Football has a numbers problem. We’re told that 1 in 3 NFL players will suffer from Alzheimer’s or dementia at younger ages than their peers; that they’re four times more likely to die from degenerative brain diseases; that 78 percent will be broke or nearly so within two years of retiring; that they may be disproportionately inclined to beat their wives; and that they live, on average, to the age of 55 or 60. It’s enough to make you wonder if banning football would be good for America.
But there’s another numbers problem here: Each of the figures cited in the paragraph above is false, misleading, or incomplete. They’ve nonetheless become ammunition for anyone who wants to criticize the sport, the NFL, or its players. If we’re going to talk about the social value of the NFL—or, more importantly, its social costs—then it’s time we reckoned with the stats. But which numbers, if any, can we trust?
Some of the most widespread statistical mistakes are simply sloppy. The recent claim that 1 in 3 pro football players will experience accelerated cognitive decline derives from actuarial reports prepared for both sides in one of the NFL’s recent concussion lawsuits. Despite headlines to the contrary in places like the New York Times, these reports pegged the rate at 28 percent—closer to 1 in 4 than 1 in 3. The actuaries predict that players will suffer from cognitive decline at younger ages than their peers, but they don’t imply—as many commentators have—that a full 28 percent will suffer from a rare, early-onset form of illness. Instead, they say that roughly 1 in 4 players will experience cognitive decline at any point during their lives, including when they’re very old. As for how those numbers might relate to those you’d find in a normal, non-football-playing population, both reports can only guess. The plaintiffs’ side admits that, for the majority of the diagnoses involved, “no peer-reviewed published research on the risk to professional football players relative to the general population [could be] identified.” Faced with the same problem, the NFL’s accountants “at each stage sought to err on the side of overstating the number of players who will develop” neurological conditions.
To make things more confusing, it’s often said that athletes in the NFL tend to live much shorter lives than other people. If that’s true, the rate of 28 percent would be extra shocking, since most people—even football players—get dementia when they’re senior citizens. But that statistic has it backward: The best available data, analyzed by researchers at the National Institute for Occupational Safety and Health on behalf of the NFL Players Association, suggests that footballers have significantly extended lives. (The claim that players tend to die at 55 or 60, which persists in stories published by outlets like the Boston Globe and the Washington Post, was never based on published data, as author Daniel Flynn points out, but rather on a pair of guesses made by former players in the late-1980s.) Since retired athletes are in fact more likely to reach old age—they’re healthier than you and me, all things considered—we might expect them to have a higher rate of Alzheimer’s and dementia than everybody else.
What about the fact that NFL players are four times more likely to die of Alzheimer’s or ALS—a health advisory that has made its way into the Today Show, Good Morning America, CNN, and Time, among many other media outlets? That figure comes from the same NIOSH analysis, which found that 0.17 percent of football players die of ALS, compared with an expected rate of 0.04 percent among controls. It’s a fourfold increase in risk, but to state it that way exaggerates the harm. Rates of the disease, for both NFL players and controls, are very low. In absolute terms, the increase in risk that corresponds to playing football—and presumably to getting bashed in the head repeatedly—is 0.13 percentage points. For every 770 men who play the sport on a professional level, we can expect one extra death from ALS. (Extra deaths from Alzheimer’s are even more unusual.)
Any extra death is cause for grave concern, but if you look at other, much more common deadly conditions, the change in risk goes the other way. The same dataset suggests that for every 770 football retirees, we should expect 13 fewer deaths from cardiovascular disease and 14 fewer from cancer. So while it’s true that Alzheimer’s and ALS rates among NFL athletes could reasonably be described as “through the roof,” the number of players’ lives saved from heart disease and cancer exceeds the number of lives lost to those diseases by 2,150 percent. And if you look at all causes of death combined, the difference is even more extreme: The NIOSH study found 8.47 percent fewer deaths among their pool of athletes than they would have otherwise expected. That corresponds to 54 football players living longer lives for every one who dies of Alzheimer’s or ALS.
The same logic can be applied to player actions off the field. Studies find that NFL players are arrested for assault at about half the rate of age-matched peers. For all the talk about a culture of violence in the league, pro athletes appear to be a relatively gentle bunch. Put these stats together and it’s tempting to conclude that football is a boon for public health. If we could press a button that turned every young man in America into a professional football player, we’d be crazy not to follow through. A nation of pro athletes would be healthier and longer-lived, and the spouses of those super-people would be safer, too.
But the methods used to find these stats raise a familiar and important question: Should football players really be compared to average men their age, of any race or body size or income level? How much does the choice of analysis affect its outcome?
Let’s start with money. Writing at FiveThirtyEight, Benjamin Morris notes that while rates of domestic abuse among football players are lower than the national average, they’re “extremely high relative to expectations” based on how much money players make. Women living in households that earn more than $75,000 per year suffer from domestic abuse at a rate that’s less than 40 percent of the national average. Meanwhile, the median salary in the NFL is around $800,000. If researchers controlled for player wealth, the league’s domestic violence stats would look far more ominous.
The same goes for player health: Rich people live much longer than poor people, so any analysis that ignores the effects of money could be said to underestimate the risks of playing football. The NIOSH analysis controls for race but not income. (Approximately two-thirds of NFL players are black.) As a result, when those numbers tell us that football players live longer than their peers, they’re matching up a group of multimillionaires with people who, on average, make a whole lot less. (By the way, there’s no empirical basis for the oft-repeated claim that 78 percent of football players are bankrupt or nearly broke within two years of retiring. It appears to have been popularized by a sociologist and former player named Ken Ruettgers, who told me that “nobody really knows where it comes from.”)
It might seem prudent to compare football players to men in their income bracket, but other factors may be more important. The rate of domestic violence against black women is 35 percent higher than the corresponding rate for white women, and black men die more than five years younger, on average, than white men. If researchers controlled for players’ race, the NFL’s domestic-violence rates could look even better than they did before, and any health “benefits” that come with playing football would go up, too.
So is it better to control for income or race, or should studies strive for both? And what about body size? Large stature has been correlated with both reduced lifespan and increased rates of violence. (The average NFL athlete is 6 feet 2 inches tall and 246 pounds, with a technically “obese” BMI of 31.6.)
These may sound like simple questions, but they’re exceedingly difficult to answer. To some extent, the best approach depends on how you think about the NFL, and what point you’d like to make.
If you’re curious about the specific risks of playing football, as compared to those affecting any elite-level athlete, then you ought to make controls as stringent as possible. Perhaps you’d compare retirees from the NFL and NBA, and factor in their race and lifetime earnings, too. That kind of analysis—which has never been done, so far as I know—might offer evidence on whether football needs to change its rules. Is the game as safe as it can possibly be? A study of former world-class athletes from Finland controlled for their subsequent career paths, among other things, and found that hockey and basketball players live longer than wrestlers and weightlifters.
But if you’re more concerned about the nature of the sport—I mean to say, if you’re evaluating football’s right to exist—then you should take a very different approach. In that case, it’s better to construe a career in the NFL as a sort of lifestyle intervention, and one that happens to be disproportionately applied to young, black men who grew up in the poorest parts of the country.
The NIOSH study, which aimed to “investigate concerns that players were dying prematurely,” did just that: Researchers treated players as if they were a group of patients, and their NFL careers as a type of treatment. They controlled for race and gender, because those traits defined their patient population before the treatment started. On the other hand, they did not control for player wealth because money can be viewed as a treatment outcome—it flows directly from the game. Football players tend to do quite well financially and rise in social status, just as they also tend to exercise and rarely smoke. These factors make them healthier, and as such they’re no less relevant to the central research question than the fact that players are disproportionately likely to get hit in the head on a daily basis or take anabolic steroids.
Such an analysis may not tell us much about how safe the NFL can be, but it does give us very useful information on the bigger question of how the sport affects the men who choose to play it, for better or for worse. Analyses that control for race but not wealth point to an important fact that’s often lost amid the outrage: Whatever the league’s policies on concussions and domestic violence, its athletes are less likely to die young, and less likely to beat their wives, than men from similar backgrounds.
That doesn’t mean the NFL is good for everyone. Each year, around 2,000 athletes enjoy the benefits of employment by the league, while those who try and fail to reach the pros could be left to pay the costs. It’s also possible that football’s evils reach beyond the players to the rabid fans, who may internalize the violence it promotes. (Domestic abuse incidents do seem to increase, slightly, in places where the local NFL team has just suffered an unexpected loss.)
Nor should the numbers mitigate our outrage at the men (and it’s pretty much all men) who run the NFL. Players, as public figures, should be held to a higher standard. The fact that most Americans are sedentary and overweight, for example, or that domestic violence is terrifyingly routine among the guys who watch football instead of play it, has little bearing on the question of whether the league’s sponsors should abandon ship or whether Roger Goodell should keep his job. It doesn’t excuse the NFL to point out that rates of domestic violence are very high overall, or that the absolute risk of Alzheimer’s and ALS is pretty low among the sport’s retirees. If league officials can figure out a way to lower rates of assault or brain injuries among its players, they have a moral obligation to do so.
Still, let’s keep the sins of football in perspective. The leaders of the NFL have behaved atrociously, and they’ve been worse than inattentive to the safety of players and their families. But the sport they represent isn’t killing its employees. It’s not abusing women. It isn’t adding to the sum total of human suffering. If you have a chance to play in the NFL, you should probably take it.