# What's the chance for one to die in a car accident in their life-time? [closed]

Considering the high number of car accidents, it must be around 5%, right?

-
The probability of dying in a car accident after one's lifetime is $0$. – Dilip Sarwate Mar 21 '12 at 12:28
Hi Alexa. Welcome to the site. Please consider adding some more detail and context to your question via an edit. This should make your interests and objectives more clear to other users and hopefully encourage more positive interaction. – cardinal Mar 21 '12 at 13:30
Please consider improving this to a form of a verifiable claim and asking on Skeptics.SE. – mbq Mar 22 '12 at 11:07

## closed as off topic by onestop, Andy W, rolando2, Greg Snow, mbq♦Mar 22 '12 at 11:05

Questions on Cross Validated are expected to relate to statistics within the scope defined in the FAQ. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about closed questions here.

I would also point out that there's going to be tremendous variation in risk across individuals due to substantial variation in exposure.

Police officers, for example, are at a very high risk because they drive around all day everyday and, in addition to that, are sometimes involved in risky high speed chases. Police officers are actually more likely to be killed in a car accident than by gunshot.

I say this merely to point out that looking at aggregate rates probably tells us little about actual individual risk, which may or may not be the number you actually care about. Things like miles driven per year, region/locale, and driving style (including seatbelt use) probably change your actual risk substantially.

This is my long winded way of warning you not to commit the ecological fallacy.

-

Statistics on cause of death, by country, are collated by the World Health Organization from national reports. I couldn't find a good summary on their site for this answer but the data are there somewhere; there is a summary in Wikipedia. Worldwide, 2.1 percent of deaths are reported there to be from road traffic accidents, in 2001 at least.

As the other answers and comments suggest, this doesn't mean any one individual's probability is 0.021, but this would be your starting point for estimating the probability for a world citizen about whom you know nothing other than their existence.

-

I don't know what the actual probability is. But I am certain it's way, way less than 5%. Five percent would mean that 1 person out of 20 would end up dying this way! If you need to know the real number, there are some registries that compile the causes of death and clearly the answer is there, somewhere.

Otherwise, you can try to estimate it. One way to approach it (not very rigorous but it's an interesting exercise) is to think of how many, among all the people you know, have been killed in a car accident. Odds are there are none or very few. I know none. And as all human beings I know quite a lot of people. Now a step further: think of how many people you know (people for which you would have known if they had been killed in a car accident). This gives you a gross denominator by which you can divide your numerator (# of people who died in a crash).

Note that this is sensitive to the age of the person doing the exercise (as you get older, you have more and more chance of knowing someone who died in a car crash... Although at the same time, the number of people you know also gets higher).

Finally, if I were to risk a number, I'd say less than 0.1%. But this relies on a personal, subjective evaluation to be taken with a grain of salt.

-
This page states that there are ~40000 deaths/year due to car accidents in the US, which has a population of ~300000000. – nico Mar 21 '12 at 21:39
Interesting. So we'd have to evaluate the population who is actually using a car, say 80%. Using a life expectancy of 80 years old, we would get something around 1.3% (40000*80/(300000000*.8)). This is far more than I would have expected. – dominic999 Mar 21 '12 at 21:53
Except a number of deaths aren't random: drink or drugged driving, not wearing a seatbelt, driving on roads with faster speed limits, etc. Also, need to incorporate exposure. – Michelle Mar 21 '12 at 23:41
Absolutely. Otherwise it is just an "average risk". – dominic999 Mar 21 '12 at 23:54
How would you account, in a statistical model, for the passengers? I mean you may have 80% of people using a car, but not all of them are driving, nor they are using the same car and in the same way all along their life... And what about pedestrians run over by cars? Do they count? OK, my head hurts now :D – nico Mar 22 '12 at 7:11
show 1 more comment