the English Rugby League was located in the North of England. Although Rugby Union removed its ban on professionalism in 1995, the two sports have remained separate. The term rugby is used in the balance of the paper to refer to Rugby Union unless otherwise stated.

2. The decision by the IRFU to centrally contract players appears to have been prompted by concerns that there would have been an exodus of top players to overseas leagues.

3. Traditionally the Irish provincial teams played each other in a provincial championship but this only involved three matches per season.

4. There have been no Romanian participants in the ERC since its first season.

5. EU Commission, Joint Selling of the Commercial Rights of the UEFA Champions League, Decision of 23 July 2003, OJ/L291/25.

6. Five points are scored for a try (touchdown) and a team gets an opportunity to score a further two points by kicking the ball over the crossbar between the uprights. The ERC and the three major European leagues award a losing bonus point when a team loses a match by seven points or less.

7. 1995/1996 is excluded as we have no measure of team quality for the first season of the competition. Our data covers 1226 of the 1228 ERC group matches played over the 18 seasons from 1996/1997 to 2013/2014.

8. If the chance of qualification were to fall from certainty to impossibility, attendance would fall by 3127 individuals. Given that the average attendance in the sample was 9111, the magnitude of this effect is similar to the 43% change predicted by the log model (see Table 3).

9. We tested for cubic terms but they proved insignificant so we dropped them for subsequent regressions.

10. If the interaction term is dropped then away team points variables become jointly significant (p-value of .0001).

11. This may also reflect an attenuation in simultaneity bias. Over time better supported teams will be able to sign better players so that attendance will influence ranking points (and the winner dummies). Arguably this is partly offset by the way ERC revenue was distributed in the past. In addition the Irish, Scots and Welsh teams receive financial support from their national associations out of revenue generated by international matches. Including team fixed effects would capture some of this reverse causality.

12. It is possible that some of the insignificance of the ranking variables could be due to multicollinearity. We examined this issue by dropping the insignificant absolute difference variable. This had virtually no impact on the two rest of the coefficients. As there is a sound prior argument for including the absolute difference in ranks, we do not include these other regressions in the paper.

13. In 2012/2013, 63 of the 72 group matches were broadcast live. Source: Retrieved from http://www.ercrugby.com/eng/erc/about/index.php. Admittedly fewer matches were broadcast live in the early years of the competition.

14. P_i is team is total points and P_q is the total points of the team occupying the relevant qualifying positions. Q has a value of 1 for all teams prior to the first round of matches.

15. An anonymous referee pointed out that our measure, Q, is similar to one developed by Janssens and Késenne (1987).