Is it sexist for the insurance industry to take gender into account when setting its premiums?

[Lio]

You've probably seen article like this one in the last few days, talking about how women could find insurance premiums (particularly car insurance) going up if the EU rules it discriminatory to take gender into account when setting its premiums.

A friend has asked whether this is discrimination against men, or just sound business sense? I've always thought the latter, given that the stastistics show that women have fewer car accidents and less spectacular ones than men. But if someone is told that the reason their premium is higher because they're a man, isn't that sexist?

I know there are other issues involved, such as age, but what do you think about the gender one?

Can I just check something here?

Are all of the people who want car insurance equalised, so that women pay a nominal amount more for their car insurance

Really happy to apply the same process to their pensions (assuming money purchase pension scheme) meaning that men's pensions will suddenly nosedive to the tune of 1000s of pounds?

TBH I'm happy for them to take sex out of the equation if it means I pay £10 more for my car insurance each year and gain 1000s at the end of my life.

Thanks guys!!!

Also interested how people intend to square this with the other risk factors that are taken into account - age, disability etc. Take them all out (end of insurance)? Or are people arguing that men having to pay slightly more than women for car insurance is much much worse than someone with an underlying health condition being denied life insurance to protect their family?

OMG

I know a lot of actuaries. They are assessing risk, they are not looking to exploit. Of course there is an eye for the margins, but they are categorically not loading male rates above and beyone what is indicated by the data.

Also, the data looked at will be accidents/deaths/whatever per thousand journeys or per thousand of the driving population etc. They will not be taking the total number of accidents of men and women and ignoring the fact (if it is a fact?) that there are more men on the road than women.

Also you say car insurance - you are happy for "sexism" as you see it to exist in other types of insurance? That doens't make sense.

Is all of this actually about outrage that statistically men are worse drivers than women, when everybody knows that women are shit drivers who can't park for toffee???

Can I just say, I'm an excellent driver and can park in just about any space in one maneouvre. Thanks.

I don think it's quite as simple as "statistically, men are worse drivers than women".

As I understood it, and I'm happy to be proved wrong, the statistics showed that women actually have a higher probability of causing an accident, but that the impact was typically much lower. The overall risk is therefore lower for women.

There may be many reasons for this difference, but it's not inconsistent to say that perhaps 'women can't park' and 'men cant keep within speed limits'. Both statements are gross generalisations and the tell you nothing about an individual. There is also nothing to suggest that either are biologically determined.

Allowing the insurance industry to continue to discriminate in this way perpetuates gender stereotypes. It says that stereotyping is fine if you can back it with statistics.

The main argument for this seems to be that pragmatism trumps principle. In truth, it isn't going to be a triumph for equality and neither will the sky fall in on the insurance industry.

Please explain how you think the sky won't fall in for the insurance industry if they are no longer allowed to use any statistical differences between things to assess risk?

The entire definition of insurance is it is agreement to take on a risk. If you are going to ban the ability of insurers to assess risk and price accordingly, then you no longer have insurance.

If you want to do away with the insurance industry and instead have everything run by the state then that is your perogative. I think it would be a shame though.

So do you think hairdressers and cosmetics retailers should have to charge the same for men's and women's haircuts/ products, Jeanine?

HerBex - I do.

I think if a woman is prepared to accept the same quality of cut for the same short back and sides type style, from the same sort of starting point, then she should be able to get it for the same price as a man.

The court case about that (back in the '80s?) was a travesty. Hopefully things like this will help to get it over turned.

SardineQueen - The insurance industry won't fall in if they aren't allowed to use sex as a factor because they can price on a whole host of other factors that are also correlated with risk.

There are plenty of factors they don't currently base their pricing on (some of which may have significant correlation with risk) and they manage just fine.

suchprospects

Please elaborate.

I would be grateful if you would tell me how you would go about pricing a life insurance policy if you are not allowed to take into account:

• Sex
• Age
• Medical history
• Medical underwriting

I am keen to find out what these "other factors" are that they could use when assessing this risk.

Thank you.

The issue is not that statistically "men are worse drivers than women."

The issue is that young men are statistically worse drivers than just about everyone else.

I really don't have any issue with charging high insurance premiums to the most dangerous, inexperienced group of drivers on the road. They're bad enough in bloody 1.1 Saxos, can you imagine what it would be like if they could afford to insure proper cars?

Well, I don't need to imagine it - lived in the US long enough where the weird insurance regulations in many states mean that you get 16 year-olds driving around in completely unsuitable vehicles with the inevitable carnage that results.

Quick question, do insurers also racially profile when assessing risk? Is for example life insurance more expensive for a young black man than a young white man? Presumably the answer is yes.

Risk is relevant to many businesses. If statistics show that black people, or women, or ginger haired people are more likely to default on loans this is presumably a valid reason to charge higher interest, or to refuse a mortgage or a loan altogether?

Perhaps this happens, but it does seem to cut close to public policy on equality if you are allowed to point to "statistics" as the reason for you discriminatory business model.

SardineQueen - I was only talking about sex and specifically in the case of driving insurance. I don't totally disagree with using protected classes in insurance pricing, but I do think causation should have to be proved in order to do so. And in particular they shouldn't be used as a proxy for lifestyle choices.

bluegrass I have not come across any type of policy where racial characteristics were used in assessing risk.

What is asked for and used: country of residence, overseas travel and the like. Obviously different countries have different risks associated residing there, ditto overseas travel.

There are certain medical conditions which are more prevalent amongst different ethnic groups, and this risk factor will come out on medical underwriting.

There are some things that they are not allowed to assume eg not allowed to assume homosexual men are at higher risk of various STDS instead they have to ask questions about specific sexual practices and condom use.

suchprospects so your argument is that only sex is a problem when used as a risk factor and only in car insurance?

That is not a logical position. You can't say that you don't want sex taken into account for car insurance (which would benefit young men) but you do want it taken into account in annuity prices (oh funny that means men benefit there too). You can't say that it's totally unfair to load a rate because of risk factors relating to someone's sex but it's totally fair to load a rate because of risk factors relating to someone's age or medical profile.

SardineQueen - No. I was simply pointing out that my post where I said "The insurance industry won't fall in if they aren't allowed to use sex as a factor because they can price on a whole host of other factors that are also correlated with risk." was specific to sex and driving insurance.

Where sex (or another protected class) has a causal relationship with risk, then I think it is a valid factor to include (though only to the extent it is a factor). However, given the insidious way discrimination has previously made it difficult for people to break out of socially assigned roles, I think it ought to be incumbent on the industry to prove that such discrimination is in fact wholly necessary to the good running of the industry. They shouldn't just get a pass because they can show a correlation.

But it makes no sense to single out one class of insurance and one protected characteristic. Either it's wrong or it isn't.

So I guess the question is why have men people got their underpants knickers in a twist about this single thing? Yet all other types of insurances and potential discrimination does not cause the same (any) outrage or alarm. Hmmmmm I wonder...

The OP (and the court case the OP is about) singles out one class of insurance and one characteristic. So that is what my specific comments were about. My general position is that of the second paragraph in my last post. Sorry I wasn't clearer.

I don't care a great deal about driving insurance, but I do care about the general principal.

That's contradictory though.

If you care about general principle then it doesn't make sense to single out one class of insurance. Or indeed one single protected characteristic.

On whether risk factors are causal or simply a correlation - again you will find that throughout the industry very few risk factors are directly causal, most show a strong correlation. Again, if you remove the ability of insurers to use risk factors based on strong correlation, then the whole thing falls over. I can think of very few situations where direct causality is in play - medical underwriting is one of them. Your plan would knacker the travel insurance industry for a start.

Why are men so keen to lose such large amounts from their pensions? Simply because they won't have it that statistically young men are a bad risk for car insurance? It's a bit silly really, the people bringing this lawsuit are cutting off their nose to spite their face.

SardineQueen - There is nothing contradictory about using a specific case as an example, but needing to be more generalized when talking about the wider general principal. I'm glad about this particular case because of the implications for other types of insurance and other protected classes. How that actually plays out in each particular case won't necessarily be the same.

On whether risk factors are causal or correlated - correlation is a useful proxy and used often by many people. But protected classes have a history of discrimination that is not justified so insisting on cause is a good way of ensuring individuals are not continually paying for the lifestyle "choices" that have been pushed on the class as a whole. Why should the insurance industry be allowed to use a protected class as a proxy for other factors when we don't generally allow other industries to? What's so special about them?

Where there is a causal relationship it could be detrimental to the industry as a whole to ignore it since selective purchase could skew risks badly (ironically less so with driving than with most types as driving insurance is mandated).

But you seem to be arguing that in some cases rating on sex may be acceptable, and on others not. While at the same time saying that you have a general principle that it is absolutely wrong to rate on sex. I see that as contradictory.

The reason that insurance companies are allowed to use protected characteristics when assessing risk is because they have agreed to abide by a strict set of conditions that mean they must be able to prove that there is an objective and statistically significant reason for it. When the equalities legislations were mooted and then enacted at EU level and enacted at local level the insurance industry was there arguing its case that it be allowed to do this and so far teh legislative bodies have agreed. As long as there is a good statistical reason then they are allowed to do it. And the reason that they have won this argument (so far), and are allowed to do it, is that without it whole lines of business will be withdrawn which will have an overall negative effect on society. If there is no possibility for obtaining life insurance then that is not a benefit to people in general, it is a loss to them. It would potentially mean that the state would have to set up many more lines than it already runs and people would not have the option to "top up" (again thinking things like life insurance). There is also the small matter of all the money and jobs sloshing around in this industry which will be lost.

I think the point is that if this "general principle" is found to be correct - that things like sex, age, medical situation - are not allowed to be considered then it will have wide ranging implications for the whole industry and so far the people legislating for this have not been inclined to cause that much disruption etc.

What questions would you ask, as a motor insurance company, in order to get at whether a person was a high risk for a crash or not, then?

You can't use age or sex, nor ask anything that is unduly intrusive, or that people will be able to easily lie about (eg do you like to break the speed limit), and you can't use any methods that are disproportionately expensive.

Assume the people we are looking at here (the young men new to driving) have no driving convictions or claims experience.

How are you going to do it?

I'm interested in your point that causal elements are easily identified in motor insurance.

So what are they and how are you going to dig them out?