NFL Crime: Female Execs May Hold Key to Reducing Arrests

·5 min read

Today’s guest columnist is Prof. Rick Burton of Syracuse University.

With the NFL’s annual college draft bearing down upon us, spreading 259 players across seven rounds and 32 teams, there is no question biometric and performance analytics will take center stage in Cleveland this week.

There will also be cursory (if not significant) mentions about each player’s character. But what if a group of highly credible researchers suggested they could predict which NFL teams have the biggest potential for off-field issues and headline-grabbing arrests in the next year?

Two Syracuse University professors and one of their former students have done just that, with recently published research in the Journal of Organizational Behavior, titled “Women Executives and Off-the-Job Misconduct by High-Profile Employees: A Study of National Football League Team Organizations.”

The research was co-authored by SU’s Dr. Mary Graham, her colleague Dr. Bhavneet Walia and Chris Robinson, a sports law attorney licensed to practice in Washington, D.C.

The surprise statistical variable that emerged as the single greatest predictor of NFL arrests? It was whether the team had a critical mass of female executives (two or more) associated with the club.

As the authors noted in their conclusion, “Serious off-the-job misconduct by high-profile employees is not an uncommon occurrence in professional sport team organizations, media and entertainment firms, and public-facing institutions.” But “firms searching for preventive and remedial solutions to misconduct should consider a basic structural solution to this problem, [by] ensuring there is a critical mass of women on the top management team.”

Graham’s academic specialty is human resources, and organizations like the NFL’s member clubs have long interested her. In Robinson, Graham was able to mentor and partner with a top student finishing his sports management undergraduate degree in less than four years. He worked to unpack nine variables linked to the significant police arrests of NFL players from 27 teams (five had incomplete staffing data) during the 12-year period of 2003-14.

Looking at each team, the trio searched for correlations between the clubs with the greatest number of violations and statistically relevant factors.

Given that consistently, a handful of NFL players were arrested each year, the researchers felt they had enough data over a long enough period of time to start analyzing factors. It was relatively easy looking at arrests in the aggregate but far more time-consuming dissecting franchise quality, the head coach’s age and tenure, team win percentages, annual payroll, and difficult-to-find information on team employee headcount.

Academic findings (or an insightful book, like Pros and Cons by Jeff Benedict and Don Yaeger, published in 1998) often cause NFL executives to discredit the research methodology (“they probably missed data”), the data set (“it’s old”), the authors (“what do they know about football?”), or the semantics of language (“Criminality? I bet they counted every little ticky-tack speed limit citation”).

Second-guessing of this nature, even stolid defensiveness, is unfortunate. But here’s a key point: The sports industry is always under massive scrutiny, and entities like the federal government remain curious whether league leaders are sufficiently proactive.

There is also the race issue, and it cannot be dismissed that black athletes in America have long suffered social injustices. Discussing data about arrests is unproductive if one fails to recognize police bias and deliberate persecution, as Graham, et al, acknowledge in their paper.

To be clear, the Graham-led research wasn’t motivated by animus or a desire to demonize football players. The purpose of this new NFL research, not unlike studying concussions, was to proactively address a problem. What she and her colleagues found may not surprise keen observers of the NFL. During the time frame studied, Minnesota (39), Denver (34), Tennessee (34) and Cincinnati (33) produced the greatest number of arrests. Dallas, Houston, New England and Philadelphia—which has long been known for hiring women to front-office jobs—had the fewest, with 11 each.

For a lot of people, negative-sounding data creates a problem. That’s because owners, general managers and coaches want (and need) proper context if the findings suggest mismanagement: “There’s no problem here. Just an unfortunate matter we’re dealing with. This is an aberration.”

Economists, on the other hand, generally want predictive models that hold up under the tightest numeric scrutiny. They want to understand clichés or generalizations and better understand the “why” of the matter. The concept has already made its way into sports, via the Moneyball analytics revolution. As Brad Pitt’s cinematic version of Billy Beane notes, Players either get on base or they don’t.

Would the same approach to statistics not work with player arrests?

Walia, an econometrics expert, made clear that some NFL teams “are doing a lot better than others in preventing player arrests, and that’s apparent because they have more women executives. We believe teams with a significant number of women executives have more inclusive team cultures and more effective player development efforts, which, in turn, leads to reduced instances of player misconduct.”

It’s easy for journalists, columnists and professors sitting in their ivory towers to suggest arrests in pro sports are destructive. That’s low-hanging fruit. But shouldn’t the team owners, the ones benefiting from billion-dollar media and gambling agreements, analyze whether their team is a statistical stalwart or a breeding ground for harmful misconduct?

The researchers ask, “Why not?” Particularly because, as this week’s NFL Draft makes abundantly clear, the owners already collect and scour the data for everything else.

Burton ( is the David B. Falk Professor of Sport Management at Syracuse University and former Commissioner of Australia’s National Basketball League.

More from