Determinants of Coaching Salaries in College Football Programs: Are There Differences across Football Bowl Subdivision Schools? ()
1. Introduction
According to the 2019-2020 National Collegiate Athletics Association (NCAA) Financial Reports (2020), the revenue sources of the NCAA Division-I Football Bowl Subdivision (FBS) public institutions’ athletics departments were as follows: Donor contributions (20%), institutional/government support (18%), ticket sales (17%), NCAA/conference distributions, media rights, and post-football (14%), student fees (14%), corporate sponsorship, advertising, and licensing (10%), other revenue (6%), and competition guarantees (2%). In contrast, the revenue sources of the NCAA Division I Football Championship Subdivision (FCS) public institutions’ athletics departments showed a quite different picture: institutional/government support (53%), student fees (21%), donor contributions (7%), competition guarantees (6%), and NCAA/conference distributions, media rights, and post-football (5%). One significant point needed to make here is that more than 70% of the total revenues of the NCAA Division-I schools relied on public money and student fees (Jewell, 2020). This fact relates to important social agenda such as equity and justice in the distribution of financial resources. When it comes to financial sustainability, the college football financial structure currently employed is not plausible, thus, more sustainable and equitable financial models, thus, should be developed and implemented.
A large portion of the revenue is concentrated among a few top programs, primarily those in the ACC, Big Ten, Big 12, Pac-12, and SEC conferences. These conferences receive the majority of the revenue distribution from the College Football Playoff (The Knight Commission on Intercollegiate Athletics, 2023). The financial projections through 2032 suggest that if current spending practices continue, the gap between revenue and expenses will widen, necessitating a reevaluation of how funds are allocated (The Knight Commission on Intercollegiate Athletics, 2023).
The expenses of NCAA Division-I FBS public institutions’ athletics departments are as follows: Coaches compensation (21%), facilities and equipment (21%), support and admin compensation (18%), athletic student aid (14%), game expenses and travel (10%), other expenses (10%), competition guarantees (2%), medical (1%), recruiting (1%), and transfers back to the institution (0%). As expected, the expenses of the NCAA Division-I FCS public institutions are similar to the peer FBS institutions: athletic student aid (29%), coaches compensation (21%), facilities and equipment (14%), support and admin compensation (14%), game expenses and travel (10%), other expenses (7%), medical (2%), recruiting (1%), competition guarantees (0%), and transfers back to the institution (0%).
In 2019, only 25 out of 130 colleges in FBS reported positive net revenues while the median athletic program in FBS showed an operating deficit of $18.8 million. FCS colleges also reported a median operating deficit of $14.3 million. The persistent deficits led colleges to increase student fees and contribute to tuition increases. Increasing student fees and tuition is also closely tied to increasing cost of attending college and student debt, estimating over $1.6 trillion in 2021. Research indicates a growing disparity between programs in Power Five conferences and those in smaller conferences. Power Five schools generate significantly higher revenues than other FBS and FCS programs, leading to disparity in facilities, recruiting budgets, and overall competitiveness.
A comprehensive body of literature explored the financial implications of college football, encompassing various aspects such as media revenue, ticket sales, and sponsorship, including advertising. Major conferences, including the Power Five conferences, had lucrative media contracts that provide member institutions with tens of millions of dollars annually, significantly impacting their budgets and allowing for increased investment in facilities and staffing. The revenue of the Power Five conference consistently grew by 10% per year between 2013-2022 with television rights generating nearly two-thirds of that revenue in 2022. The rest of the revenue was generated from postseason bowl games (21.2%), distributions from the NCAA (8.9%.), and other sources (4.2%). In 2022, the Power Five conferences distributed 91.0% of their revenues to member schools (Morones & Heide, 2023).
However, coaching salaries of college football have rarely been addressed. When it comes to the issues in college football in the US, coaching salaries have been discussed as one of controversial topics. Coaching salaries are normally determined by the amount of revenue the football programs make, meaning that they heavily rely on a variety of revenue sources such as ticket sales, media contracts, sponsorships, donations, and gifts. Colbert and Eckerd (2015) investigated the correlation between football coaching salaries and success of the programs and found that higher coaching salaries are associated with greater success. Zimbalist (2018, 2023) pointed out that coaching salaries have escalated dramatically, consuming a substantial portion of athletic budgets. It is not difficult to find stark polarization across conferences and football programs. Some colleges with huge markets and alumni can afford the salaries, however, most colleges are not able to make enough revenues to afford it. While some people think that it is an outcome of the market system in a capitalist economy, others criticize the inequitable use of resources by public institutions subsided by taxpayers’ money.
The purpose of the study was to identify the financial status of the NCAA Division-I FBS football programs. Moreover, the study specifically compared the financial characteristics of the football programs between the Power Five conferences - the Atlantic Coast Conference, Big Ten Conference, Big 12 Conference, Pac-12 Conference, Southeastern Conference - and non-Power Five conferences. Another purpose of the study was to investigate the relationship between football coaching salaries and key revenue sources. Among various financial issues in college sports, college football coaching salaries have been a controversial issue lately. Thanks to extraordinarily commercialized college football events, football coaching salaries have been escalating especially in the FBS football programs. It is rare, however, to address the relationship between revenue sources and coaching salaries. The findings of the study will help a variety of stakeholders in the intercollegiate athletic programs better understand the financial conditions and gain insights on a variety of financial issues in the highly commercialized college football industry.
2. Analysis Method
Panel data analysis refers to a research method that combines times series and cross-sectional data to examine changes over time and entity. Panel data is a dataset in which the behavior of entities is observed across time. Panel data analysis not only enables the researcher to study the dynamics of change with time series but also enhances the quality and quantity of data. Fixed effects models and random effects models are widely used analytical models in the context of panel data. The rationale of random effects models is that the variation across entities is assumed to be random and uncorrelated with the independent variables included in the model. Random effects models allow the researcher to include time-invariant variables (i.e., race, gender) and generalize the inferences beyond the sample in the models. Fixed effects models assume that unobserved heterogenous variables are constant over time. Thus, in fixed effects models, the time-invariant variables are absorbed by the intercept. The difference between fixed effects and random effects is that “whether the unobserved individual effect embodies elements that are correlated with the regressors in the model” (Greene, 2008: p. 183).
To examine how a variety of revenues sources in college football programs affect football coaching salaries, the following model was used for panel data analysis. In this model, independent variables are revenues from media and conference distribution (MED), donation (DON), government and institutional support (GIS), sponsorship (SPO), ticket sales (TKT), including Power Five Conference (PFC) as a dummy variable.
3. Model
Total Football Coaches Salaries (TFCS) = Media (MED) +Donation (DON) + Total Government and Institution Support (GIS) +Sponsorship (SPO) + Ticket (TKT) + Power Five Conference (PFC)
Where,
: Football coaching salaries
: Constant
MED it: revenue from media and conference distribution
DON it: revenue from donation
GIS it: revenue from government and institutional support
SPO it: revenue from sponsorship
TKT it: revenue from ticket sales
PFC it: Power Five conference
: between entity error
: error
The study examined the financial characteristics of the NCAA Division-I FBS football programs and compares the schools in the Power Five conferences with the schools in the non-Power Five conferences. Panel data analysis including fixed and random effects model employed. The dataset used in the study was the NCAA College Football Program Financial Data from 2005 to 2021. The dataset comprising 1,904 observations and 25 variables was panel data containing financial information such as revenues and expenses across the NCAA Division-I FBS colleges. The Power Five conferences accounted for 45% of total observations and the non-Power Five conferences accounted for 55%. The dataset was unbalanced panel data. Systematic adjustments to conference and division alignments resulted in the inherent presence of missing values within the dataset. Thus, the study used missing values treatment functions in Stata to address the issue.
To compare financial characteristics between the two groups in the FBS, the results of descriptive statistics are presented. Table 1 reported that there are considerable differences between the two groups in revenue and expense associated with the football programs.
Table 1. Mean of total revenue and expense.
Type |
Conference |
Mean ($) |
Std. Err. |
Total revenue |
Non-Power Five Power Five |
3.09e + 07 9.34e + 07 |
448484.8 1325467 |
Other revenue |
Non-Power Five Power Five |
1,507,766 7,911,529 |
50475.94 192966.2 |
Total expense |
Non-Power Five Power Five |
3.06e + 07 8.96e + 07 |
439450.1 1172415 |
Other expense |
Non-Power Five Power Five |
3,506,562 1.13e + 07 |
82546.34 198235.6 |
Table 2. Means of key revenue source.
Revenue Source |
Conference |
Mean |
Std. Err. |
Media/conference |
Non-Power Five Power Five |
3,129,954 2.63e + 07 |
78774.14 524254.8 |
Sponsorship |
Non-Power Five Power Five |
1,694,970 7,452,919 |
63869.45 216694.2 |
Donation |
Non-Power Five Power Five |
3,617,588 2.17e + 07 |
111343.7 516238.2 |
Ticket sales |
Non-Power Five Power Five |
2,950,753 2.06e + 07 |
107474 418845.6 |
Government / Institution Support |
Non-Power Five Power Five |
9,836,447 2,896,104 |
245501.6 168306.6 |
Student fee |
Non-Power Five Power Five |
6,896,162 2,308,215 |
187265 113192.7 |
Table 2 presented the mean of key revenue sources of the football programs. As expected, revenue from media contracts showed the biggest source of the revenue for the Power Five conferences while government/institution support is the biggest revenue source for the non-Power Five conference schools. The average student fees of non-Power Five conference schools are 33.4% more than those of Power Five conference schools. The average government and institutional support of non-Power Five conference schools is 29.3% more than that of Power Five conference schools. Consequently, it is reasonable to assert that non-Power Five conference football programs are heavily reliant on governmental and institutional support, including student fees. The programs also show lower revenue from ticket sales, donation, and sponsorship. Thus, it is readily discernible that the majority of Division-I FBS football programs would cease to operate without governmental and institutional support. Furthermore, the current financial structure employed is unsustainable and inequitable in allocating resources.
4. Results
This study performed the Hausman test and the Breusch and Pagan Lagrangian Multiplier test (LM test) to find the appropriate model fitted the data among the pooled ordinary least square (OLS), fixed effect model (FE), and random effect model (RE). After conducting the Hausman test, the random effect model was employed, since p is greater than 0.05 (probability > chi-square = 0.2346). The LM test confirmed the presence of random effect, meaning evidence of significant differences across entities (probability > chi-bar square = 0.000, Chi-bar square (01) = 1601.22). In addition, the OLS model was refused since p is less than 0.05. According to the modified Wald test, heteroscedasticity exists (probability > chi-square = 0.000). Consequently, the random effects model was selected.
5. Random Effects Model
Random-effects GLS regression |
Number of obs: 1,771 |
Number of groups = 109 |
R-Squared |
Within |
0.7527 |
Wald chi2(6) = 5844.24 |
|
|
Between |
0.8661 |
corr(u_i, X) = 0 (assumed) |
Prob > chi2 = 0.0000 |
|
Overall |
0.8171 |
sigma u = 784840.25 sigma e = 1143648.1. rho = 0.3201687 |
Table 3. Results of random effect regression analysis.
Variable |
Coefficient |
Std. Err. |
z |
P > z |
MED |
0.1460248 |
0.0034201 |
42.70 |
0.000 |
DON |
0.0239706 |
0.0029367 |
8.16 |
0.000 |
GIS |
0.0751549 |
0.0051926 |
14.47 |
0.000 |
SPO |
0.1610699 |
0.0115245 |
13.98 |
0.000 |
TKT |
0.0086317 |
0.0055423 |
1.56 |
0.119 |
PFC |
220401.5 |
17,590 |
1.25 |
0.210 |
Constant |
227012.1 |
136767.5 |
1.66 |
0.097 |
Note: dependent variable: football coaching salaries.
As shown in Table 3, football coaching salaries have positive relationships with media revenue, donation, government and institutional support, and sponsorship. However, ticket revenue and conference have no statistically significant relationships with football coaching salaries. The finding supports that, regardless of the conference participated, the FBS football programs show that football coaching salaries have been increasing during the period (2005-2021). Ticket revenue is also not a significant factor influencing football coaching salaries. The result implies that football coaching salaries are determined by the college football market generally influenced by supply and demand, not by individual football program’s financial condition.
It is worth noting that government and institutional support contributes to football coaching salaries. The size of the contribution is greater than donation and ticket sales. Most of the non-Power Five conference football programs receive substantial support from their own colleges, including student fees. Therefore, it is prevalent to transfer the money collected from the general student body to athletics programs across the football programs.
6. Fixed Effects Model
This study also employed an analysis model with the logarithm of each variable to compare the relative influence of independent variables on the dependent variable. After conducting the Hausman test, the fixed effect regression model was recommended, since p is less than 0.05 (probability > chi-square = 0.003). According to the modified Wald test, heteroscedasticity exists (probability > chi-square = 0.000), since p is less than 0.05. Consequently, the robust fixed effect regression model was employed to examine whether time-variant variables of the football programs affect the coaching salaries.
Fixed-effects (within) regression |
Number of observations = 1629 |
Number of groups = 107 |
R-squared |
Within 0.7546 |
F(6, 106) = 288.53 |
|
|
Between 0.8827 |
Corr (u_i, Xb) = −0.4685 |
Prob>F = 0.0000 |
|
Overall 0.8429 |
sigma_u: 0.23432703; sigma_e: 0.20702889; rho: 0.5616149.
Table 4. Results of fixed effect regression analysis.
Variable |
Coefficient |
S. E. |
t |
p > t |
MED |
0.4929888 |
0.0236697 |
20.83 |
0.000 |
DON |
0.1010132 |
0.0254641 |
3.97 |
0.000 |
SPO |
0.1354944 |
0.0189339 |
7.16 |
0.000 |
TKT |
−0.0396694 |
0.0083351 |
−4.76 |
0.000 |
GIS |
0.0854784 |
0.0214149 |
3.99 |
0.000 |
PFC |
−0.0785351 |
0.0445643 |
−1.76 |
0.081 |
Constant |
2.996077 |
0.460838 |
6.50 |
0.000 |
Note: dependent variable: football coaching salaries.
As shown in Table 4, football coaching salaries have positive relationships with media revenue, donation, government and institutional support, and sponsorship. However, unlike the random effect model, ticket revenue shows a negative association with football coaching salaries. This finding shows that ticket revenue is not a major determinant of football coaching salaries, implying that the football programs heavily relying on ticket revenue may have relatively low coaching salaries. Additionally, there was a difference between Power Five and non-Power Five conference schools.
The fixed effects analysis model used the logarithm of each variable, therefore, it is useful to compare the relative influence of independent variables on the dependent variable. NCAA/conference distributions and media rights (MED = 0.493), contributes the most to the football coaching salaries followed by corporate sponsorship (SPO = 0.135), donor contributions (DON = 0.101), and government/institutional support including student fees (GIS = 0.085).
The results of the fixed effect regression analysis examining the relationship between total revenue and key revenue sources confirmed the findings of the random effect regression analysis except for ticket revenue. NCAA/conference distributions and media rights (MED = 0.314), contributes the most to the total revenues followed by donor contributions (DON = 0.154), government/institutional support including student fees (GIS = 0.122), corporate sponsorship (SPO = 0.105), and ticket sales (TKT = 0.022).
It is worth noting that government and institutional support contributes to football coaching salaries. The size of the contribution is higher than donation and ticket sales. Most of the non-Power Five conference football programs receive substantial support from the colleges including student fees. Therefore, it is prevalent to transfer the money collected from the general student body to athletics programs across the football programs. Humphreys (2006) found that a college’s athletic success may positively affect state appropriations. Denhart and Vedder (2010) analyzed the effect of institutional characteristics on allocated revenue by concentrating on the direct and indirect costs to students. The authors found that allocated revenue is higher at schools with fewer students, schools with lower overall expenditure, and schools with higher percentages of low-income students. Jones et al. (2018) focused on the growth of student fees for athletics at Division I universities from 2004 to 2016, finding that the growth rate had been positive, had decreased over time, and varied over institutional affiliation within Division I. The authors maintained that given the substitutable nature of student fees for athletics and university funding for athletics, it was likely that universities were allocating more tuition and general funds to athletics as opposed to including specific athletics fees. Bradbury and Pitts (2018) pointed out that cost-of-attendance scholarships increases apply to all students making particularly nonathletes take on more financial burden including student loan debt.
Student fees have soared in the last decade, rising even faster than the overall cost of a public university education. In the fiscal year 2018, Division 1 schools garnered a substantial revenue of $1.2 billion, representing a significant increase of 51% compared to a decade earlier. Conversely, the average annual tuition fees at a four-year public college have surged by 37% during the same period. Additionally, four in five of the 230 Division 1 public universities charged students a fee to finance athletic programs (Enright, Lehren, & Longoria, 2020). However, there was no statistically significant relationship between conferences and football coaching salaries. This finding supports the fact that regardless of the conferences that schools belong to, the FBS football programs showed that football coaching salaries have increased between 2005 and 2021. Furthermore, it is worth noting that ticket revenue showed a negative relationship with coaching salaries. This outcome implies that football coaching salaries are likely determined by the college football market influenced by the supply and demand, not by an individual football program’s financial conditions, including revenue from ticket sales. In addition, as previously mentioned, revenue from ticket sales is a small portion compared with other key revenue sources to total revenue, the variable is not even a significant factor (see Table 3).
Numerous non-Power Five institutions, including those in the FBS, have been experiencing difficulties in selling their tickets. In response to this challenge, some colleges have utilized student fees to purchase their own tickets, thereby meeting the NCAA’s requirement for a certain number of tickets to be sold (Enright et al., 2020). This factor may contribute to the negative correlation between ticket revenue and football coaching salaries.
7. Conclusion
The study identified the financial status of the NCAA Division-I FBS football and compared the financial characteristics of the football programs between the Power Five conferences and non-Power Five conferences. Revenue from media contracts showed the biggest source of revenue for the Power Five conferences while government/institution support is the biggest revenue source for the non-Power Five conference schools. This financial characteristic leads to a significant disparity in financial resources in college football programs. Thus, it should be addressed critically for the sustainable finance of the college sports.
The study also examined the relationship between football coaching salaries and key revenue sources. Football coaching salaries have positive relationships with media revenue, donation, government and institutional support, and sponsorship. However, ticket revenue shows a negative association with football coaching salaries in the fixed effect model. This finding shows that ticket revenue is not a major determinant of football coaching salaries.
The highly commercialized college football industry in the United States has generated numerous financial challenges that necessitate stakeholders to thoroughly assess the current financial framework and identify appropriate solutions. Furthermore, there exists a huge financial disparity between the two groups despite they are even in the same subdivision.
Controversial college athlete compensation, escalating coaching salaries in football and men’s basketball, and substantial investment in sports facilities are significant financial challenges facing college athletics. Furthermore, there exists a significant financial disparity between college football programs, despite their shared subdivision. These issues are intricately linked to equity in the distribution of financial resources across colleges, competitive balance, and equity among academic and athletic programs (Bradbury & Pitts, 2018).
It is important to acknowledge the limitations of the study. Firstly, while this study examined the assumptions underlying the panel data analysis models employed, there is a possibility of endogeneity between coaching salaries and revenues. Secondly, this study did not include some key variables in the analysis, such as unobserved coaching quality, conference culture, and stakeholders’ characteristics due to the absence of relevant data. Therefore, an analysis of incorporating these variables is essential for future research endeavors.
Appendix
Description of Variables
Total Revenues |
Total revenues for the athletics program minus “Less Transfers to the Institution” |
Other Revenues |
Revenue from the following categories: Compensation and
benefits provided by a third party; game program, novelty,
parking and concession sales; sports camps and clinics; athletics restricted endowment and investments income; and, other
operating revenue. |
NCAA/Conference Distributions,
Media Rights,
Post-Season
Football (MED) |
Revenue received from the NCAA (including championships) and athletics conferences, media rights, and post-season football bowl games. |
Ticket Sales (TKT) |
Revenue received from ticket sales for all NCA-sponsored sports at an institution. |
Government/
Institutional
Support (GIS) |
Revenue received from governments, direct funds from the
institution for athletics operations, and costs covered and services provided by the institution to athletics (and for athletics debt but not charged to athletics. |
Student Fees |
Fees paid by student and allocated for the restricted use of the
athletics department |
Corporate
Sponsorship,
Advertising,
Licensing (SPO) |
Revenue generated by the institution from royalties, licensing.
advertisements and sponsorships. |
Donor
Contributions (DON) |
Funds contributed from individuals, corporations, associations, foundations, clubs or other organizations external to the athletics program above the face value for tickets. |
Total Expenses |
Total athletic operating expenses, including athletic scholarship costs. |
Other Expenses |
Expenses related to the following categories: Sports equipment, uniforms and supplies, fundraising, marketing and promotion, sports camps, spirit groups, direct overhead and administrative
expenses, indirect institutional support, membership and dues, student-athlete meals, and, other operating expenses. |
Total Academic Spending |
Total expenditures for the direct role and mission activities of an institution. It includes functional classifications of expenditures for instruction, research, public service, academic support, student services, institutional support, scholarships and fellowships, and operations and maintenance. |
Total Football Coaching Salaries |
Total compensation reported for all football coaches, including salaries, benefits and bonuses paid by the university, and
contractually-guaranteed amounts paid by third parties. |
Power Five
Conference (PFC) |
ACC, BIG 10, BIG12, SEC, PAC-12 |