Domestic Violence Is A Giant Problem. Is It Also a Football Problem?

In the comments to the latest Linky Friday post, Chris links to this post by Dan Diamond at Forbes, which in turn draws heavily from this post and this post by Benjamin Morris at 538. In the post, Diamond argues for scrutinizing the role of football in American life by arguing that the risk of concussions in football is massively higher than in other sports, with severe consequences, and more dramatically that the NFL’s domestic violence problem is unique and concludes that “[s]omething about football, or at least about the men who play it, is a factor in causing violence at home.” More specifically, he suggests that the higher domestic violence rate is either caused by “brain injuries reducing players’ ability to regulate emotion, or the culture of a game that rewards men to be violent on the field, and empowers them off it.”

He may indeed well be right in the end, but if so, he hasn’t proven it. First, the Benjamin Morris article on which he relies for the argument that NFL players are significantly more likely to commit domestic violence is, at most, inconclusive. Morris – to his credit – makes very clear that his findings, which claim that NFL players may be somewhat less likely to commit domestic violence than the average 25-29 year old male, but that they are substantially more likely to do so than would be expected after controlling for income. Specifically, those assumptions are (1) that current income, rather than childhood socioeconomic status, is a particularly useful factor for comparing NFL players’ domestic violence rates with the general population; and (2) that the household income of general population victims of domestic violence is a reasonably accurate approximation of perpetrators’ incomes that can be meaningfully compared with the incomes of NFL arrestees.

I think Morris understates the potential effects of the latter assumption, but probably not enough to meaningfully alter his conclusions. However, the use of current income rather than childhood socioeconomic data as an appropriate adjustment seems highly problematic, particularly to the extent it is used to make assumptions about 25-29 year old males making that income. In the context of analyzing professional athletes, this comparison effectively assumes that one who grows up in a difficult or at least non-privileged environment can largely escape the impact of those conditions in just a short period of time making an obscene amount of money. That of course is not how humans operate. Moreover, this assumes that the average high-income household comes from a background comparable to the average NFL player, which seems unlikely, to say the least. Finally, there is a “lottery effect” element of the NFL that we would want to find a way of controlling for before using income as a comparison – the average person making $75,000 or more a year does not do so at a particularly young age, and even if he does, does not usually do so overnight, but instead gradually and incrementally builds to that.

None of which means that NFL players are less violent than they theoretically should be, just that it’s difficult to draw much of a conclusion from the data without making highly questionable assumptions with the potential to significantly swing the analysis. In any event, my guess is that a study that was somehow able to properly control for all variables would indeed find NFL players are somewhat more likely to engage in domestic violence than the average person, but I’m skeptical that it would be a particularly large difference, and certainly not on the scale that Morris suggests.

That said, if we’re concerned about whether football generally and the NFL particularly has a unique problem with domestic violence, the best way to come up with a definitive answer here would be to compare domestic violence arrest rates with other sports. In particular, a comparison between domestic violence arrest rates in the NFL and NBA would be quite usefual, as the NBA is the only other Big 4 sport where the overwhelming majority of athletes are both born and raised in the US, and in which the athletes just about always go from making nothing to making obscene amounts of money overnight.

It’s quite possible that the NBA’s domestic violence rates are substantially below the NFL’s – it certainly seems fairly rare to hear about NBA players getting arrested for anything involving violence, though it’s possible that’s just a function of it having a fraction of the players the NFL has. [UPDATE: In the comments, Nob links to statistics indicating that, at least for domestic violence, NBA players in fact seem to have a somewhat higher arrest rate. It should be noted that the NBA’s raw numbers for any year are in fact quite small due to the relatively few players in the NBA, but the higher rate is consistent across several years’ of data, so this does not seem to be a question of statistical variance. This also strongly undermines the argument that concussions or the violent on-field nature of football are a statistically significant factor in off-field violence by football players.].

But even then, it is not clear whether we’d be able to say that the problem of off-field violence is inherent in the game of football, as Diamond seems to insist, or is specific to the particular structure and rules of the NFL, which could easily be altered without any cultural impact whatsoever. Specifically, there’s another little-discussed but clearly important variable: the NFL is almost certainly the most stressful professional sport by far, and it goes without saying that people who are emotionally stressed are more likely to violently lash out.

No other sport – really, no other profession in the world – has athletes with so much on the line on a daily basis. The average career is only either 3 or 6 years, depending on how you make the calculations. There are 53 people on the in-season roster, but at the start of training camp, there are 90. That means that only a handful of players on each team can truly be said to be guaranteed a roster spot from season to season.

Unlike other sports, contracts are not guaranteed; combined with the possibility of getting cut, this means that a player can go from dirt poor college student to young one percenter with the world at his feet to bankrupt with limited career potential, all within the span of just a few months. Even a moderately successful veteran, at the ripe old age of 27 or 28, can easily go from making the league minimum to making millions to making absolutely nothing overnight, just months after signing his first big contract.

Unlike the other major sports, there’s not even much in the way of viable options in lesser leagues. A basketball player out of the NBA can, if need be, go play in Europe or China and make a respectable middle-class to upper-middle class income. A baseball player cut from the big league roster might just go down to AAA (and even some of independent minor leagues, for whom a former big leaguer can be a big draw), where he can at least make a middle class income, and may have other opportunities in Japan or, for less money, Latin America. And a hockey player can go to Europe. A football player’s options are basically the Arena League, which pays about $14,000/year, or the CFL, where he can make a respectable upper middle class income, but which only has 9 teams.

And just making a roster at the outset of the season hardly ends the stress – listen to enough current and retired players talk and you’ll often hear some variation on the phrase “you have to prove yourself every week if you want to survive in the NFL.” There are only 16 games to make an impression and build a reputation that will help a player stick around or get a better contract with more guaranteed money – every single play can substantially impact a player’s job security or earning potential. For players on the roster fringe, a freak injury to someone who plays a completely different position could mean that you will get released so the coach can make roster space to replace the injured player. And, of course, someone gets at least a one or two week injury in more games than not. Worse, several players per team per year are likely to wind up with a season-ending injury at some point, and in the world of the NFL’s business model, a season-ending injury can easily wind up causing your team to give you a one-time injury settlement and release you, with that season-ending injury functionally becoming a career-ending injury.

Without much in the way of a fallback option, for many the rational response to being cut, whether during the preseason or regular season, may even be to spend all of one’s time working out and training by oneself in the hope that another NFL team will pick you up. If one gets cut in the preseason, he might work out in the hopes of being picked up midseason when another player gets injured – the off-chance of even getting two or three game checks might well be more than what a player could hope to earn outside of football for several years combined. If one gets released midseason or at the end of a season, there’s always an excellent chance that someone will want to bring you in for training camp next year, giving you a chance to compete for another roster spot.

The NFL’s business model, then, seems like it not only encourages players to constantly fear being cut but perhaps inadvertently also discourages players from seeking out permanent fallback options should they be cut.

It’s hard to imagine a more stressful career path that still involves large or even upper middle class incomes. Stress causes violence.

It’s very possible, perhaps likely, that pro football players are more violent when they walk off the field than their most comparable peers. But that may or may not mean that off-field violence is a football problem so much as it is an NFL problem.

(Post image via FiveThirtyEight.com)