Hacker News new | ask | show | jobs
by VLM 4713 days ago
A realistic fact based estimate is around one hundred, around 1 or 2 orders of magnitude lower.

http://www-nrd.nhtsa.dot.gov/Pubs/811753.pdf

Around page 6, a little under 10K fatalities in 2011 somehow involved speeding. (about 1/3)

Around page 7, 87% of fatalities involving speeding were not on interstate highways. So about 10% of about 10K is about 1K people died from driving "too fast" on the highway.

Around page 7, around half of drivers who killed someone while speeding were drunk. so about 500 people were killed by sober drivers who were going over the current speed limit. Drunks are going to kill someone anyway, no matter what the law is. You can consider this a fixed constant death rate regardless of speed laws.

In other words, if you raised the speed limit such that no one ever got cited for speeding during a fatal crash, aka we go total full motion autobahn, about 500 people would die legally instead of illegally.

Of course a lot of this involves police high speed chases, driving too fast for conditions, suicide by car, road rage, mental illness, stuff like that. I feel confident that out of a total sample of about 500, a realistic change in speed limit up to perhaps 75 would probably kill at most 100 people. I think this is a overestimate, but feel confident it would not actually turn out to be 10x higher.

This analysis pays no attention to the death toll from increased car exhaust fumes due to reduced mileage, or increased soldier death rates in the middle east due to needing to burn more oil, or any of the other secondary effects, which may very well be greater than the primary death toll, I really can't estimate those very well. For example, if you raised the death toll in Iraq by 20% on both sides to get 20% more oil, that would swamp the relatively minor increase in death toll on the interstates back home.

This helps you identify political axes to grind... someone claiming "10" or "zero" is obviously distorting reality in one direction, and "100s" is obviously equally distorted in the other.

Frankly I wouldn't be concerned. The odds of death by lightning are about 1 or 2 orders of magnitude higher. The odds of death by poor diet are around 5 orders higher. Death by poor exercise habits, probably the same around 5 orders higher. Higher highway speeds are not a very serious threat in the big picture of risks. I'd worry a lot more about "not talking a walk after dinner" or "drinking too much corn syrup"

1 comments

To be clear, I'm not in favor of altering speed limits. Just pointing out how little sense it makes to turn our whole society upside down for terrorism deaths, when we clearly don't do that for non-terrorism deaths.

Thanks for the details though. :)

I wasn't attacking you, I just wanted some sources because driver education is really the only solution to this :D