I must say I am disappointed with what the New York Times has placed front and center on their home page. A story that begins with this: 

"Finally, some happy news for President Obama. Despite a common assumption that life in the Oval Office prematurely ages its occupants and speculation that it may even shorten life spans, a new statistical analysis has found that most presidents have actually lived longer than other American men their age."

First of all, it may be false. According to the paper in JAMA it reports on, the 34 presidents who have died of natural causes died at an average age of 73.0. Men their age at the time of inauguration lived to 73.3. Let's say we include the living presidents. That 73.0 might increase to more than 73.3, but probably not enough to make the difference statistically significant. 

Second, that opening doesn't even make logical sense. Here's how it's meant to read: "Despite the assumption that X, a new statistical analysis has found that not-X." Here's how it reads to me: "Despite the assumption that X, a new statistical analysis has found that Y," where X and Y are compatible. We cannot conclude that being president doesn't hasten aging based on evidence that presidents don't die younger than other men. To assess whether being president hastens aging, presidential age of death should not be compared to non-presidential age of death; it should be compared to what age the presidents *would* have died, which is impossible to know. 

Presidents can't be compared to non-presidents because too many factors other than serving in the Oval Office are at play. As the paper notes, "all but 10 presidents were college educated, had considerable wealth, and had access to the best medical care in their era. Level of completed education and its related social and economic status correlates have documented powerful effects on longevity today and probably had even more powerful effects centuries ago." It's quite possible that these men of means and stamina would have outlived their peers by five years, but that they lost these five years from the stress of serving as president. So in the end, based on the data presented in the paper, we have no way of knowing whether the Oval Office shortens life spans.

I can't place all of the blame for this story on the New York Times writer. The researcher also fails to bring up the possibility that being president does in fact increase aging but that it's countered by their otherwise strong health and good medical care. But, to use a phrase I came across recently, it is our duty to be not just science reporters but science critics.


Addendum: I note above that the ideal, although impossible, analysis would be to compare how old presidents were when they died to how old they *would* have been if they hadn't been presidents. While this is true, a decent substitute—one still better than comparing presidents to all other men their age—would be to compare them to men matched for age, health, socioeconomic status, and perhaps a few other relevant factors. 

You are reading


A Con Artist’s Best Accomplice Is You

Getting hustled is not a spectator sport.

Why Candy Crush Is Like Life

The cognitive science of Candy Crush Saga

Computers Judge Personality Better Than People Do

A few Facebook Likes say a lot.