GM Declares Bankruptcy

[KrazyHorse]

Dan, it's pretty obvious that you didn't understand. Insurance makes people pay for the accidents THEY CAUSE (causation in a liability sense). Pigovian pricing is completely on TOP of this and makes people pay for the additional accidents (which they are not liable for) which their presence on the road leads to.

[Ben Kenobi]

As an extreme example, imagine a driver who never makes any mistakes. His insurance costs nothing because he will never cause an accident. On the other hand, there's a non-zero chance that he will be involved in an accident with somebody else!

Wouldn't that be a function of driver density?

I'd go with that. In larger cities you are more likely to get in an accident just by virtue of driving because there is more traffic.

By your reckoning, insurance costs should be higher the more people you live near. You wouldn't have to raise it by very much either, but it would work out to be a discount living in less congested areas.

[KrazyHorse]

THIS is the 2000$ figure that I'm quoting, by the way. When I choose to drive in California I cost OTHER DRIVERS 2000$ per annum by increasing their insurance costs. These increased insurance costs are an EXTERNALITY. I don't care about them because I DON'T PAY THEM.

The choice I make is between driving and not driving. When I drive I impose (say) 1500$ in insurance costs on myself and 2000$ in insurance costs (in aggregate) on other people. In other words, if I choose not to drive for a year then society as a whole saves 3500$ in damaged vehicles, injuries etc, but I only save 1500$. The difference between the two is the externality.

[Kuciwalker]

KH: Assuming that the average compensation paid for each accident is roughly correct*, then in aggregate we must pay the correct amount for accidents. Since that's all funded by insurance, in aggregate our insurance rates must be sufficient to cover the externality, even if our individual insurance rates aren't correct.

Essentially this should result as a subsidy for good drivers...

*not a great assumption, but whatever

[Kuciwalker]

THIS is the 2000$ figure that I'm quoting, by the way. When I choose to drive in California I cost OTHER DRIVERS 2000$ per annum by increasing their insurance costs. These increased insurance costs are an EXTERNALITY. I don't care about them because I DON'T PAY THEM.

But they each have the same effect on you, so it should even out.

[DanS]

Dan, it's pretty obvious that you didn't understand. Insurance makes people pay for the accidents THEY CAUSE (causation in a liability sense). Pigovian pricing is completely on TOP of this and makes people pay for the additional accidents (which they are not liable for) which their presence on the road leads to.

Maybe I'm not understanding what you're saying, but I think I do. The additional accidents that you cause based on your presence is picked up by somebody else's insurance carrier. This assumes that people carry all types of coverage, which admittedly, some don't. Also, this assumes that people are covered unlimited, which admittedly, is rare.

[rah]

Car insurance is traditionally higher in higher population density areas. It is cheaper in less congested areas. (on average)

[Ben Kenobi]

It's a hidden cost.

Right now, drivers who live in low density areas are paying the same as those who live in high density areas.

I'd suspect that the function would go up with the square root of population density.

[Jon Miller]

It's a hidden cost.

Right now, drivers who live in low density areas are paying the same as those who live in high density areas.

I'd suspect that the function would go up with the square root of population density.

Actually, where you live and how much you drive all go into determining your insurance cost.

Admittedly, they should probably play a larger roll than they do.

JM

[rah]

No if you live in a less dense area, the odds are you pay less. (unless there are other mitigating risks, like a blighted area in a city)


It's simple logic. More traffic density, the more accidents, the higher the risk of accident, the higher cost of car insurance premiums because of the risk. I doubt it's any different in Canada.

It's been that way in the US since the first actuary figured it out.

Gee we found another topic that BK knows nothing about. What a suprise.

[Ben Kenobi]

hmm, squart root isn't a high enough function. I get 20x the accidents in say NYC as in where I live now.

I could see twice as likely, but no more then that.

x^0.1 gives me about the right function. You'd be about 6x as likely to get in an accident if you lived in New York city, then if you lived in a town of 1 person, assuming all else was equal. In a town of 70k you would be about 3x as likely, and about 6x as likely in New York city.

Those are pretty big numbers! I don't think folks in New York pay 2x what I do!

[Ben Kenobi]

I doubt it's any different in Canada.

We have socialised insurance here.

Everyone pays the same unless they have a 10 year accident free history as I do. I get about 1/4th what I would normally pay. Young men under 25 have huge premiums. I pay about 500 a year in insurance, I don't know what your rates are down in the US.

[rah]

The cost is based on multiple factors you idiot. Accident probability is just one of them.

[Imran Siddiqui]

I like the non-moderator rah very much

[KrazyHorse]

KH: Assuming that the average compensation paid for each accident is roughly correct*, then in aggregate we must pay the correct amount for accidents

Kuci, it's pretty obvious that you haven't thought very hard about Pigovian taxation.

Think about it this way: ignore for now differential drivers. All drivers are equally bad, all drive the same amount at similar times in similar places etc. The only choice people have is whether or not to drive a car (if they don't drive they take the subway). Call the number of drivers d. The number of accidents per year ("a") is some function of d which increases faster than linearly with d. For simplicity's sake say

a = (10^-6)d^2

and say that each accident costs 1000$ total in damages and injuries. The total cost of accidents per year is (0.001)d^2 $

Since drivers are indistinguishable they each get charged 0.001d $ per year in insurance. Fine, you say. All damages are paid for. This is NOT THE POINT. The point is that the MARGINAL DRIVER (the one to whom it was barely worthwhile to drive when insurance cost him 0.001d $ per year) is NOT making the efficient decision when driving. The TOTAL COST imposed by him driving is the difference between him driving and him not driving!

(0.001)*(d^2 - (d-1)^2) $ per year which is approximately 0.002 d $ per year

Since he experiences a benefit of only slightly greater than 0.001 d $ per year by driving (once all other costs are figured in) but this costs society as a whole 0.002 d $ per year then he is making an inefficient decision!

With the simplifying assumption that all drivers are identical then insurance charges the MEAN cost per driver while efficient pricing demands that drivers be charged the MARGINAL cost for an additional driver. The difference between the two is the externality. With heterogenous drivers it's not as easy to put into a nutshell as that; in that case, you can describe the pigovian pricing as the additional accidents which other drivers are at fault for which my driving causes.