Too much thought required for a Monday, but I would point out that changing the price will change people's behaviour when it comes to claims (rather than accidents). Losing your no-claims bonus is very expensive. More expensive than say replacing the bumper on a car for a 22 year old male, but might not be for a 50 year old female. Thus charging more for insurance makes insurees less likely to claim.
So that would tend to widen the gap - women are still having fewer accidents as before, but now they're *also* making fewer claims than even before, because they're paying more than before.
Basically that's the opposite of the possibility that men (somehow) cost more to insure because they're charged more. And makes a lot more sense intuitively.
I read today that the insurer Sheila's Wheels will not be abandoning their marketing and branding towards women just because they can't offer them better deals. They don't anticipate much change in their business, since they're a direct insurer (that is, they are a branch of Esure, and don't offer their services through brokers.) Presumably, they don't expect the average 18-25 year old man to opt for a pink girly insurance company, despite the price difference. I can't imagine that they're much wrong there either.
Do 18-25 year-olds use price comparison sites, and if so do those sites include Sheila's Wheels?
If all insurers keep their prices for women the same, and drop their prices for men to match, then it's all good news for drivers, and the end of a great lark for insurance underwriters.
No, as I said, Sheila's Wheels, (as part of Esure), is a direct insurer, who do not offer their products through brokers, which essentially, is all a price comparison website is, an automated broker. They seem to do this, primarily, so that you *can't* work out that Joe Q. Nobody Insurers who don't have an epic, epic marketing budget has a more appropriate policy for you, personally
( ... )
insurance is a field which is almost entirely about […] discrimination
Indeed so. I heard an interesting suggestion from theoclarke the other day: that insurers should be compelled to disregard demographic factors and charge everyone a flat premium: they could then vary that premium (within some defined parameters) as and when each person made any claims, but not otherwise. So they would be managing risk only on the basis of actual demonstrated individual exposure to it. And if the driver wanted to change insurer, they would take their claims record with them.
(I guess he would allow them to charge differential premiums for different cars, on the basis of new shiny ones being more expensive to replace than old grotty ones. But it should (I think) be possible to allow that without sneaking demographics in by the back door.)
Now obviously this is a bit hard on the poor old insurer. But in my ideal world it would be OK to say "tough!" to them.
Is that really any harder on the insurers than it is on the customers? Presumably insurers would charge a flat rate equal to whatever the average currently is for a new driver, plus a bit to account for the expected consequence that under the new system the customer base will be slightly more weighted towards costlier drivers
( ... )
Still, as _kent has said, removing the current discriminants might encourage insurers to put more effort into developing better ones, such as measuring more direct risk factors.
That is sort of my view on this. The declaration of gender on the form might be a nice short cut saving them spending time on working out what is really going on but really they could just develop better risk models and work out what the factors actually are. In theory that would be better for everyone, insurers and customers.
It might be better. It might be worse - the industry might spend millions of pounds on research, and end up with predictors of future claim costs that are only slightly better (or even slightly worse). The costs of the research will of course be passed on to customers.
I doubt that actuaries currently are lazy or stupid, so if they don't have better predictors it's probably either because there's some impediment to them using better predictors to generate quotes (e.g. brokers can't handle the complexity), or because they simply don't gather the information needed to evaluate better predictors.
So they might well come up with predictors which are more *accurate*, leading to cheaper insurance for those who "deserve" it, but with externalized costs. _kent's suggested Spy-O-Tron 3000 for example has a cost associated with the risk that Alexander the Meerkat accidentally lets Wikileaks get hold of everybody's movements for the past year.
Comments 21
Reply
Basically that's the opposite of the possibility that men (somehow) cost more to insure because they're charged more. And makes a lot more sense intuitively.
Reply
Reply
If all insurers keep their prices for women the same, and drop their prices for men to match, then it's all good news for drivers, and the end of a great lark for insurance underwriters.
Reply
Reply
(The comment has been removed)
Indeed so. I heard an interesting suggestion from theoclarke the other day: that insurers should be compelled to disregard demographic factors and charge everyone a flat premium: they could then vary that premium (within some defined parameters) as and when each person made any claims, but not otherwise. So they would be managing risk only on the basis of actual demonstrated individual exposure to it. And if the driver wanted to change insurer, they would take their claims record with them.
(I guess he would allow them to charge differential premiums for different cars, on the basis of new shiny ones being more expensive to replace than old grotty ones. But it should (I think) be possible to allow that without sneaking demographics in by the back door.)
Now obviously this is a bit hard on the poor old insurer. But in my ideal world it would be OK to say "tough!" to them.
Reply
Reply
That is sort of my view on this. The declaration of gender on the form might be a nice short cut saving them spending time on working out what is really going on but really they could just develop better risk models and work out what the factors actually are. In theory that would be better for everyone, insurers and customers.
Reply
I doubt that actuaries currently are lazy or stupid, so if they don't have better predictors it's probably either because there's some impediment to them using better predictors to generate quotes (e.g. brokers can't handle the complexity), or because they simply don't gather the information needed to evaluate better predictors.
So they might well come up with predictors which are more *accurate*, leading to cheaper insurance for those who "deserve" it, but with externalized costs. _kent's suggested Spy-O-Tron 3000 for example has a cost associated with the risk that Alexander the Meerkat accidentally lets Wikileaks get hold of everybody's movements for the past year.
Reply
Leave a comment