Car insurance

[Jon Corby]

The European Court of Justice has ruled that insurers can no longer use gender in premium calculations.

Is this a good thing? Is it forcing equality where equality doesn't really exist or make sense? I'm undecided - seduce me with your opinions.

[Michael Wallace]

It's an odd one. I've wondered for years why it was apparently accepted without question, when it seemed a bit weird, but now the question is why can they still discriminate on age? Seems a bit odd.

[Jon Corby]

It's an odd one. I've wondered for years why it was apparently accepted without question, when it seemed a bit weird, but now the question is why can they still discriminate on age? Seems a bit odd.

Yeah, I'm in two minds. It seems that if you want to force gender equality in other areas (e.g. employment), then by extension it has to be applied to everything, even where you can actually see the sense in it.

As a stats-based person, do you think they're dismissing relevant or irrelevant statistics?

[Jon O'Neill]

I work in Insurance Statistics and it's pretty retarded that you're not allowed to discriminate based on gender but you are on stuff like age.

[Michael Wallace]

As a stats-based person, do you think they're dismissing relevant or irrelevant statistics?

I think the trouble is that from an individual's perspective, it's not fair - simply being a man doesn't make you more likely to have a crash in the same way that it might make you more at risk of a certain disease. But it seems inconsistent (as Jono just said) to be allowed to discriminate on age, but not on gender.

[Jon Corby]

As a stats-based person, do you think they're dismissing relevant or irrelevant statistics?

I think the trouble is that from an individual's perspective, it's not fair - simply being a man doesn't make you more likely to have a crash in the same way that it might make you more at risk of a certain disease. But it seems inconsistent (as Jono just said) to be allowed to discriminate on age, but not on gender.

I agree, but are you not then tending towards a situation where either the cost is spread equally across everybody (i.e. completely indiscriminate), or you do away with insurance altogether (where the cost is accurately paid by specifically those individuals who incur it)? How can you have any "fair" system that lies between these two extremes if it isn't discriminatory in some way?

[Michael Wallace]

I agree, but are you not then tending towards a situation where either the cost is spread equally across everybody (i.e. completely indiscriminate), or you do away with insurance altogether (where the cost is accurately paid by specifically those individuals who incur it)? How can you have any "fair" system that lies between these two extremes if it isn't discriminatory in some way?

I think it would be better if everyone starts off with the same insurance costs, and then they change based on your driving. If you have crashes it goes up, if you don't, it goes down. I don't really know anything about driving, or car insurance, so despite stats-powers I'm probably not particularly well informed on this one, but I'm not how it's unfair to treat everyone identically until they demonstrate bad driving. (As I said, it's not like medical insurance where simply being a particular gender/age/whatever will carry a risk you can't do anything about.)

Ideal world ideas would inclue making you go to a centre that forces you to do reaction tests would be good, or make everyone retake their driving test vaguely regularly (you could even grade people, so if you're super good at your driving test you get a lower rate of insurance). (I accept these are pretty impractical ideas.)

[Jon Corby]

but I'm not how it's unfair to treat everyone identically until they demonstrate bad driving.

I didn't say that was unfair, that's pretty much option b) - you're covering your own costs (though partly in arrears) with a bit of option a) thrown in to split the costs already incurred just in case you're priced out of driving (or at least with insurance) altogether.

[Michael Wallace]

but I'm not how it's unfair to treat everyone identically until they demonstrate bad driving.

I didn't say that was unfair, that's pretty much option b) - you're covering your own costs (though partly in arrears) with a bit of option a) thrown in to split the costs already incurred just in case you're priced out of driving (or at least with insurance) altogether.

O rite, I misread, soz.

[Jon O'Neill]

Hopefully this means no more Shiela's Wheels adverts.

[Ian Fitzpatrick]

Hopefully this means no more Shiela's Wheels adverts.

The ONLY benefit to this crackpot idea!

[Charlie Reams]

Ideal world ideas would inclue making you go to a centre that forces you to do reaction tests would be good, or make everyone retake their driving test vaguely regularly (you could even grade people, so if you're super good at your driving test you get a lower rate of insurance). (I accept these are pretty impractical ideas.)

I think if you take Pass Plus then your premiums go down, so we already have a limited version of that system. And as a general defender of meritocracy I think that would be the fairest system. Score people in their driving test, everyone has to meet a minimum standard and beyond that, the better you do, the lower your premiums. Then the existing no-claims-bonus system reinforces the benefits of good driving. This would also encourage people to practise a bit more before taking their test, which sounds beneficial to me.

Within the current system though, there is a bit of a difference between age and gender -- with age, if you're annoyed by high premiums now, at least you'll get low premiums later; but you're stuck with the same gender permanently. Not sure how relevant that really is, but it's an argument I can see someone making. Nevertheless I wouldn't be surprised if the ECJ squash age discrimination too.

[Jon Corby]

but you're stuck with the same gender permanently.

O RLY?

[Jon O'Neill]

Driving tests in general are total gash and have no bearing on how likely you are to have an accident. For example, I passed my driving test with 1 minor fault, and I've had two crashes in under a year. One was completely my fault for driving too aggressively, which I obviously didn't do in the test or I would've failed. The other was completely the fault of this doddering old 70+ who want through a give way without giving way, which made my right wing give way. He probably passed his test in the fifties when all you had to do was turn the engine on and wink at the tester. My mum passed her test about twenty years ago at the thirteenth (13th) attempt, and she's never had so much as a scratch.

I suppose what I'm getting at is that it would be difficult to make a test that gave an indication of how likely you were to make a claim on your insurance.

[Jon Corby]

I suppose what I'm getting at is that it would be difficult to make a test that gave an indication of how likely you were to make a claim on your insurance.

I agree, and just adjusting premiums after the event (which of course, already happens) won't cover everything because the worst offenders will just be priced completely off the roads. So who picks up the tab then, if it's not everybody else? How can you spread the cost in any way other than this, that isn't discriminatory?

[James Robinson]

My mum passed her test about twenty years ago at the thirteenth (13th) attempt, and she's never had so much as a scratch.

At last, someone who has taken more goes than me to pass my driving test 8 goes, if anyone is interested, or cares.

[Michael Wallace]

Not directly relevant to car insurance, but that's something else I've found odd. Is it really a good sign if someone passes their driving test on the 20th attempt? What do people think of a system where you can only get a licence if you've passed more times than you've failed. Would make it a lot harder for someone to just get a bit lucky.

(That said, from what I know of driving tests, they don't sound particularly good to begin with.)

[Jon O'Neill]

Not directly relevant to car insurance, but that's something else I've found odd. Is it really a good sign if someone passes their driving test on the 20th attempt? What do people think of a system where you can only get a licence if you've passed more times than you've failed. Would make it a lot harder for someone to just get a bit lucky.

(That said, from what I know of driving tests, they don't sound particularly good to begin with.)

They're SO subjective, to the tester and to the roads around the test centre and to how retarded everyone else you come across decides to be that hour.

[Jon O'Neill]

I suppose what I'm getting at is that it would be difficult to make a test that gave an indication of how likely you were to make a claim on your insurance.

I agree, and just adjusting premiums after the event (which of course, already happens) won't cover everything because the worst offenders will just be priced completely off the roads. So who picks up the tab then, if it's not everybody else? How can you spread the cost in any way other than this, that isn't discriminatory?

We use statistics to determine the price based on certain risk factors!

[Michael Wallace]

Not directly relevant to car insurance, but that's something else I've found odd. Is it really a good sign if someone passes their driving test on the 20th attempt? What do people think of a system where you can only get a licence if you've passed more times than you've failed. Would make it a lot harder for someone to just get a bit lucky.

(That said, from what I know of driving tests, they don't sound particularly good to begin with.)

They're SO subjective, to the tester and to the roads around the test centre and to how retarded everyone else you come across decides to be that hour.

Do you think they have to be, though? How hard would it be to design a simulator that could throw 'random' events at you, and couple that with tests of technical ability. Like I said, I really don't have a clue about driving, so I'd be interested to know if there are any good reasons for the way tests are done as they are now.

[Jon Corby]

We use statistics to determine the price based on certain risk factors!

Yeah, so do you feel, with this ruling, that you're now not allowed to use a relevant statistic?

I can see the argument that, as a man, why should I penalised if some other men are idiots behind the wheel. But it doesn't it then apply to any of your other measures that aren't based on the individual (e.g. convictions, claims)? Why is it not fair to be lumped in with all men, yet fair to be lumped in with all the people that drive the same car as you?

[Jon O'Neill]

We use statistics to determine the price based on certain risk factors!

Yeah, so do you feel, with this ruling, that you're now not allowed to use a relevant statistic?

Yes.

They're SO subjective, to the tester and to the roads around the test centre and to how retarded everyone else you come across decides to be that hour.

Do you think they have to be, though? How hard would it be to design a simulator that could throw 'random' events at you, and couple that with tests of technical ability. Like I said, I really don't have a clue about driving, so I'd be interested to know if there are any good reasons for the way tests are done as they are now.

Seems like a simulator would be the fairest thing. Might not prepare you for actually having to do things properly or dying though.

[Mark James]

Just get robots to drive the cars and put sensors on the cars so they can detect where anything they could crash into is so that they don't. If they can make programmable vacuum cleaners that go around the living room on their own without bumping into the furniture then they can do it with cars too surely.

[John Bosley]

I suppose it's only unfair the first time you take out insurance because, as someone said, it then depends on your claims record. It would probably go up immediately after you made your first claim and if you had a major write off it could go up a huge amount.....or have I missed the point?

[David Williams]

Men are more likely to make motor insurance claims than women. Fact. Men's claims are on average higher than women's. Fact. So, other things being equal, a man's insurance premium should be higher than a woman's? Not necessarily.

For example, it may be that men drive more miles than women, and that's the reason they have more accidents. It may be that a greater proportion of their mileage is on faster roads, where an accident will cause more damage. If so, women who drive a lot are having their premiums subsidised by men who only drive a little.

[JimBentley]

I thought it was a really bizarre decision when I heard about it. As I understand it, people are put into a certain actuarial class depending on their sex, age, marital status, place of residence, etc. and the insurance premium is calculated based on statistics of previous claims within these groups. Men make more claims than women, so premiums are higher for men. This seems fair enough to me, just as it's fair that someone living in a postcode with a high auto crime rate has a higher premium than someone living in the middle of nowhere with a negligable crime rate.

However, this ruling also affects life insurance premiums, which makes things very interesting. Women live on average about five years longer than men, so their life insurance premiums have always been lower. But now they must be equalised, which is completely absurd. A man of 65 taking out life insurance would on average be paying premiums for another 13 years, but a woman would on average be paying for another 18 years, so when the ruling takes effect next year, women will effectively be being discriminated against as they'll have to pay more in total for the same overall benefit.

All that's going to happen is that the insurance companies will look to cover their backs by increasing women's premiums to the same (higher) level as men's, without passing on any benefit to the men.

[Mark James]

Yeah, but women tennis players now get the same amount of prize money as men even though they play less tennis.

[Michael Wallace]

Yeah, but women tennis players now get the same amount of prize money as men even though they play less tennis.

Which means they have more time to drive their cars and CRASH

[Paul Howe]

I thought it was a really bizarre decision when I heard about it. As I understand it, people are put into a certain actuarial class depending on their sex, age, marital status, place of residence, etc. and the insurance premium is calculated based on statistics of previous claims within these groups. Men make more claims than women, so premiums are higher for men. This seems fair enough to me, just as it's fair that someone living in a postcode with a high auto crime rate has a higher premium than someone living in the middle of nowhere with a negligable crime rate.

However, this ruling also affects life insurance premiums, which makes things very interesting. Women live on average about five years longer than men, so their life insurance premiums have always been lower. But now they must be equalised, which is completely absurd. A man of 65 taking out life insurance would on average be paying premiums for another 13 years, but a woman would on average be paying for another 18 years, so when the ruling takes effect next year, women will effectively be being discriminated against as they'll have to pay more in total for the same overall benefit.

All that's going to happen is that the insurance companies will look to cover their backs by increasing women's premiums to the same (higher) level as men's, without passing on any benefit to the men.

On the other hand women currently receive lower annuity rates than men due to their longer expectation of life, this will be equalised under this ruling. For most people this is likely to be a far bigger item of expenditure than pure life insurance.

[Rhys Benjamin]

Well.... I feel like being the go compare MAN.

At last, someone who has taken more goes than me to pass my driving test 8 goes, if anyone is interested, or cares.

What about Lee Graham?

[David Williams]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

[Jon Corby]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

Yeah, this is the other area I thought we'd get into in this discussion. I really am torn as to whether it should be okay, if there really genuinely is a strong correlation between such attributes and the amount that they cost the insurance company, surely it's reasonable for them to include it in their calculations? Isn't there enough data that it's statistically sound? Personally, I don't see why race or sexual orientation would make any difference, but if it is shown to, then why not? It seems to be an inescapable fact that year on year young males factor disproportionately in motor insurance payouts.

(I guess part of the issue might be that it's kind of tricky to measure somebody's race or sexuality in the same way it is gender. I'd probably camp it up on the phone to Churchill if it made my premium cheaper.)

[Jon O'Neill]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

Yep.

[Michael Wallace]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

Yep.

Except race and sexual orientation aren't as well-defined?

Edit: Oh ok, Corby already said that. Stupid words, expecting me to read them. OOI, what happens if you tell an insurance company you self-identify as a woman when your name is David and you have a beard?

[Jon O'Neill]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

Yep.

Except race and sexual orientation aren't as well-defined?

Edit: Oh ok, Corby already said that. Stupid words, expecting me to read them. OOI, what happens if you tell an insurance company you self-identify as a woman when your name is David and you have a beard?

Good question. And how much does Caster Semenya pay?

[Michael Wallace]

I suppose it's a case of what you gender is legally, so this 'self-identify' business doesn't come into it. Does kinda highlight the problem with race or homo-ocity, though. I suppose for the former you could use paint swatches or something?

[David Williams]

If it's OK to use gender as a rating factor, presumably it's OK to use race, or sexual orientation?

Yep.

Except race and sexual orientation aren't as well-defined?

The fact that they are less well-defined is a matter of practicality, rather than principle. Insurers use plenty of ill-defined information. Is your car garaged overnight? What, every single night? Who is the main driver? The one who goes to the shops every day, or the one who drives to Penzance once a month? Or the one who'll attract the lowest premium. How many miles a year do you drive? Well, I know roughly how many I did last year, but I really haven't decided how many I'll do this year.

[Michael Wallace]

The fact that they are less well-defined is a matter of practicality, rather than principle. Insurers use plenty of ill-defined information. Is your car garaged overnight? What, every single night? Who is the main driver? The one who goes to the shops every day, or the one who drives to Penzance once a month? Or the one who'll attract the lowest premium. How many miles a year do you drive? Well, I know roughly how many I did last year, but I really haven't decided how many I'll do this year.

The only one of those that seems potentially ill-defined is main driver, which I don't know about (do policies not at least specify what that means to them?). Regardless, I would argue the point is about whether you have control over it. Do insurers use any ill-defined factors that one has no control over?

[Jon Corby]

Except race and sexual orientation aren't as well-defined?

The fact that they are less well-defined is a matter of practicality, rather than principle. Insurers use plenty of ill-defined information. Is your car garaged overnight? What, every single night? Who is the main driver? The one who goes to the shops every day, or the one who drives to Penzance once a month? Or the one who'll attract the lowest premium. How many miles a year do you drive? Well, I know roughly how many I did last year, but I really haven't decided how many I'll do this year.

True, but I can see how it's a sufficient barrier to using such measurements, considering how controversial it would be. I genuinely can't decide what I make of all this. Should sexual(/any) equality be forced through in cases where it clearly doesn't exist?

[David Williams]

Suppose it was considered discriminatory to use information about health when pricing life insurance. Terminally ill people would take out massive policies, the price of life insurance would rise, such that fewer healthy people would consider it value for money, and a downward spiral would result. Life insurance, which is a useful product, would cease to exist. (Unless it was compulsory, like national insurance, but that's a very different thing, more akin to taxation.) So you have to allow discrimination, but maybe only where it makes a significant difference, and where you can prove that it does.

I believe that the disabled used to be rated up for motor insurance, and the law was changed such that this was deemed to be OK so long as the insurer could justify the increase from its experience (which seems the right tack to me). And they couldn't, so they stopped. In theory the marketplace should have achieved this, but it didn't, which suggest that a bit of legislation is a good thing.

Quite how you would prove that gender is a root driver of insurance claims, rather than a hundred other things which are loosely correlated to gender, I really don't know. But if you can't, maybe it isn't. I have a sneaking suspicion that a few years down the line the notion of using gender will seem bizarre.

[Soph K]

The European Court of Justice has ruled that insurers can no longer use gender in premium calculations.

Is this a good thing? Is it forcing equality where equality doesn't really exist or make sense? I'm undecided - seduce me with your opinions.

Jon, I have absolutely no idea what you're on about. ...Then again, why did I look at this topic when I have no idea about car insurance things? I'm weird. I'm random. I'm tired. I'm bored. I'm hungry. I'm Sophie. ...

[David Williams]

And there was me wondering why I haven't been able to raise the enthusiasm to post for a couple of months.

[Soph K]

And there was me wondering why I haven't been able to raise the enthusiasm to post for a couple of months.

[Gavin Chipper]

In the past I always thought it was completely wrong to discriminate based on what sex you are - just because other men crash more on average, why should you as an individual be penalised? So I always thought you should base it on an individual's driver record.

And I do still largely think this. But there are other things to take into account. Even if they don't discriminate on grounds of sex, people are still going to be penalised because of the driving records of others. The starting rate for a new driver is still going to be based on average statistics - it's just that it will be based on all new drivers rather than just male or female ones. So even if you are a great driver, you are still being penalised for being a member of the human race.

On Mars, on the other hand, people crash far less often so Martians have much better insurance deals. But when people move to Mars, the anti-discrimination laws would mean that the Martians would suffer as soon as the humans turned up. And we'd do very well out of the Martians' arrival here. So I can see why the Martians would be annoyed by this.

With the other stuff, I suppose one problem with stuff like sexual orientation is that it seems rather intrusive to be asking such a thing. Where would you draw the line about what you could ask about someone?

But since insurance is a legal requirement, I'm not sure if these charges should be in the hands of profit-making companies. I wonder if it would be feasible to just have a flat rate for basic 3rd party insurance and then leave the companies to do the rest.