3 ms·
Before we venture into whether it's good or bad, here's a couple of observations. 1) The point of insuring is to reduce tail risk. If you have a 1/1000 years c
by lordnacho 11y ago
Before we venture into whether it's good or bad, here's a couple of observations.
1) The point of insuring is to reduce tail risk. If you have a 1/1000 years chance of your house getting flattened in an earthquake, you want to spread that bill over 1000 years. If you don't you might get away with it most years, but you'll have a new house to buy if you're unlucky.
To contrast, if your house is flooded every other year, someone insuring you will ask you for roughly half the likely damages each year. That's not smoothing it out by much, and you might as well save up for the inevitable.
2) If you're going to smooth out risk, you need to know what the risk is. For that, you'll need information. For instance if women are less likely than men to get themselves hurt while driving sports cars, you will want to know whether someone is a man or a woman. If old people are more likely to get cancer, you want to know how old someone is.
3) If you do not heed the above observation, there's something called adverse selection that will get you. Suppose you have a bunch of high risk people mixed with low risk people. You'll calculate a premium that's somewhere in between the two pools. The problem is the low risk people will look at it and conclude it's expensive for them, and they'll be less likely than you'd realised to sign up. So you then get a bunch of high risk people who think it's a great deal signing up, getting killed in accidents, and costing you more than you thought likely. It also means if you as a low risk person got into an accident anyway, you'd be without insurance.
So is it good that insurance companies want more and more info? For the most part, yes. It lets them price the premium more precisely, which helps everyone get insured. We'll be able to pay for unlikely events at a cost that reflects the likely average. And you wouldn't be paying for people who are more risky, and you wouldn't be receiving anything from people in lower risk groups.
One thing to look out for is how we treat expensive-to-insure people. Even if there's going to be a more accurate price for them, they may be very small groups whom the insurance companies decide aren't worth the effort. For instance if we know that 80% of people with the BRCA2 gene will get cancer by age 50, you're going to calculate a huge cost for them. You'd be collecting money from 5 people to hand it out to 4 of them, plus the expense of handling a claim.
One interesting thing about this new information fountain is you can lose money by having tests done. Say you have a parent with BRCA2, and people who aren't tested have a 40% chance. WTF do you do then? You will either pay a huge premium or barely anything.