Log in

No account? Create an account
laptop, geek, MacBook, bursting, breakout


Matt's LiveJournal

Random ramblings of a programmer, photographer, journalist and runner

Previous Entry Share Next Entry
laptop, geek, MacBook, bursting, breakout


Just heard that today is the day there's a ruling on whether charging blokes more for car insurance is a form of sex discrimination.

My first thought on hearing this was, "D'uh! Of *course* it's a form of sex discrimination. But that doesn't actually mean it's a bad -- or even unfair -- thing to do..."

But then perhaps I've been blinded by being in the insurance trade for so long. It's been about fifteen years now, come to think of it...

What d'you think, people? Should it be illegal for insurers to charge women less for car insurance because they cost insurers less?

What about my line of work? Is it a *bad* form of age discrimination to charge an eighty-year-old more for medical insurance than a twenty-year-old?

  • 1
You do end up with the logical end-point that it should be illegal to charge anyone more for any sort of insurance if the reason is something they can't do anything about (age, sex, perhaps even prior driving record if you want to be really deterministic about human behaviour). Depends how widely you want insurance to spread risk, I suppose. I certainly don't want to pay car insurance at the same rate as a boy racer with the premiums set high enough to cover our combined chances of having an accident.

But, presumably, in that system, you'd also have been charged less in your younger days as a boy racer because your premium would have been offset by all the sedate older drivers paying more... It all evens out in the end!

so in summary is science unfair -yup it is one of the few ways in which we as a species might be evolving much better if we speed up the devolving

this is absolutely barking and I can't believe they won't sort it out before Dec when it is due to come in otherwise I see the dole queues being added to be a whole load of actuaries

as most of my pension is final salary I don't even get that benefit - madness


But it's NOT a form of sex discrimination.

It's a "you're in a group of people that statistically have more claims" 'discrimination', which I don't think is at all.

Of course I might be lack-of-biased due to not owning a car I need to insure....

(Deleted comment)
(Deleted comment)
But in terms of insurance, is whether it's a cause important? We're just dealing with statistics and correlations, not cause and effect.

Actuaries, in my experience -- and I work with them all day long -- really do want their divisions, of any kind, to have a clear predictive power, and will drop stuff if it's not working. They experiment with whatever data they've got, all the time, looking for new ways of provisioning for their risks. I think it's remarkably unlikely, say, that sex *doesn't* make a clear, statistically significant difference to the cost of insurance claims, thus giving it predictive power. I'm sure it's measured very carefully. Probably by people like me :) And at that point, from an insurance point of view, it makes a good rating factor, whether it's a *cause* of the risk or not.

I honestly don't see any difference between discriminating between the sexes on car insurance and discriminating between the old and the young on medical insurance. You're splitting your market based on where there are observed correlations between an easily-measurable customer characteristic and the amount of money you fork out, on average, and charging them appropriately.

It's possible that there are only two completely fair approaches to insurance -- either nobody has any, and you take the risk of having to buy a new car all of a sudden every time you drive, or everyone pays the same amount -- or at least the same amount per mile -- based on the average cost of a claim across the whole population.

Obviously I'm not suggesting either of those is remotely tenable; I'm just pointing out that anything we do in the middle is likely to be both arbitrary, and done solely because we're in a capitalist society and insurance companies want to differentiate themselves from each other based on how they divide up and assess risks...

(Deleted comment)
I wonder if men tend to write off more cars[1] because they're generally taller -- maybe it's the distance between the brain and the brake pedal...

I'm not sure where I stand on your last point. Car insurance isn't one of the insurance industries I've been involved in, so I've not seen just how well-divided the figures are.

There's a correlation in the health insurance market (women are a smidge more risky than men) but not enough for the companies I've worked for to bother rating one sex higher than the other, even if the question is easy to ask. Some other health companies *do* charge women more, and I wonder if they'll have to stop now, even though there may be a more obvious causal link in health insurance than in car insurance?

(Incidentally, some reports I've just read are claiming that women get charged more for health insurance because they live longer. Which isn't the correlation I'm talking about; PMI is paid for annually, and average claims costs *per year per person* are higher for women *of the same age* as male comparison groups. In the UK market, where contracts are renewed annually, saying women pay more for health insurance because they live longer is a bit like someone paying more for car insurance because they drive a car for more years. True, but not necessarily relevant...)

[1] I've often heard it said that the main difference in costs between sexes is not the number of accidents, but because men tend to smash cars up a lot more seriously when they crash, but I can't see any available stats on that...

(Deleted comment)
Interesting -- none of the places I've worked do half-hour interviews. You've just filled out a form or two. Generally speaking, detailed questions are more likely to be used for medical underwriting -- personal exclusions, in other words -- than as rating factors. Is it possible those questions were to draw you out on historic medical problems which would then have been excluded from the new policy, rather than to weight the premium?

I would give a few details of our rating factors, but I suspect we'd be straying into commercially confidential territory, as opposed to the more general talking we've been doing. But I'd say that many of the larger PMI providers in the UK will rate using virtually nothing but age for many of their products. There are definitely some car insurers who use more rating factors than some medical insurers, even if they just rate based on age, no claims bonus, sex, postcode and annual mileage.

(Deleted comment)
Well, it would certainly be interesting if car insurance worked the same way. "What, you ran into the back of someone last year? Well, yes, certainly we'll insure you for the coming year. But we won't pay out any claims related to you running into the back of someone. You're on your own if that happens..."

  • 1