Estimating the market value of collegiate football players from professional factor shares

ABSTRACT

Although the level and growth of revenue in college football has drawn increasing public attention, reliable estimates of the value of college and associated rents remain elusive. Many players do not have easily observable performance statistics and there are significant complementarities. Using a simple estimation strategy based on professional factor shares and rooted in economic theory, we generate estimates of market value for college players across seven National Collegiate Athletic Association conferences using professional football and basketball salaries from 2011 to 2013 as shadow prices. With this method, market value of an average starting football player in a Bowl Championship Series (BCS) conference ranges from $120,000 to $1.7 million per year. The value of superstar players may exceed $4 million per year. We find that rents can flow from the institutions to players in some cases at relatively low-revenue institutions outside the BCS.

KEYWORDS:

• College sports
• shadow prices
• marginal revenue product (MRP)

JEL CLASSIFICATION:

• J24
• L30
• L83

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Fleisher, Goff, and Tollison (1992), Quirk and Fort (1992), Kahn (1992) and Zimbalist (1999) provide detailed discussions of the structure of various sports associations. See also Barro (2002) and Becker (1987).

² These data are from http://sports.usatoday.com/ncaa/finances/

³ For instance, see O’Bannon, Jr. v. National Collegiate Athletic Association et al. and the unsuccessful attempt by players at Northwestern University for recognition as a labour union by the National Labor Relations Board.

⁴ While Scully (1974) is the original contribution, he expands on these methods in Scully (1989, 1995). The method has been adjusted and altered, but has remained similar in its methodology.

⁵ Using performance regression methods, Brown (1993, 1994) and Brown and Jewell (2006), the average market value of high-quality college football players was estimated to be to be $495 000, $1.4 million for men’s basketball players and $263 000 for women’s basketball players. Kahn (2007) provides a summary of estimates of impacts of college athletes and related issues. Zimbalist (1999) and Fleisher, Goff, and Tollison (1992) are earlier, broader treatments.

⁶ Berri and Schmidt (2010) note large interaction effects within football due to season-to-season inconsistencies. The interaction effects are much lower in basketball due to much higher consistencies over time.

⁷ The ε and ν are assumed to be randomized error terms; however, systematic industry over or under valuing of positions is possible (see Berri and Schmidt 2010).

⁸ McDonald and Solow (1981), Arpaia, Perez, and Pichelmann (2009) and Cahuc and Zylberberg (2004) provide theoretical discussions. Bentolila and Saint-Paul (2003) find empirical support for the importance of bargaining power in OECD countries. In a review of several empirical estimates, Dickens (1995) finds wages generally close to marginal product. Manning (2010) finds a wide range of outcomes in his survey of evidence.

⁹ Each league defines their sport-related revenues with slight differences. Owners have a significant incentive to bargain over such small percentages. Due to the relatively small number of owners, increased percentages of revenue can make several million dollars of difference to each owner. MLB shares were near 50% until recently. For more discussion, see Zimbalist (2010).

¹⁰ This compares to Brown’s (1993) estimates for elite, future professional football players of $495 000. However, Brown’s estimates are not position specific and do not apply to those not drafted to the professional level.

¹¹ Kahn (2007) discuss the problems of using university budgetary figures to estimate economic costs attributable to athletics.

¹² An extensive literature exists regarding employer training. Acemoglu and Pischke (1999) summarize some of the more complex models of on-the-job training.