An NMA member’s insurance rates were about to go up because of a low speed rear-ender on slush-covered ice, the same as if she spun off the road at 100 mph. It seems unfair, but it’s a natural consequence of predicting the future based on rare events.
The system assumes that for every time you get a ticket, there were 99,999 times you disobeyed a sign and didn’t get caught. For every accident, you had a hundred near misses.
Ever heard of margin of error in a poll? If you ask 1,000 people you can tell a 55-45 split from a 45-55 split, but not 51-49 from 49-51.
If you grab one guy walking down the street and ask which megalomaniac he is going to vote for in November, you know nothing. The margin of error is basically ±100%.
Predictions based on a few events are unreliable. One speeding ticket, or one accident, is that one man poll.
But if all you know is Fred was hit by a meteorite, and Sally was not, you have to assume he is a rock magnet and raise his insurance rates.
And if you say that excusable icy slip and slide doesn’t count as an accident, you’re left with even less data to work with.
So what’s the solution? Collect more data.
Insurance companies want to track your driving. Mostly not your speed; they know that doesn’t matter. Each time you brake hard they say you had a close call because you weren’t paying attention. That gives them a bunch of those 100 near misses to work with.
And here’s another problem with using anything but actual accident history.
Some drivers who brake hard weren’t paying attention. Others brake hard by design, when it is safe. How does a spy in the dashboard tell them apart? It may not need to.
Insurance cares about correlation, not cause and effect.
They want to charge you extra if you have bad credit, for example, even though that doesn’t have any effect on your driving. A college classmate anticipated a rate reduction after he got married because married men under 25 got the same good car insurance rates as women. Actuaries figured out that the people who do these things also drive differently, not that doing these things makes people drive differently.
Same with cell phones. We used to assume that cell phones turned people into bad drivers. A more recent study found that bad drivers tend to use cell phones, and if you take away their cell phones they still drive badly.
An Australian study sometimes cited for the proposition that speed is dangerous suffers from a similar problem. It found drivers speeding in one place tended to have accidents in other places. That is, bad drivers tend to speed.
Insurance companies decided that hard braking is bad, when I doubt the evidence supports a stronger statement than “bad drivers tend to brake hard.”
I said speeding is not generally considered an indicator of risk by insurance companies. Yet insurance companies (through their lobbying group, the IIHS) want speed limits lowered. Speeding tickets are an indicator of risk, just like bad credit or a Y chromosome.
Insurance companies want everybody to be speeding. More specificially, they want police to be able to ticket anybody. Not everybody, anybody. The goal is for a few cases of “you’re driving like a jerk but I can’t prove reckless driving” in a 70 zone to become 65-in-a-55 tickets in a 55 zone.
Make everybody a criminal and hope police can pick out the dangerous lawbreakers from the ordinary lawbreakers.
Or as a Crusader was reported to have said of a sacked city, “kill them all and let God sort them out.”
God — or somebody — took care of the NMA member I mentioned. The prosecution failed to show at the insurance surcharge appeal hearing. Not at fault.
The opinions expressed in this post belong to the author and do not necessarily represent those of the National Motorists Association or the NMA Foundation. This content is for informational purposes and is not intended as legal advice. No representations are made regarding the accuracy of this post or the included links.
