[sublupo]

If that is what you want then that is okay. You can pay for that type of insurance. I prefer to buy life insurance from a place that will charge me a lower rate because I don't drink or smoke. I assume that most people who don't drink or smoke would join my type of insurance as well, leaving your type of insurance full of drinkers and smokers. That would make your insurance cost more (the insurance is paying more) while mine would be cheap.

Now, if you are talking about requiring everyone to use the same insurance for the same price (maybe a government insurance) then I would want to ban drinking and smoking. Because now their actions directly impact my by rasing my insurance cost.

[mikeash]

You can't have both types of insurance. If you try, the one that discriminates less just becomes the highest-risk pool. So telling someone to choose the one they prefer doesn't make any sense.

[sublupo]

That's my point. How would you like it if you had to pay just as much in flood insurance as someone with a beach front property in Florida? If I had to, I would try to make beachfront properties in Florida illegal.

[mikeash]

The other extreme is a scenario where insurers magically have perfect knowledge about your individual risk and insurance becomes completely pointless.

You will not receive exactly what you pay. Some people will receive much more, and their payouts are subsidized by the others who receive much less. That is the nature of insurance.

What's the difference between your question and "How would you like it if you had to pay just as much flood insurance as someone down the street whose house actually gets flooded next year?" The only difference is that you can't reliably predict things with that level of accuracy.

Insurance companies and low-risk insurance buyers are both incentivized to predict individual risk profiles as accurately as possible, but this gradually drives the system towards uselessness.

[SamReidHughes]

The only way insurance becomes pointless is under absolute certainty of future events. If you merely have personalized calculated risk using all the information available, insurance still makes sense. For example, you have total information about yourself, and there's a minimum price you'd be willing to sell car insurance to a copy of yourself, and it makes sense for you to buy it, too.

[sjjshzvuiajhz]

One could argue that the social value of insurance is to incentivize you to not take risks we know about, while protecting you from risks we don’t know about.

[c1sc0]

This assumes that conformity is the best way to increase value for society which I would argue is wrong.

[sublupo]

The more information they have the better it is for me. The reason that I use insurance is to lower the variance significantly while only slightly lowering the expected value. Now, if they had an oracle that could see the future, then there would be no point in using them. But that is not the case now, nor do I expect it to be the case any time soon.

Currently I want the insurance company to know what my expected value is properly so that I can get a rate that is appropriate for my needs. Of course I would prefer that they make a mistake and assume that my expected value is lower than reality, but if they make to many mistakes like that then either they will go out of business or raise their premiums.