Introduction

According to the NCAA (Fulks, 2001), contributions from alumni and others, or fund raising, is the second-largest revenue source for Division I-A athletic programs, trailing only ticket sales. Fund raising accounts for nearly five million dollars of the typical Division I-A athletic programs’ $25 million of total revenue, and, as such, is clearly a vital source of funding for intercollegiate athletic programs. Therefore, the ability to forecast fund raising revenues is crucial for college athletic departments. This study will create a model to predict annual fund raising revenues in NCAA Division I-A intercollegiate athletics, thus aiding practitioners in predicting these revenues on their own respective campuses.

Related Literature

Numerous authors have examined the relationships between intercollegiate athletic programs and higher education. These studies have focused on the relationship between college sports performance and variables such as applicants to universities (Allen & Peters, 1982; Chressanthis & Grimes, 1993; Murphy & Trandel, 1994; Toma & Cross, 1998; Zimbalist, 2001), SAT scores of incoming students (Bremmer & Kesserling, 1993; McCormick & Tinsley, 1987; Mixon, 1995; Tucker & Amato, 1993), and university fund raising (Baade & Sundberg, 1996; Brooker & Klastorin, 1982; Budig, 1976; Gaski & Etzel, 1984; Grimes & Chressanthis, 1994; McCormick & Tinsley, 1990; Sack & Watkins, 1985; Sigelman & Carter, 1979). Few studies, however, have investigated athletic fund raising in this regard.

Sigelman and Brookheimer (1983) examined the relationship between 11 predictor independent variables and contributions to both athletics and university fund raising programs at 57 NCAA Division I-A institutions in major athletic conferences using a multiple regression analysis. Football success (r = .335) and traditionalism (r = .242), a scaled measurement of the social and political culture towards civic responsibility and philanthropy, were determined to be significant predictors of giving to intercollegiate athletics annual fund raising programs, albeit not overly strong predictors given their Pearson’s coefficient values.

Coughlin and Erekson (1984, 1985) utilized multiple linear regression analysis to model contributions to athletic fund raising program using 16 independent variables. Coughlin and Erekson utilized 1980-1981 athletics fund raising data published in the Omaha World-Herald as their measurement of the contributions dependent variable, as Sigelman and Brookheimer (1983) had done previously. Coughlin and Erekson’s final regression model accounted for 58% of the variance in predicting athletic contributions. The authors identified football attendance, conference affiliation, bowl participation, state population, men’s basketball winning percentage, and professional competition to be significant determinants of athletic contributions.

Daughtrey and Stotlar (2000) investigated the relationship between contributions to both athletics and the university and winning a national championship in football at the Division I-AA, Division II, and Division III levels. Daughtrey and Stotlar found significant relationships between football championships and increased contributions to athletics with Division II and Division III schools and between football championships and increased contributions to the university with Division III institutions. The authors’ delimitations of studying only national champions and not examining Division I-A institutions prevent useful comparisons between their work and the study conducted here.

No other published studies were identified that attempted to use a variety of variables to predict fund raising contributions to NCAA Division I-A intercollegiate athletics programs. As such, the only published works investigating the ability to predict athletic fund raising contributions are currently 20 years old and each relies upon contributions data collected in 1980-1981. Obviously, much has changed in intercollegiate athletics since then. If an athletic fund raising practitioner today tried to understand and predict fund raising contributions based upon the existing body of literature, they would be relying upon considerably outdated research. Clearly then, there is a need to re-examine the prediction of athletic fund raising contributions, as is the purpose of this study.

Methods

Subjects

The population for this study was defined as all 119 NCAA Division I-A athletic programs and their athletic fund raising contributions for each of the five-year span from 1998-1999 to 2002-2003. Questionnaires were sent to the athletic fund raising director at each of the 119 institutions in performing a census of the population. Thirty-five questionnaires were returned, representing 171 usable subjects, for a usable response rate of 28.7%.

Variables

Based on the work of Coughlin and Erekson (1984, 1985) and Sigelman and Brookheimer (1983), 13 predictor variables were selected to use in explaining the variation in annual athletic fund raising contributions: football and men’s basketball winning percentages for the year examined, the change in football and men’s basketball winning percentages from the previous year, average home attendance for football and men’s basketball in the year examined, whether the school is a member of a “major” athletic conference, whether the school is a public or private institution, state population, and four categorical variables to control for fixed-effects in the time-series regression analysis. Each of these variables is described further in Table 1.

Procedures

Questionnaires were sent to athletic fund raising directors at all 119 NCAA Division I-A athletic programs to collect dependent variable data. Data collection on each of the predictor variables was performed as discussed in Table 1. A fixed-effects ordinary least squares (OLS) multiple regression equation was developed to empirically explain annual athletic fund raising contributions. The fixed-effects model is used to control for changes over time due to the use of panel data. Four indicator variables were used to represent the five years of data from 1998-1999 to 2002-2003. A significance level of .01 was established a priori to reduce the risk of Type 1 error common with time-series regression analyses and large sample sizes.

Results

Table 2 provides descriptive data for the continuous variables included in the regression equation. The results show that the average annual athletic fund raising contributions total was $4,065,616. Additionally, the average home football game attendance was 43,119 and the average men’s basketball game attendance was 8,749.

Five of the 13 independent variables were found to be significantly related to athletic fund raising contributions at the .01 level, including football home attendance (r=.721), conference affiliation (r=-.621), football winning percentage (r=.322), type of institution (r=-.302), and men’s basketball home attendance (r=.237). In examining the correlation coefficients between the independent variables, only the relationship between football attendance and conference affiliation was above .500 or below -.500 (r=-.651), thus providing evidence that multicollinearity was not problematic.

Table 3 summarizes the multiple regression results. The model was a statistically significant estimator of annual athletic fund raising contributions. The model F-statistic equaled 18.647 and was significant at the .01 level. In addition, the model explained 60.7% of the variation in spectator attendance and the adjusted R2 was .574. The R2 and adjusted R2 findings were similar to those found in Coughlin and Erekson (1984, 1985). Additionally, this type of regression analysis allows for an estimation of the magnitude of change in annual contributions based upon a change in values of the independent variables. For example, the results suggest that membership in one of the six conferences with automatic bids to the Bowl Championship Series in football is worth more than $2.5 million per year in athletic fund raising contributions to conference members. Also, the data suggests that annual athletic fund raising contributions would increase by $70 for each average attendee increase at home football games.

Discussion

The purpose of this study was to predict annual athletic fund raising contributions in NCAA Division I-A intercollegiate athletics, providing a needed re-examination of this issue given the dated works in this area of the literature. Despite the passing of two decades and major changes in intercollegiate athletics since the studies of Sigelman and Brookheimer (1983) and Coughlin and Erekson (1984, 1985), this study supports the findings of those previous works, particularly Coughlin and Erekson. As with their work, this investigation found both home football attendance and conference affiliation to be statistically significant predictors of annual athletic fund raising contributions. Additionally, the amount of variance explained in annual athletic fund raising contributions in this study (R2=.607) was extremely close to that of Coughlin and Erekson (R2=.58). None of the similarities between the findings of these studies are, in and of themselves, overly surprising; however, these similarities are somewhat surprising given the radical changes in intercollegiate athletics since the early 1980’s. These changes include a dramatic increase in media/television coverage, rapid increases in revenues and expenses among athletic programs, the creation of the Bowl Championship Series, conference realignment, and progress towards gender equity. It is in the context of all of these major changes that the similarities between this study and previous dated works are noteworthy.

These results indicate that, assuming conference affiliation does not change, an athletic fund raising practitioner should carefully track home football attendance as an indicator of fund raising contributions. A fairly strong positive relationship (r=.721) was found between these two variables. Other changeable variables of interest to practitioners in this regard are football winning percentage (r=.322) and men’s basketball home attendance (r=.237), however, neither approaches home football attendance in the ability to predict athletic fund raising contributions.

References

  1. Allen, B. H., & Peters, J.I. (1982). The influence of a winning basketball program upon undergraduate student enrollment decisions at DePaul University. In M.J. Etzel & J.F. Gaski (Eds.), Applying marketing technology to spectator sports (pp. 136-148). Notre Dame, IN: University of Notre Dame.
  2. Baade, R. A., & Sundberg, J. O. (1996). Fourth down and gold to go? Assessing the link between athletics and alumni giving. Social Science Quarterly, 77, 789-803.
  3. Bremmer, D. S., & Kesserling, R. G. (1993). Advertising effects of university athletic success. Quarterly Review of Economics and Business, 33, 409-421.
  4. Brooker, G., & Klastorin, T. D. (1981). To the victors belong the spoils? College athletics and alumni giving. Social Science Quarterly, 62, 744-750.
  5. Budig, J. E. (1976). The relationships among intercollegiate athletics, enrollment, and voluntary support for public higher education (Doctoral dissertation, Illinois State University, 1976). Dissertation Abstracts International, 37, 4006A.
  6. Chressanthis, G. A., & Grimes, P. W. (1993). Intercollegiate sports success and first-year student enrollment demand. Sociology of Sport Journal, 10, 286-300.
  7. College Football Data Warehouse (2004). Retrieved from cfbdatawarehouse.com.
  8. Coughlin, C. C., & Erekson, H.O. (1984). An examination of contributions to support intercollegiate athletics. Southern Economic Journal, 66, 180-195.
  9. Coughlin, C. C., & Erekson, H.O. (1985). Contributions to intercollegiate athletic programs: Further evidence. Social Science Quarterly, 66, 194-202.
  10. Fulks, D. L. (2002). Revenues and expenses of Division I and II intercollegiate athletics programs: Financial trends and relationships-2001. Indianapolis: National Collegiate Athletic Association. Retrieved from http://www.ncaa.org.
  11. Gaski, J. F., & Etzel, M. J. (1984). Collegiate athletic success and alumni generosity: Dispelling the myth. Social Behavior and Personality, 12(1), 29-38.
  12. Grimes, P. W., & Chressanthis, G. A. (1994). Alumni contributions to academics: The role of intercollegiate sports and NCAA sanctions. American Journal of Economics and Sociology, 53, 27-40.
  13. McCormick, R. E., & Tinsley, M. (1987). Athletics versus academics: Evidence from SAT scores. Journal of Political Economy, 95, 1103-1116.
  14. McCormick, R. E., & Tinsley, M. (1990). Athletics and academics: A model of university contributions. In B. L. Goff & R. D. Tollison (Eds.), Sportometrics (pp. 193-204). College Station, TX: Texas A&M University Press.
  15. Mixon, F. G. (1995). Athletics v. academics: Rejoining evidence from SAT scores. Education Economics, 3, 277-283.
  16. Murphy, R. G., & Trandel, G. A. (1994). The relation between a university’s football record and the size of its applicant pool. Economics of Education Review, 13, 265-270.
  17. National Collegiate Athletic Association. (2004). Attendance data. Retrieved from http://www.ncaa.org.
  18. Sack, A. L., & Watkins, C. (1985). Winning and giving. In D. Chu, J. O. Segrave, & B. J. Becker (Eds.), Sport and Higher Education (pp. 299-306). Champaign, IL: Human Kinetics.
  19. Sigelman, L., & Bookheimer, S. (1985). Is it whether you win or lose? Monetary contributions to big-time college athletic programs. Social Science Quarterly, 64, 347-359.
  20. Sigelman, L., & Carter, R., (September, 1979). Win one for the giver? Alumni giving and big-time college sports. Social Science Quarterly, 60, 284-293.
  21. Toma, J. D., & Cross, M. E. (1998). Intercollegiate athletics and student college choice: Exploring the impact of championship seasons on undergraduate applications. Research in Higher Education, 39, 633-661.
  22. Tucker, I. B., & Amato, L. (1993). Does big-time success in football or basketball affect SAT scores? Economics of Education Review, 12, 177-181.
  23. Zimbalist, A. (2001). Unpaid professionals: Commercialism and conflict in big-time sport. Princeton, NJ: Princeton University Press.

Table 1

<th”>Variable <th”>Description and Sources

Variable Descriptions and Data Sources
CONTRIB Dependent variable representing one year’s annual athletic fund raising contributions, not including capital campaigns. Collected via questionnaire self-reporting from each institution’s athletic fund raising director.
FBWINPCT Represents a school’s football winning percentage for that respective year. Collected from ncaa.org.
BBWINPCT Represents a school’s men’s basketball winning percentage for that respective year. Collected from ncaa.org.
FBWINCH Represents the change in a school’s football winning percentage from the previous year. Collected from ncaa.org.
BBWINCH Represents the change in a school’s football winning percentage from the previous year. Collected from ncaa.org.
FBATTEN Represents a school’s average home football attendance for that respective year. Collected from ncaa.org.
BBATTEN Represents a school’s average home men’s basketball attendance for that respective year. Collected from ncaa.org.
CONFERNC Represents whether or not a school is a member of one of the six major Division I-A conferences that receives an automatic bid to the Bowl Championship Series in football. Coded as a “0” for BCS conference, “1” for non-BCS conference.
INSTTYPE Represents whether the school is a public or private institution. Coded as a “0” for public institution, “1” for private institution.

Table 2

<th”>Variable <th”>Mean <th”>Standard Deviation <th”>Minimum <th”>Maximum

Descriptive Statistics
CONTRIB $4,065,615.92 $3,502,701.71 $201,791 $14,363,913
FBWINPCT .513 .212 .000 1.000
BBWINPCT .575 .155 .222 .892
FBWINCH -.007 .206 -.727 .492
BBWINCH -.005 .151 -.419 .398
FBATTEN 43,118.94 25,737.64 6,595 111,175
BBATTEN 8,748.54 4,975.17 935 22,248
CONFERNC .40 .491 .00 1.00
INSTTYPE .16 .371 .00 1.00
POPULATE 8,596,946.04 8,142,659.46 1,808,344 33,871,648

Note: N=171

Table 3

<th”>Variable <th”>Unstandardized Beta Coefficient <th”>Standard Error <th”>T-statistic <th”>P-value

Regression Results
FBWINPCT 934.813 1,259.942 .742 .459
BBWINPCT 1,474.402 1,631.419 .904 .368
FBWINCH -658.047 1,064.320 -.618 .537
BBWINCH 434.665 1,386.728 .313 .754
FBATTEN 70.567 12.129 5.818 .000
BBATTEN -137.200 53.892 -2.546 .012
CONFERNC -2,587,336.224 529,184.751 -4.889 .000
INSTTYPE -636,637.927 530,003.528 -1.201 .231
POPULATE -.0253 .023 -1.105 .271
TIME0203 1,417,825.114 567,331.928 2.499 .013
TIME0102 1,207,864.865 568,200.487 2.126 .035
TIME0001 916,870.951 560,734.574 1.635 .104
TIME9900 392,457.586 565,992.692 .693 .489
Constant 1,443,097.013 1,260,510.634 1.145 .254

Note: R2=.607, Adjusted R2=.574, F-statistic=18.647, P-value=.000

Print Friendly, PDF & Email