Cost spreading in college athletic spending in the United States: estimates and implications

ABSTRACT

With rising costs, mounting student debt, and many schools experiencing financial hardship, the higher education industry faces unwanted scrutiny from the popular media and political sector. College athletics too have come under close examination because of rising costs and internal subsidies. In this paper, we provide estimates of the per-student costs of college athletic programs for US colleges and universities by the number of undergraduate students enrolled, National Collegiate Athletic Association division, and whether the institution is public or private. These estimates find significant potential for cost spreading, so that costs per-student fall as the number of students rises.

KEYWORDS:

• Cost spreading
• fixed costs
• higher education
• college athletics

Acknowledgement

The authors thank Frank Mixon, Tobin Turner, participants at a session of the 2015 Missouri Valley Economic Association meeting, two anonymous referees, and the editor of this journal for helpful comments on earlier drafts. We are responsible for any remaining errors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The Power Five conferences are the Atlantic Coast Conference, the Big Ten, the Big Twelve, the Pacific Twelve, and the Southeastern Conference. The schools in these conferences generally have the most prestigious athletic programs in the country.

2. In the US, the terms ‘college’ and ‘university’ are used interchangeably because they both refer to institutions of higher education.

3. The NCAA divides Division 1 schools into two subsets, based on the scale of their football programs. Schools with large-scale football programs are known as Football Bowl Series (FBS) schools, and schools with smaller scale football programs are known as Football Championship Series (FCS) schools. Throughout this paper, we will refer to these schools by their former designation as D1A and D1AA, respectively.

4. As Shulman and Bowen (2001) acknowledge, the EADA data may face limitations and shortcomings, but they are the best available, and improvements should make them better each year.

5. Division 3 schools offer no athletic scholarships.

6. It is possible that the number of students is a function of athletic expenditures. We ignore this possibility in our specification because if other schools follow suit with increased athletic expenditures and the demand for higher education is inelastic, the effect of athletic expenditures on enrollment is likely to be small.

7. The Breusch-Pagan Lagrangian Multiplier test for random effects indicates the random effects model is preferred to simple OLS (χ²=14,720).

8. The log-log form provides the best overall fit and significance, but results are qualitatively the same in the reciprocal form and when estimates include squared values of total undergraduates and total athletes.

9. We note that the simple correlation coefficients between total athletes and NCAA division are small, ranging from −0.01 to 0.38.

10. e^0.513−1=0.670.

11. Complete results may be obtained from the authors upon request.

12. In the figures, PR designates private, PU designates public, F indicates a football program, and NF indicates no football program. All D1 and D1AA schools in this sample have football programs.

13. e^0.568− 1 = 0.764.

14. Complete results may be obtained from the authors upon request.