Live Broadcasting, Gate Revenue, and Football Club Performance: Some Evidence

Abstract

This paper presents new evidence concerning the main argument of the English Premier League for collective arrangements in selling television broadcast rights. Matches that are broadcast live on television can be considered as a substitute for watching at the stadium. New panel-data show that broadcasting live has a small negative effect on gate revenue for the best performing clubs and a much larger effect for the worst performing clubs. However, the marginal cost is outweighed by the marginal revenue, concluding that restrictions on quantity stemming from collective arrangements are detrimental to the consumers, broadcasters, and clubs themselves.

Key Words:

• Premier League
• Football
• Attendance
• Broadcasting Rights
• Collective Selling

JEL classifications:

• D43
• D71
• K21
• L13
• L83
• L88

Notes

12001–2004 contract was for 66 games and up to 40 on a pay-per-view basis (PremPlus).

22004–2007 contract was for 88 games and up to 50 on a pay-per-view basis (PremPlus).

3At the time, there were no data on how much ESPN bought Setanta’s rights for. Source: Adapted from OFCOM.

1Excludes payments for international broadcast rights sales. Source: Adapted from Premier League Annual Report 2008.

1. Factors such as a team’s allocation or uncertainty of the match outcome are researched by Buraimo et al. (2009), Dobson et al. (2001), Forrest et al. (2004), Noll (1974), amongst others.

2. In a study of the Spanish football league, Garcia and Rodriguez (2002) find evidence of a significant reduction in stadium attendance when a match is televised. The effect is increased for matches shown on free-to-view terrestrial television.

3. By Carmichael et al. (1999), Forrest and Simmons (2002), and Forrest et al. (2004).

4. Of Neale (1964), Rottenberg (1956), and Sloane (1971).

5. To allow for a more flexible functional form, this paper uses the interaction of the TV variable with time-varying regressors. This is particularly valuable in a fixed effects model, as unchanging variables will be otherwise washed out by time de-meaning. This method does change the interpretation of the coefficients. It would now represent the marginal effect of the time-invariant regressor evaluated when the match is broadcast.

6. Number of broadcast matches 2007–2008 season: Manchester United, 25; Arsenal, 23; Liverpool, 21; Chelsea 18.