The effect of sanctions on external job promotions of NCAA division I-FBS College Football Head Coaches.

Soebbing, Brian P. ; Tutka, Patrick ; Seifried, Chad S. 等

Introduction

Sport has been found to be a good empirical setting to examine organizational (Day, Gordon, & Fink, 2012) and labor market (Kahn, 2000) phenomena. One of the most common topics using sport as an empirical setting examines succession of head coaches. Specifically, such research examines the determinants (e.g., d'Addona & Kind, 2014; Frick, Barros, Prinz, 2010; Holmes, 2011) and consequences of head coach succession (e.g., Allen, & Chadwick, 2012; Soebbing & Washington, 2011). These topics formed an active area of sports research over the past five decades since Grusky's (1963) research sought to examine team performance following a change in Major League Baseball (MLB) managers.

Within the organizational succession literature, examining sport job promotions received limited attention. This lack of attention in the literature could be due to the difficulty in defining and tracking movement of individuals throughout their career and defining what a job promotion would be for an individual that is already in a head coaching position. Generally, research examining job promotion focuses on issues such as salary, human and social capital, discrimination between men and women, career progression, and availability of positions in a greater context of job mobility (e.g., Adler, Berry & Doherty, 2012; Baron, Davis-Blake, & Biebly, 1986; Gibbons, & Waldman, 1999, Rosenfeld, 1992).

The purpose of the present research is to explore the role sanctions play in external job promotions of NCAA Division I-Football Bowl Subdivision (FBS) head college football coaches. We choose NCAA Division I-FBS football due to its financially lucrative nature (Brown, Farrell, & Zom, 2007; Otto, 2005). Division I-FBS teams generate large revenues for their athletic departments and may provide additional benefits to the university in terms of increased donations (e.g., Humphreys & Mondello, 2007; Rhoads & Gerking, 2000), state appropriations (e.g., Alexander & Kern, 2010; Humphreys, 2006), and student admission rates (e.g., Toma & Cross, 1998). In addition, many college football coaches are the highest paid employees of their university and the state (Greenberg & Smith, 2007), and are presented with opportunities for them to move to other universities to serve a similar role of head football coach or to become a National Football League (NFL) head coach.

We examine head football coach movement from the 1973-1974 season through the 2012-2013 season, operationalizing external job promotions based upon university status, bowl revenue generation, and the coach receiving a position in the NFL. Estimating a logistic regression model, we find sanctions do not impact the likelihood of external job promotion. However, other factors such as on-field performance, university characteristics, and the availability of coaching positions influence external job promotion of current head coaches.

Literature Review

Within the academic literature, job promotion has been defined multiple ways (Longnecker & Fink, 2008). For instance, Winter-Ebmer and Zwelmuller (1997) defined job promotion as a movement to a job seen by peers and executives as higher in terms of job quality compared to the individual's previous job. Within, there are two types of job promotions: internal and external. Internal job promotion involves movement upwards within the same organization. Theoretically, internal job promotion has been positioned as incentive based, where a tournament is created and superior performance is rewarded, or as an optimal assignment, where higher talented individuals are assigned to the higher positions within the organization's structure mechanisms. External job promotion involves movement from one company to another (Fee, Hadlock, & Pierce, 2006). Even though little theoretical work has focused on external job promotions, Fee et al. (2006) stated similar factors determining internal promotions should also determine external promotions. Generally, job promotions involve significant increases to an individual's wage. This element of job promotions has received attention in the literature (DeVaro & Brookshire, 2007; Gibbons & Waldman, 1999).

Many other broad factors also influence job promotion. These factors include human capital factors such as education (e.g., Smith & Abbott, 1983), social capital (e.g., Day & McDonald, 2010), demographic factors such as age and physical attractiveness (e.g., Morrow, McElroy, Stamper, & Wilson, 1990), market factors such as vacancies, and the structure of the organization (e.g., Rosenfeld, 1992). Additional research examined specific aspects of job promotion. For example, Longnecker and Fink (2008) surveyed managers as to specific reasons why they promoted employees. The main reasons included the ability to get desired results, possessing strong networks and connections, excellent communication skills, and strong career-based knowledge (Longnecker & Fink, 2008). Fee et al. (2006) also mentioned the value of talent and ability as being important to job promotion when they examined both internal and external job promotions of coaches using data from the NFL. In their study examining internal and external job promotions of head coaches and offensive and defensive coordinators from 1970 until 2001, they found differences in the determinants of internal and external promotions for assistant coaches to head coaches. Factors affecting external promotions, such as team performance, did not necessarily impact internal promotion.

One omitted factor from previous research is the role that misconduct by the coach or the organization imposes on a coach's likelihood of receiving an external promotion. To examine misconduct, we look at actions investigated and punished in NCAA Division I-FBS, which is the highest classification for universities in terms of athletics. The main objective of the NCAA is "[...] to maintain intercollegiate athletics as an integral part of the educational program and the athlete as an integral part of the student body and, by so doing, retain a clear line of demarcation between intercollegiate athletics and professional sports" (National Collegiate Athletic Association [NCAA], 2014, p. 1).

The NCAA enforcement process is a four-step process (NCAA, 2013). First, a possible violation is reported to the NCAA enforcement office by either the university that committed the possible violation or by another member university. Second, this possible violation is investigated by the enforcement office. Third, the NCAA sends a notice of allegations to the university if evidence is found to support the claim that a violation of NCAA rules has occurred. The notice of allegations includes what rules or bylaws the enforcement staff believes the school has broken, along with supporting evidence (NCAA, 2013). The university's decision-makers can decide to accept the charges stated within the notice of allegations or request a hearing to defend its organization to the Committee on Infractions (NCAA, 2013).

Fourth, the Committee on Infractions levies the penalties (commonly called "sanctions") it deems necessary after the hearing (NCAA, 2013). Prior to August 2013, violations were defined as major or secondary. The NCAA defined a secondary violation as "a violation that is isolated or inadvertent in nature, provides or is instead to provide only a minimal recruiting, competitive or other advantage and does not include any significant impermissible benefit [...]" (NCAA, 2012, p. 323). All other violations are classified as a major violation (NCAA, 2012).

After August 2013, the NCAA classifies violations by four levels. Level I and II sanctions include compensation penalties (e.g., "limitations on the institution's participation in postseason play"), financial penalties (e.g., fines, return of revenue received or the "reduction in or elimination of monetary distribution by the association"), scholarship reduction during a specific period, show-cause orders, head coach restrictions, recruiting restrictions, and probation (NCAA, 2014, p. 324-325). Level III and IV sanctions include public reprimand, suspension of head coach, institutional recertification, fines, forfeiture/vacation of contests, and the termination of recruitment regarding specific student-athletes (NCAA, 2014, p. 326). Upon receiving the sanctions, the university can appeal the penalties to an infractions appeal committee (NCAA, 2013).

Previous research on sanctions can be partitioned into two streams of research.1 The first stream is in the context of cartel behavior both from a legal perspective (e.g., Otto, 2005) and an economics perspective (e.g., Humphreys & Ruseski, 2009; Winfree & McCluskey, 2008). Winfree and McCluskey (2008) found universities who selfreported their violations to the NCAA received more lenient punishments compared to universities that did not self-report. Humphreys and Ruseski (2009) examined the likelihood of football teams cheating. Their results found better on-field performance (defined by a team's winning percentage) increased the likelihood of cheating while the more years' experience of the head coach and higher the university's funding of students decrease the likelihood of the football team cheating. The second area looks at the consequences of sanctions. Grimes and Chressanthis (1994) found sanctions slightly reduced the donations received by the academic unit of the university. Holmes (2011) found infractions increased the likelihood for dismissal of Division I-FBS head coaches. Inoue, Plehn-Dujowich, Kent, and Swanson (2012) concluded sanctions did not have any impact on head coaching salaries.

In summary, the literature on job promotion has determined that many factors influence the likelihood of job promotion. However, previous research has not looked at the impact that misconduct has on the likelihood on job promotion. The availability of infraction data in the NCAA provides an opportunity to look at the impact that sanctions have on job promotion. While we find that infractions do have impact in certain areas (coaching dismissals, contributions to the universities) but not in others (coaching compensation), the examination of the potential impact that misconduct has on a coach receiving an external promotion warrants attention.

Methods

In order to investigate the effect sanctions have on external promotions of NCAA Division I-FBS head coaches, the present research examines both head coaching movement and sanctions from the 1973-1974 season through the 2012-2013 season. The unit of observation is a coach-season. During the sample period, there are 4,680 coach-season observations.

The dependent variable in the present research is if a coach accepted an external job promotion (Promo) following the observed season. The variable takes the value of 1 if the coach accepted an external promotion after the observed season; otherwise, 0. The information regarding coaching movement within the NCAA was gathered from the College Football Data Warehouse.

We operationalized external promotions in three instances in order for promotion to be consistent across all the organizations during the sample period. The first instance was a coach moving to a non-Bowl Championship Series automatic qualifying (AQ) school (i.e., schools located in Conference USA, Mid-American Conference, Mountain West Conference, Sun Belt Conference, and Independents Army, Navy, and Brigham Young University) to a BCS AQ institution (i.e., members of the American Conference [former Big East Conference], Atlantic Coast Conference, Big 12 Conference, Big Ten Conference, Pacific 12 Conference, Southeastern Conference, and Notre Dame). In general, BCS AQ schools report significantly higher revenue compared to non-AQ institutions. For example, AQ conferences receive a reserved spot in the postseason bowl games as a result of their position as a charter member of the BCS and previous conference tie-in agreements with the various BCS bowl sites (i.e., Tostitos Fiesta Bowl, Discover Orange Bowl, Rose Bowl, and Allstate Sugar Bowl). Seifried and King (2012) found during their bowl history investigation (i.e., 1902 to 2010) that 1,077 of 1,090 bowl games involving Division I-FBS institutions produced a total payout of $3,570,360,540 with current BCS-charter conference schools collecting "roughly 90% of all bowl game money in 1,633 appearances whereas non-BCS institutions secured about 9% from 434 appearances" (p. 209).

The second instance was a coach moving from one university to another university within the same BCS AQ distinction (i.e., non-AQ school to non-AQ school or AQ school to AQ school). Under this instance, we examined the five-year average nominal bowl payout from the coach's current university and compared that number to the five-year average nominal bowl payout for the university that the coach was moving to for his new position. If the payout was higher to the university the coach was moving to, it was considered a promotion. The final instance was a coach leaving to become an NFL head coach, offensive coordinator, or defensive coordinator. The information on NFL coaching rosters was gathered from multiple websites.

There are two independent variables in the present research that explore the impact sanctions have on external job promotion. For the sample period, there are 451 university reports published by the NCAA for all Division I institutions involving all sports. Only examining violations involving football for Division I-FBS schools, the number of reports reduces to 144.

The public reports contain information regarding when the violations occurred, the summary of the violations, the sport(s) within the university that were included in the violation report, whether the NCAA determined these violation to be major, and the resulting penalties. From that report, the coach that was managing the team when the violation(s) were accused to have occurred according to the report was coded with a 1 in the year in which the letter of allegations was given. The first independent variable is Violations, which represents the cumulative number of violations the observed coach has committed during his career until the observed season. The public use reports state whether the NCAA governing body declared the violation to be major or secondary. Similar to the Violations variable, the variables Major and Secondary are the cumulative number of major and secondary violations the observed coach has committed during his career until the observed season. In order to account for the variation in experience among the observed coaches in the sample, these three counts are divided by the coach's career head coaching experience.

When examining executive promotions, a number of other variables can affect the likelihood of an external promotion. The first is the performance of the organization under the coach's lead. Research examining performance effects in college athletics has documented the positive effect that on-field athletic success has on an overall university such as an increase in state appropriations (Alexander & Kern, 2010; Humphreys, 2006), donations (Grimes & Chressanthis, 1994; Humphreys & Mondello, 2007), higher quality students (Chressanthis & Grimes, 1993), and coaching salaries (Grant, Leadley, & Zygmont, 2013). Thus, one would also expect that strong recent on-field performance by the coach's team increases the likelihood that a coach would earn an external promotion.

In the present research, on-field performance variables are included that reflect the belief that strong on-field performance leads to an external promotion. As a result, we include the coach's team winning percentage (Winpct) in both the current and previous seasons. (2) The long-term performance of the organization that the observed coach is currently leading may also affect whether the coach is externally promoted. Thus, the variable SchoolWinpct is the winning percentage over the previous four Division IFBS seasons. Four years is the amount of time an individual player is eligible in the NCAA. Hence, four seasons is a recruiting cycle for a team. The use of the organization's performance over the past four seasons is consistent with previous research (e.g., Soebbing & Washington, 2011). Fee et al. (2006) discussed the importance of controlling for team performance as higher sustained team performance will lead to the head coach staying in his current position with the team.

Coach characteristics may impact whether our independent variables affect job promotions. The present research includes four coaching factors. The first characteristic is the age of the coach (CoachAge) expressed as the difference between the observed year and the coach's year of birth. A coach's year of birth was obtained and confirmed from multiple secondary sources. Previous research on promotion of coaches used age as a control variable (Fee et al., 2006). The second characteristic is the number of years the observed coach has been a head NCAA football coach at any level (CarExp), which reflects coaching competence (e.g., Pfeffer & Davis-Blake, 1986; Soebbing & Washington, 2011). The third coaching characteristic is whether a coach is a visible minority (Minority), which has been shown to affect the career progression of coaches (Bozeman & Fay, 2013). The variable is dichotomous with a value of 1 indicating the coach is a visible minority and 0 if the coach is not a visible minority. In order to determine minority status, the authors examined head shots of each coach within the sample. This method is consistent with previous research by Holmes (2011). The final coaching characteristic variable is the reputation of the coach (Rep). We operationalize reputation by the cumulative number of bowl appearances by the coach at the end of the observed season. The number of bowl appearances by the coach is divided by his career coaching experience.

There are two important university characteristics to account for in the present research. The first university characteristic is the number of years the coach's university has been playing football in Division I-FBS (FootballAge). The age of the football program was calculated with data from College Football Warehouse. The second characteristic is if the coach is currently coaching at a private school (Private). Fulks' (2012) annual reports on the revenues and expenses of NCAA intercollegiate athletics show coaching salaries as distinct observations regarding those supported at public and private institutions. Furthermore, Fulks (2012) noted it was critical for individuals reading his report to understand "there are inherent differences in fiscal demands and resources of public institutions and those of private institutions" (p. 12). This variable, obtained from the National Center for Education Statistics through the U.S. Department of Education, is equal to 1 for a private institution and 0 for a public institution. The final variable captures the number of head coach position openings at both Division I-FBS and the NFL (Opening). It reflects the number of potential opportunities for a coach to change positions. Given that the total number of Division I-FBS and NFL teams varies during the sample, this variable is divided by the combined number of teams in both leagues.

Model

The formal model is expressed below:

[PROMO.sub.it] = f ([WinPct.sub.it], [WinPct.sub.i(t-1)], [SCHOOLWinPct.sub.it], [COACHAGE.sub.it], [MINORITY.sub.it], [CAREXP.sub.it], [REPi.sub.jk], [FOoTBALLAGE.sub.it], [PRIVATE.sub.it], [COURT.sub.it], [OPENINGS.sub.it], [SANCTIONS.sub.it], [[epsilon].sub.it]) (1)

where i indexes coaches, t indexes seasons, and e is the equation error term.

Estimation Issues

There are two estimation issues that may affect the reliability and validity of the results. The first is the major and secondary violations are highly correlated with the total number of sanctions as the as the correlation coefficients are between 0.83 and 0.89. The correlation coefficient between major and secondary violations is low at 0.42. As a result, we estimate two models, one model includes total sanctions and the other model includes both major and secondary violations. The second estimation issue is with the equation error term within Equation 1. As we do not know specific coaches' desires for changing positions, let alone a promotion, we cluster the standard errors by coach.

Results

Table 1 presents the summary statistics for the sample period. The final sample size for the present research is 3,913 coach-season observations as 767 observations were eliminated. These observations were eliminated for two reasons. The first reason was the observed coach may be in his first year in the observed season and did not have a previous win-loss record. The second reason variables were eliminated was because a school was not a member of Division I-FBS for the four previous seasons. Thus, these schools have missing values for the SCHOOLWinpct variable.

The average winning percentage for the sample period is 0.524. The average age of the coach during the sample period is 49 years old. Three percent of coach-season observations are visible minorities. Within the sample, coaches average 10 years of NCAA head coaching experience. The average football age for universities in the sample period is 73. On average, about 17% of the total number of NFL and Division IFBS teams has an open head coaching position any given year.

Due to the dependent variable being dichotomous, a limited, dependent estimator (probit or logit) should be used (Maddala, 1983). Table 2 presents the logistic regression results with robust standard errors and the odds ratios for the significant variables found in Equation 1. In Model 1, we operationalized sanctions by the cumulative number of separate violation incident reports (Violations). This variable is insignificant at the level of 0.05. In Model 2, we examine the severity of those incidents with the cumulative number of incidents that were labeled by the NCAA as major (Major) and secondary (Secondary) violations. The results from Model 2 show the cumulative number of major and secondary violations do not impact external promotions.

The sign and significance for all of the control variables included in Equation 1 are consistent in both models. Winning percentage in both the observed season (Winpct) and the prior season ([Winpct.sub.(t_ij]) increase the likelihood the observed coach receives an external job promotion. The performance of the school over the previous four seasons (SchoolWinpct) at which the observed coach is currently employed decreases the likelihood of receiving an external job promotion.

The age of the coach (CoachAge), his career experience (CarExp), and the age of the football program (FootballAge) has a negative and significant effect on external job promotions. A coach that is a visible minority (Minority), employed at a private school (Private), and the number of head coach openings in Division I-FBS and the NFL (Openings) increases the likelihood of an external job promotion. Finally, a star coach (Rep) does not have any significant effect regarding an external job promotion at the 0.05 level.

Two additional models were analyzed to access the robustness of the results reported in Table 2. We substituted the career coaching experience for the number of years being the head coach at the university in the observed season. This change did not result in any changes with regard to the sign and significance of our main independent variables as well as the experience variable. Since promotions occur 3% of the time, it may constitute a rare event. When events are rare, a rare events logistic model does not have to be used to addresses biases in predicted probabilities when sample sizes are large and events are few (King & Zeng, 2001). When estimating a rare event logistic regression model, the sign and significance of the parameter estimates do not change when comparing to the results in Table 2. Thus, we are confident that the results reported in Table 2 are robust.

Discussion

We find the cumulative number of sanctions--major and secondary violations--do not have any significant impact on the likelihood of a coach receiving an external promotion. This result differs from the previous research by Holmes (2011), who found sanctions lead to an increased chance of being dismissed. Thus, sanctions affect dismissals but not external promotions.

Within this empirical setting, these insignificant findings may provide additional insight regarding behavior by both the NCAA and its member institutions. As Humphreys and Ruseski (2009) stated, the NCAA is known for a slow response of the enforcement apparatus (The Committee on Infractions) in response to the misbehavior of a member. Thus, the time the punishment occurs compared to when the violation took place may be critical in regards to any potential effect for the coach. The NCAA has attempted to speed up the enforcement process in recent years, but no significant testing of its effectiveness has occurred to this point.

Another reason for an insignificant result could be that violations of NCAA rules are more widespread than what is known through the enforcement data provided by the NCAA. Given the incentives for schools to cheat and the difficulty in trying to enforce the rules when there are many member organizations within the association, it could be that rule violations are considered normal within that environment and, thus, would not have any impact on coaching promotions. An insignificant result might also mean the punishment for violating NCAA rules are not strong enough to deter coaches and their programs from potentially committing violations. Previous acknowledgments by Boxill (2003) and Seifried (2004) similarly supported such a business-ethic for competitive athletics to prompt an egocentric approach to coaching. Thus, coaching legitimacy involves norms centered on efforts to get away with as much as one can (Boxill, 2003; Seifried, 2004). Anecdotally, this rationale is enforced by current Big 12 Commissioner Bob Bowlsby. He stated: "[...] I think it's not an understatement to say cheating pays presently. If you seek to conspire to certainly bend the rules, you can do it successfully and probably not get caught in most occasions" (Trotter, 2014).

We find coaching performance, measured by its win-loss record in the recent completed season and the season prior, matters in the labor market for external promotions. Examining the odds ratios in both models, a one-unit increase in Winpct in the observed season (e.g., losing all games in the season to winning all their games in the season) results in the odds of a coach moving to a head coaching position that would be considered an external job promotion increasing by about a factor of 66. Examining winning percentage in the previous season, the odds of an external job promotion increases by a factor of 5.3 in both models. These results are consistent with Fee et al.'s (2006) findings for external promotions for NFL coaches.

The better the organization's performance over the last four years decreases the likelihood that the coach is externally promoted, albeit not to the degree of individual performance as the odds ratios in both models are approximately 0.08. This result intuitively makes sense. As Fee et al. (2006) remarked in regards to organizational success for internal promotion "a number two individual at a unit that is performing well may be very unlikely to have the opportunity to be promoted internally, because the likelihood that his superior will leave is particularly low during times of strong unit performance" (p. 825). Thus, an organization's sustained success influences a leader's decision to stay rather than move to a new job. However, a subordinate's likelihood of leaving may be increased due to the leader's desire to not move. The prospects of such a question present a viable future study that was also established in Greenberg and Smith (2007) during their review of assistant basketball and football coaches' contracts, which positioned those contracts as getting more complex due to the "increased recognition of the importance of good assistant coaches" (p. 27).

For minority coaches, there is a significant increase in the likelihood of receiving a promotion. A minority coach increases the likelihood of an external job promotion by a factor of 2. This result is not consistent with previous research by Bozeman and Fay (2013), who found that visible minorities negatively affects the career progression of coaches. In addition, the result in Table 2 is not consistent with previous research examining Division I-FBS coaching dismissals (Holmes, 2011) and compensation (Inoue et al., 2012). Still, this finding in the present study is consistent with recent research by Hill, Upadhyay, and Beekun (2014) examining the role that an executive being ethnically diverse or a female has on compensation. Their results found minority executives receive higher compensation when compared to white males. Hill et al. (2014) attributed this finding to the belief that minority status is something that is rare, valuable, and cannot be imitated. This rationale might explain why being a visible minority positively impacts the likelihood of a promotion as minority status is something the university can use as a competitive advantage for the organization.

Career experience was found to have a negative and significant effect on external job promotion. The rationale for this result is that as a coach gains in experience, he might become comfortable in his current job and might not be looking to take a new job, let alone a promotion. Reputation was found to be insignificant in the both models of Table 2. One would expect the more bowl games a coach appears in relative to his career experience would increase the likelihood of a promotion as this serves as a signal for quality of the coach. During the sample period, we do not find that it is the case. As the number of bowl games increased during this time period (Seifried & King, 2012), it may make it harder for university decision-makers to make decisions on a coach's quality based upon bowl game appearances.

We find the age of the program decreases the likelihood of a coach receiving an external promotion by a factor of 1. Given not only the history of the football program but also the status that these older programs have within Division I-FBS college football, this result is not surprising. It is also not surprising that coaching at a private school increases the likelihood of an external promotion. According to Table 2, coaching at a private school increases the likelihood of a promotion by a factor of 1.5. Generally, private institutions do not provide the resources for athletics in terms of funding football facilities, recruiting, and coaching staffs. Thus, coaches at private institutions may be actively looking for opportunities to leave. Finally, the variable controlling for the number of vacancies compared to the number of teams in Division I-FBS and the NFL increases the likelihood that a coach receives an external promotion. When examining the corresponding odds ratio in both models, the impact is relatively small with a factor of 1.08. This result is consistent with the idea that an increase in the number of openings provides an increase in the opportunity that a coach may be contacted in regards to changing universities or moving to the NFL. The impact, however, of increased opportunities is not as important compared to other variables such as recent on-field performance.

Conclusion

The present research attempted to understand how these actions impact head coach external job promotion. Examining coaching movement from the 1973-1974 through 2012-2013 seasons, the present research finds that sanctions did not impact the likelihood of a head coach's external job promotion. Results from the other variables provide information to the role that performance, age, experience, and the job market have on job promotions. These results, taken together, provide further evidence to the role that different factors have on promotion of executives, something that has received limited attention within both the sport and non-sport research regarding executive and upper management succession. Furthermore, we contribute to the existing research stream examining the consequences that sanctions have on universities, athletic departments, and coaches.

The present study is not without its limitations. First, in examining where the dependent variable and the independent variables were both non-zero, we find these occurrences to be small. While this might raise concern regarding Type II error, we feel the results provide initial interesting insight regarding external job promotion both in terms of sanctions and other factors. Second, the present study only examines the sanctions as whole and not each individual violation that the sanction includes. There may be many violations that compose the overall sanction that may provide additional information regarding the type of violation and if the cumulative number of those individual violations may impact promotion. Thus, future research should exploit the variation regarding the number and type of individual violations and the category(ies) in which these violations occur such as academic, recruiting of players, and improper benefits from alumni and boosters. The third limitation of the study is it only examines external promotion of head coaches. As found in the present study, strong sustained performance of an organization deters external job promotion of the head coach. However, strong sustained organizational performance may not deter assistant coaches to become head coaches at other universities. Future research should look at the determinants of external job promotion of assistant coaches and the impact sanctions may have on these job movements. Finally, the present research only looks at individuals that moved to a new job. It did not examine instances where a coach could have received an external promotion but chose to stay at his current university for an increase in salary or other benefits such as an increased recruiting budget. Future research could attempt to look at opportunities for individuals to move regardless of the actual final decision made by the coach regarding whether he moves to another university or the NFL.

References

Adler, E., Berry, M. J., & Doherty, D. (2012). Pushing "reset": The conditional effects of coaching replacements on college football performance. Social Science Quarterly, 94, 1-28.

Alexander, D. L., & Kern, W. (2010). Does athletic success generate legislative largess from sports-crazed representatives? The impact of athletic success on state appropriations to colleges and universities. International Journal of Sport Finance, 5, 253-267.

Allen, W. D., & Chadwick, C. (2012). Performance, expectations, and managerial dismissal: Evidence from the National Football League. Journal of Sports Economics, 13, 337-363.

Baron, J. N., Davis-Blake, A., & Biebly, W. T. (1986). The structure of opportunity: How promotion ladders vary within and among organizations. Administrative Science Quarterly, 31, 248-273.

Boxill, J. (2003). The ethics of competition. In J. Boxill (Ed.), Sports ethics: An anthology. (pp. 107-115). Malden, MA: Blackwell Publishing.

Bozeman, B., & Fay, D. (2013). Minority football coaches' diminished careers: Why is the "pipeline" clogged? Social Science Quarterly, 94, 29-58.

Brown, T., Farrell, K. A., & Zorn, T. (2007). Performance measurement & matching: The market for football coaches. Quarterly Journal of Business and Economics, 46, 21-35.

Chressanthis, G. A., & Grimes, P. W. (1993). Intercollegiate sports success and first-year student enrollment demand. Sociology of Sport Journal, 10, 286-300.

d'Addona, S., & Kind, A. (2014). Forced manager turnovers in English soccer leagues: A longterm perspective. Journal of Sports Economics, 15, 150-179.

Day, D. V., Gordon, S., & Fink, C. (2012). The sporting life: Exploring organizations through the lens of sports. Academy of Management Annals, 6, 397-433.

Day, J. C., & McDonald, S. (2010). Not so fast, my friend: Social capital and the race disparity in promotions among college football coaches. Sociological Spectrum, 30, 138-158.

Devaro, J., & Brookshire, D. (2007). Promotions and incentives in nonprofit and for-profit organizations. Industrial & Labor Relations Review, 60, 311-339.

Fee, C. E., Hadlock, C. J., & Pierce, J. R. (2006). Promotions in the internal and external labor market: Evidence from professional football coaching careers. Journal of Business, 79, 821-850.

Frick, B., Barros, C. P., & Prinz, J. (2010). Analysing head coach dismissals in the German "Bundesliga" with a mixed logit approach. European Journal of Operational Research, 200, 151-159.

Fulks, D. (2012). 2004-2012 NCAA Division I intercollegiate athletics programs report. Indianapolis, IN: National Collegiate Athletic Association.

Gibbons, R., & Waldman, M. (1999). A theory of wage and promotion dynamics inside firms. Quarterly Journal of Economics, 114, 1321-1358.

Grant, R. R., Leadley, J. C., & Zygmont, Z. X. (2013). Just win baby? Determinants of NCAA Football Bowl Subdivision coaching compensation. International Journal of Sport Finance, 8, 61-74.

Greenberg, M. J., & Smith, J. S. (2007). A study of Division I assistant football and men's basketball coaches' contracts, Marquette Sports Law Review, 18, 25-71.

Grimes, P., & Chressanthis, G. (1994). Alumni contributions to academics: The role of intercollegiate sports and NCAA sanctions. American Journal of Economics and Sociology, 53, 27-40.

Grusky, O. (1963). Managerial succession and organizational effectiveness. American Journal of Sociology, 69, 21-31.

Hill, A. D., Upadhyay, A. D., & Beekun, R. I. (2014). Do female and ethnically diverse executives endure inequity in the CEO position or do they benefit from their minority status? An empirical examination. Strategic Management Journal, 36, 1115-1134.

Holmes, P. (2011). Win or go home: Why college football coaches get fired. Journal of Sport Economics, 12, 157-178.

Humphreys, B. R. (2006). The relationship between big-time college football and state appropriations to higher education. International Journal of Sport Finance, 1, 119-128.

Humphreys, B. R., & Mondello, M. (2007). Intercollegiate athletic success and donations at NCAA Division I institutions. Journal of Sport Management, 21, 265-280.

Humphreys, B. R., & Ruseski, J. E. (2009). Monitoring cartel behavior and stability: Evidence from NCAA football. Southern Economic Journal, 75, 720-735.

Inoue, Y., Plehn-Dujowich, J. M., Kent, A., & Swanson, S. (2012). Roles of performance and human capital in college football coaches' compensation. Journal of Sport Management, 26, 73-83.

Kahn, L. M. (2000). The sports business as a labor market laboratory. Journal of Economic Perspectives, 14, 75-94

King, G., & Zeng, L. (2001). Logistic regression in rare events data. Political Analysis, 9, 137-163.

Longenecker, C. O., & Fink, L. S. (2008). Key criteria in twenty-first century management promotional decisions. Career Development International, 13, 241-251.

Maddala, G. S. (1983). Limited-dependent and qualitative variables in econometrics. Cambridge, UK: Cambridge University Press.

Morrow, P. C., McElroy, J. C., Stamper, B. G., & Wilson, M. A. (1990). The effects of physical attractiveness and other demographic characteristics on promotion decisions. Journal of Management, 16, 723-726.

National Collegiate Athletic Association. (2012). 2012-13 NCAA Division I manual. Retrieved from http://www.ncaapublications.com/productdownloads/D113.pdf

National Collegiate Athletic Association. (2013). Enforcement. Retrieved from http://www.ncaa.org/wps/wcm/connect/public/ncaa/enforcement

National Collegiate Athletic Association. (2014). 2014-15 NCAA Division I manual. Retrieved from http://www.ncaapublications.com/productdownloads/D115AUG.pdf

Otto, K. A. (2005). Major violations and NCAA 'powerhouse' football programs: What are the odds of being charged? Journal of Legal Aspects of Sports, 15, 39-57.

Pfeffer, J., & Davis-Blake, A. (1986). Administrative succession and organizational performance: How administrator experience mediates the succession effect. Academy of Management Journal, 29, 72-83.

Rhoads, T. A., & Gerking, S. (2000). Educational contributions, academic quality, and athletic success. Contemporary Economic Policy, 18, 248-258.

Rosenfeld, R. A. (1992). Job mobility and career processes. Annual Review of Sociology, 18, 39-61.

Seifried, C. S. (2004). Gamesmanship: An acceptable strategy in athletic competition. Journal of Contemporary Athletics, 1, 341-351.

Seifried, C. S., & King, B. (2012). Examining the history of Division I (FBS) bowl payouts: Current Bowl Championship Series (BCS) and non-BCS institutions and the growth of bowl games into BCS controlled contests. Sport History Review, 43, 206-228.

Smith, D. R., & Abbott, A. (1983). A labor market perspectives on the mobility of college football coaches. Social Forces, 61, 1147-1167.

Soebbing, B. P., & Washington, M. (2011). Leadership succession and organizational performance: Football coaches and organizational issues. Journal of Sport Management, 25, 550-561.

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.

Trotter, J. (2014, July 22). Bob Bowlsby sees bleak landscape. ESPN.com. Retrieved from http://espn.go.com/college-sports/story/_/id/11243234/bob-bowlsby-big-12-commissioner-says-cheating-pays-ncaa- enforcement-broken

Winfree, J. A., & McCluskey, J. J. (2008). Incentives for post-apprehension self-punishment: University self-sanctions for NCAA infractions. International Journal of Sport Finance, 3, 196-209.

Winter-Ebmer, R., & Zwelmuller, J. (1997). Unequal assignment and unequal promotion in job ladders. Journal of Labor Economics, 15, 43-71.

Endnotes

(1) There was a recent special issue in 2011 on NCAA infractions in Criminology and Public Policy.

(2) Ties occurred in the early years of the sample and count a 0.5 wins.

Brian P. Soebbing (1), Patrick Tutka (2), and Chad S. Seifried (2)

(1) Temple University

(2) Louisiana State University

Brian Soebbing, PhD, is an assistant professor in the School of Tourism and Hospitality Management. His research interests include the strategic behavior of sports leagues and teams, as well as the social and economic impacts of gambling.

Patrick Tutka is a doctoral candidate in sport management in the School of Kinesiology. His research area involves sport facility management and focuses on innovation with special emphasis on college and professional venues through debates about renovation and new construction and stadium/ballpark leases.

Chad Seifried, PhD, is the L.M. "Pat" and Mildred Harrison Endowed Professor in the School of Kinesiology. His major research interests embrace the use of historical methods to review the management history associated with organizations, individuals, and sport facilities to help highlight the contribution of history within contemporary management settings.

Table 1. Summary Statistics

Variable        Mean    Std. Dev.    Min     Max

Promo          0.034      0.182       0       1
Winpct         0.524      0.224     0.000   1.000
Winpct(t-1)    0.531      0.221     0.000   1.000
SchoolWinpct   0.520      0.178     0.034   0.946
CoachAge       48.887     7.738      30      85
Minority       0.031      0.172       0       1
CarExp         9.478      6.991       2      46
Rep            0.321      0.313     0.000   1.000
FootballAge    72.864    35.322       5      138
Private        0.183      0.387       0       1
Openings*100   17.563     3.464     9.09    24.29
Violations     0.021      0.067     0.000   0.800
Secondary      0.010      0.042     0.000   0.600
Major          0.010      0.035     0.000   0.400
n = 3,913

Table 2. Logistic Regression Results and Odds Ratios

                                  Model 1

Variable           Coef.     Robust     p-value    Odds
                            Std. Err.             Ratio

Winpct             4.187      0.588     < 0.001   65.841
Winpct(t-1)        1.671      0.717      0.020    5.317
SchoolWinpct       -2.582     0.784      0.001    0.076
CoachAge           -0.036     0.016      0.028    0.965
Minority           0.666      0.325      0.041    1.946
CarExp             -0.053     0.023      0.019    0.948
Rep                -0.143     0.363      0.694    0.867
FootballAge        -0.007     0.003      0.018    0.993
Private            0.441      0.207      0.033    1.554
Openings*100       0.080      0.028      0.004    1.084
Violations         -1.163     1.228      0.344    0.313
Secondary           ---        ---        ---      ---
Major               ---        ---        ---      ---
Constant           -4.465     0.934     < 0.001    ---

                                 Model 2

Variable           Coef.     Robust     p-value    Odds
                            Std. Err.             Ratio

Winpct             4.188      0.588     < 0.001   65.883
Winpct(t-1)        1.672      0.718      0.020    5.322
SchoolWinpct       -2.592     0.784      0.001    0.075
CoachAge           -0.036     0.016      0.028    0.965
Minority           0.673      0.325      0.038    1.961
CarExp             -0.053     0.023      0.021    0.948
Rep                -0.142     0.364      0.696    0.867
FootballAge        -0.007     0.003      0.017    0.993
Private            0.441      0.207      0.033    1.554
Openings*100       0.080      0.028      0.004    1.084
Violations          ---        ---        ---      ---
Secondary          -1.901     2.177      0.382    0.149
Major              -0.382     2.595      0.883    0.682
Constant           -4.327     0.948     < 0.001    ---