Will Your Spare Tire Save Your Life?

A new paper (gated version here) by Michael Sivak, Brandon Schoettle, and Jonathan Rupp takes a look at what keeps people alive in fatal traffic accidents. The authors find that women drivers involved in fatal crashes are more likely to die than men, and, not surprisingly, unbelted drivers are 2.1 times more likely to die than belted drivers. The effect of body mass index (BMI) is more complicated. For men wearing seatbelts, a higher BMI seems to offer some protection: “For example, the probability of being killed, if involved in a fatal crash, is 22 percent lower for belted male drivers with BMI between 35 and 50 as opposed to those with BMI between 15 and 18.4.” The opposite is true for unbelted males. For belted females, having a normal BMI gives you the best shot at survival.[%comments]

f tripoli

a good argument for eating while driving... especially when the roads are slick !!

Fritz Mills

Does this imply anything about passengers' relative survival rates? What about the relative frequency of an accident being described as "fatal" (as a subset of all accidents), with respect to the gender of one or more of the drivers involved? Is there really any practical, useful information in this study? Who paid for it?


BMI range of 35-50 is very high. 25-30 is overweight, and 30-35 is obese, so we're talking extremely obese people here.

Likewise, BMI range of 15-18.4 would be considered underweight, although not quite as much.

I'm more interested in the more realistic ranges of 19-35, where a majority of males will lie.


It seems like there are a wide range of factors to consider. I don't see how it could be used to answer the question I am really interested in: Will my extra pounds help save me in a crash?


A plane crashed right on the border between two countries. In which country did they bury the survivors?

If you survive a car accident, isn't it by definition not fatal?


Are the results corrected for vehicle type? I'm thinking the the BMI factor could just be a correlation because high BMI people are more likely to buy and drive gigantic land barges and trucks, whereas low BMI people are more likely to be in a compact or sports car?


Dan - Not if you survived but somebody else didn't.


from my boots-on-the-ground experience:

-for women, high BMI correlates with junky cars, poor basal fitness, lower intelligence, and other poor health habits.

-for men, also the above, but less strength of the correlation.

-six inches of extra fat tissue and a seatbelt protects everyone in an MVA


So the study is of accidents in which at least one person died, but in which not necessarily one person survived? Seems like a silly way to qualify accidents. Something like "survival rates of multi-car collisions involving relative velocities of at least 30mph" would be more less arbitrary.

And look at me now, I'm usually the one sitting at my computer thinking "why do these commentors nit-pick these studies, how useless." But I guess at some point the stupidity of it all kinda gets to you....



I'm not sure if that is entirely accurate. I know many smaller girls who prefer SUVs and other big cars because they feel safer in them and prefer the higher vantage point provided by them. Obviously, this is only anecdotal, but I think it is hard to correlate body type to vehicle type, as intuitive as the connection might be.


Pretty ridiculous generalizations to draw, eh?


Without seeing the whole study, it's hard to know, but I agree with Dan. Picking accidents where someone was killed, and someone else may or may not have been is fraught with complicating factors. If someone drunk doesn't wear a seatbelt and T-bones a car at 35mph he could easily be killed while the belted passengers in the Volvo he hit are not killed. That's hardly the same analysis as a head-on collision at 70mph where everyone is belted but killed anyway.


I'm usually pretty good at deconstructing these kinds of statistical statements, but "women drivers involved in fatal crashes are more likely to die than men" had me scratching my head for a while.

Obviously, a fatal crash, in this case, means a crash in which at least one person was killed and there may or may not have been survivors. And I can assume that "men" is short for male drivers involved in a fatal crash. But I'm still not sure what we're comparing? Does this mean men are more often among the survivors of fatal crashes? Are we only comparing mixed-gender crashes? How is this statement even arrived at?


I knew all those Snickers, extra Big Macs, and fried apple pies would come in handy! Not only can I outlast everyone if they play a Survivor in Antarctica, but it is extremely unlikely that I can be successfully kidnapped!

And now to find that I am actually SAFER in an accident than my vegan snob friends. Oh, the sweetness of it all!

In fact, I think I'll go out right now and take on a train. I am big, soft...and INVINCIBLE!


I wonder if gender is irrelevant and what we're actually seeing is an effect of height -- where things tend to hit you, whether the extra padding is more likely to cushion or crush.

Eric M. Jones


1) Way too many uncontrolled-for variables here.

2) My daddy always said the seatbelts' secondary duty is to keep your butt planted firmly in a position to control the car.

3) I have always figured your chance of survival was lowered by wearing an expensive watch or jewelry. Those paramedics don't make much money. Your chance of survival being carried out of the woods from a crash if you are wearing a $99,000 Breitling Bentley GMT watch is zero...particularly if I carry you.

4) Favorite humor: "I want to die peacefully in my sleep like my Dad, not terrified and screaming like the other people in his car."


could be more to do with age than with the supposed shock resistance of adipose tissue.

I'm basing this on the assumption that younger people tend to be skinnier, and *also* tend to get in more accidents.


Could the bit about women drivers be related to another recent post on the Freakonomics blog? When a couple is riding together, the woman is most likely the passenger. Therefore, the fraction of the time that women are driving with passengers would seem to be lower. If there are no passengers and she survives, it's not a fatal crash and it's not considered.

Of course, I don't care about how to better survive fatal crashes - I care about how better to survive crashes in general (or even better, to avoid crashes altogether and have no problem surviving the driving).

Another View

A crash is fatal if anyone, and not necessarily the driver, dies, so, yes, surviving a fatal crash is possible.

Seat belts save lives, but cause secondary injuries as yet another object with which the human body will, and, in fact, is meant to, collide in the event of a crash. That having additional fat between the belt and the vital organs might be helpful is not surprising.

The limitation as to the helpfulness of the fat factor in the case of women is curious. I wish I could read the paper, to learn whether the study distinguished between at-fault drivers and others, as well as whether drivers under the influence of alcohol or drugs were included, what types of crashes (e.g., head-on, side impact, one vehicle, etc.) were considered, and how those whose later deaths were attributed to injuries received in the crash were classified.

Perhaps some sort of mini-airbag on the back side of the belt might be in order, particularly for those without much natural padding.

I don't relish the idea of colliding with a seat belt, but, as a short person, I am more frightened of being killed by the airbags stored in the dashboard or in the steering wheel.



i'm interested in the survival differential between NORMAL BMI males and overweight males, rather than between underweight males and morbidly obese males. In addition, there may be so few males in the low BMI range that the comparison statistic is not statistically significant. regardless, it's not terribly surprising that body fat *protects* someone in a car crash.


This statistic can mean that current belts are designed to best protect seriously obese men, and normal-sized women. Did we really need another example of society encouraging men to eat, and women - to stay trim?

Perhaps not. Instead, we need better belts.