I agree with all your points re. the way insurance works.
But your last line that “Life’s not fair” is exactly the issue… the argument is that it’s not “fair” that now women pay more to subsidise the risk of male drivers. Insurance as you’ve noted is inherently unfair to an individual (as it treats them as part of an aggregate), so why is spreading the load across a wider pool of people any worse?
This is only partially true. It’s societal convention combined with competitive pressures and marketplace realities.
IOW, if everyone in the world would truly eliminate use of gender as a factor, then the insurance companies wouldn’t care, and they would make as much money from the unisex pricing as they would from gender-specific pricing. The question is whether that can happen. In the absence of the court ruling, it definitely could not, and the question is whether the court ruling is enough.
As an example, suppose Company A in the US, or in the EU without this ruling, decided to unilaterally institute unisex pricing. So if they were charging men $2K for auto insurance and women $1K and they had a 50/50 mix, they would now charge everyone $1.5K. Meanwhile other companies continue to differentiate by gender. At this point, for women, it would now be a very foolish decision to insure with Company A and pay the blended rate, while for men it would be a great deal. What would happen is that virtually all women would insure elsewhere, while all men would choose this company, with the result that Company A’s rates would eventually end up being close to the old Male rate, and gender-specific pricing would carry the day, with separate companies insuring males and females.
Under the EU ruling this theoretically can’t happen, since all companies are in the same unisex boat. The question is whether companies can find other proxies for gender, and whether consumers can vary their investment choices.
As an example of the first, you might look at occupation. Nurses are predominently female, for example. If you were an insurer offering auto insurance, you might want to discount the rates you offer to nurses, because nurses are cheaper. Which is correct, but not because of anything inherent about nurses, but because they happen to be mostly female. You might want to offer higher rates to lumberjacks. And so on.
And the same goes for consumer choices. Auto insurance is mandatory, but others are not. If you are a man, annuities have just become a much worse deal for you, and if you are a woman they’ve gotten a lot better. I would expect that a lot of men will be inclined to consider other investments, while women will be drawn to annuities, with the result that the female percentage of annuity buyers will rise significantly, which will drive the unisex rates closer to the current female rates (which will make it an even worse deal for men, which will drive the female percentage even higher and so on).
Insurance is tricky in this way, and it’s not always easy to control things by judicial fiat.
That is a good point.
It is no more fair to charge an individual male good driver extra because other males get into accidents, than it is to spread out the costs of the bad male drivers among both male and female good drivers.
This is especially true since gender is mainly used because of the ease of data collection, rather than because it is an important factor in safe driving. Males getting into more accidents tells you nothing about whether they drive less safely; perhaps women are more risky drivers, but spend less time driving.
Certainly time spent driving, income, education, spatial awareness, etc, are valuable information, but they are much harder to accurately collect.
But “fairness” is not the goal. The efficiency of the system is better using the gender data. It would be great if better data could be collected in a cost effective manner, but unless you can figure out how it only makes sense to use gender.
I think this would be easy, using existing categories.
So you could assume that:
So you could adjust your premiums to customers whose vehicles fall into certain categories. This already happens, of course, but you can widen the gap so customers with large, fast cars who cover a lot of miles on business (which tend to be men) pay more than someone with a small low-mileage runabout (which tend to be women).
If it were so easy, you can “assume” blithely as well that someone would leverage the opportunity.
So you could assume that:
On waht actuarial data is that based?
Same question.
Same question.
Same question.
Generic observations do not actuarial tables make.
It’s already leveraged. All cars can easily break the UK national speed limit, but you still pay more for cars with a higher CC even though that in itself doesn’t raise risk.
Are actuarial tables independently validated, or do insurers keep their own records?
(This is supported by actuarial tables)
The other factors - engine size, usage etc - are identified as risk factors in actuarial tables, but I don’t know whether they are specifically linked to gender (I presume they are, but could be wrong).
Interestingly for the main argument (re. who should pay) it is already the case that young men currently pay less than their true actuarial premium with other groups paying over the odds (so effectively subsidising young men). The ECJ ruling will increase this disparity, but it does already exist.
To clarify.. this clearly does raise risk, which is why it’s a factor, but there’s nothing intrinsic in higher engine CC which would cause more risk apart from the habits of those who choose such cars?
“Life is not fair” is an argument that you can’t rectify all the unfairnesses of life. It’s not an argument that you should create new ones. In this case, the concept of insurance is a trade-off which has winners and losers, and you can call that unfair if you like, but the unfairness is a naturally arising one. Unisex pricing is an artificial unfairness, and ironically one that is being imposed for the very reason of making things fairer.
I don’t know about the specific factors cited, but in general your argument is a logical error (assuming I’ve understood it correctly).
The important thing to remember about rating factors is that they have a lot of overlap. For example, sports cars have higher accident rates and young men have higher accident rates, but you can’t just multiply these factors to get the factor for young men driving sports cars. Because a lot of the reason that sports cars have such high accident rates is that they tend to be driven by young men, so you’d be double counting the same factor.
But this changes if you outlaw use of one factor. As it stands now, you would not rate a sports car simply by computing the increased likelihood due to sports cars, because you are also capturing some of that likelihood in a more refined manner by seeing if this sports car is being driven by a male or female. So you would currently use a factor that reflects the increased accident likelihood of a male/female sports car driver over a male/female non sports car driver. But if you outlaw gender, then the sports car factor would have to - or at least could - be adjusted to reflect the overall average, since you would not be capturing that elsewhere, and the male accident rates would be partially accounted for in a higher sports car rate.
In addition, there are a lot of factors that are not worthwhile to use at all under the current system, since they have a lot of overlap with gender, which is already reflected. There’s a point of dimishing returns to creating ever more ratings categories, and they also create practical difficulties. But if you outlaw gender, then there will inevitably be a lot of factors that become more useful, since they will be capturing an important factor that would otherwise be ignored. (The occupation one I’ve suggested above might be an example.)