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.
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.
I think actually he was envisaging that they would be obliged to charge initially only what was appropriate for a 'safe driver', and then they could only ramp it up for an individual if they started making claims and so were proven non-safe.
So in comparison with today's system, safe drivers will receive recognition of that form day one. Dangerous drivers will experience an initial period of tolerance but then be hit heavily as they reveal their true nature. If this ends up forcing them off the road as the costs rise, then that might actually be a bonus. (Of course, they might prefer to drive uninsured, which would be bad. But I suspect in the Orwellian nightmare Theo proposes here, there would be ways of preventing that :-)
As well as the value of the car insured, there other
( ... )
obliged to charge initially only what was appropriate for a 'safe driver'
OK, so that would sort of fall under the same heading as the uninsured drivers fund: "we, the government, mandate that drivers must buy insurance. You, a private company, may only sell that product if you agree to pay for that privilege as follows...".
Currently the payment is into the fund, theoclarke would in addition mandate a benefit to drivers of unknown quality, that they should be treated the same as drivers of known good quality. (where "quality" means "the property of not making claims").
Sounds like a lot of fuss, presumably with the end that people can feel they're "innocent until proven guilty" in a realm unrelated to criminal law. What's next, Tesco has to treat you as though you have lots of clubcard points, until you've proven you aren't loyal by failing to accumulate more
( ... )
Sounds like a lot of fuss, presumably with the end that people can feel they're "innocent until proven guilty" in a realm unrelated to criminal law.
He felt that that sort of general emotional buy-in might be important to acceptance of the scheme. But yes, this is a slightly doubtful aspect of the proposal.
It also massively increases the number of drivers with an interest in concealing their claim history.
Mm, another fruitful area for the Orwellian nightmare to operate.
Insurance companies surely would respond by making penalties for making a claim permanent.
Yes.
The trick would be to ensure that as many of their customers as possible are off the state-mandated charging scheme
Probably, although I think that he was proposing pretty stringent state circumscription of charges even for those.
If you're going to sit on insurance companies' shoulders telling them exactly what they can charge, will there be any benefit at all from competition among providers, or will they all just charge the maximum, all the time? If private companies make a profit selling insurance at a flat rate, doesn't that just make insurance beyond third-party less worthwhile for more people? Insurers naturally can't afford to give as big no-claims bonuses to the cheaper drivers, if they have to give the same bonuses to all new drivers too.
The more the government regulates prices, the more that the difference between the legal maximum and what it actually costs to provide the insurance, looks like a fee paid to the private companies to do the admin. The government might as well roll insurance into road tax, and go direct to the underwriting market if it feels the need to lay off the risk (which I suspect it shouldn't need to).
Actually, I neglect the other important thing that insurance companies do apart from set prices and sell the products, which is contest claims in court. There might still be a point in letting private companies do that, and try to maximise their profits by minimising claims paid rather than by having any control over revenues.
Mm, I would be happy for it to be a govt project, I can't see any free market benefit really.
(But I pretty much feel that about any market which in practice has to be heavily regulated to prevent the participants screwing customers/the nation over.)
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
I think actually he was envisaging that they would be obliged to charge initially only what was appropriate for a 'safe driver', and then they could only ramp it up for an individual if they started making claims and so were proven non-safe.
So in comparison with today's system, safe drivers will receive recognition of that form day one. Dangerous drivers will experience an initial period of tolerance but then be hit heavily as they reveal their true nature. If this ends up forcing them off the road as the costs rise, then that might actually be a bonus. (Of course, they might prefer to drive uninsured, which would be bad. But I suspect in the Orwellian nightmare Theo proposes here, there would be ways of preventing that :-)
As well as the value of the car insured, there other ( ... )
Reply
OK, so that would sort of fall under the same heading as the uninsured drivers fund: "we, the government, mandate that drivers must buy insurance. You, a private company, may only sell that product if you agree to pay for that privilege as follows...".
Currently the payment is into the fund, theoclarke would in addition mandate a benefit to drivers of unknown quality, that they should be treated the same as drivers of known good quality. (where "quality" means "the property of not making claims").
Sounds like a lot of fuss, presumably with the end that people can feel they're "innocent until proven guilty" in a realm unrelated to criminal law. What's next, Tesco has to treat you as though you have lots of clubcard points, until you've proven you aren't loyal by failing to accumulate more ( ... )
Reply
He felt that that sort of general emotional buy-in might be important to acceptance of the scheme. But yes, this is a slightly doubtful aspect of the proposal.
It also massively increases the number of drivers with an interest in concealing their claim history.
Mm, another fruitful area for the Orwellian nightmare to operate.
Insurance companies surely would respond by making penalties for making a claim permanent.
Yes.
The trick would be to ensure that as many of their customers as possible are off the state-mandated charging scheme
Probably, although I think that he was proposing pretty stringent state circumscription of charges even for those.
Reply
If you're going to sit on insurance companies' shoulders telling them exactly what they can charge, will there be any benefit at all from competition among providers, or will they all just charge the maximum, all the time? If private companies make a profit selling insurance at a flat rate, doesn't that just make insurance beyond third-party less worthwhile for more people? Insurers naturally can't afford to give as big no-claims bonuses to the cheaper drivers, if they have to give the same bonuses to all new drivers too.
The more the government regulates prices, the more that the difference between the legal maximum and what it actually costs to provide the insurance, looks like a fee paid to the private companies to do the admin. The government might as well roll insurance into road tax, and go direct to the underwriting market if it feels the need to lay off the risk (which I suspect it shouldn't need to).
Reply
Reply
Reply
(But I pretty much feel that about any market which in practice has to be heavily regulated to prevent the participants screwing customers/the nation over.)
Reply
Leave a comment