The European Court of Justice’s ruling regarding gender disparity in insurance and pension is surely a dangerous precedent. It’s almost always a good thing to remove gender disparity – that I agree with – but is ‘almost always’ = ‘always’?
For those not in the know, insurers in the UK will no longer be able to charge women less for car insurance than they do for men. An associated change will affect pensions – as women typically live longer than men, an annuity bought by a woman is typically worth less than an annuity bought by a man, and that is also about to change.
Are these changes correct? Is it right to demand that gender is never allowed to be an issue over which people are treated differently? Should the law be focusing on being blind to objective, correctly analysed and unbiased evidence that shows men and women live and act differently?
The reason that insurers give women discounts is not because women are not ‘better’ drivers. Statistically, they are safer – this data taken from the Scottish Government shows this in a loose sense:
Typically, women are injured less frequently in accidents than men. Also, the fraction of serious injuries that result in fatalities is lower for women than it is for men. Yes, I know, this doesn’t adjust for the relative frequency of each gender in cars in Scotland typically at any time, but it’s probably safe to say that it’s close to 50-50. Car insurers are keen to do the stats – they want to save money, and attract customers. If they can identify a demographic that’s cheap, they can price out the competition by offering cut price premiums to those of lower risk. A risk, incidentally, that is established by evidence rather than some unwarranted prejudice.
The fact that this does not always translate well to individual cases is at the heart of statistics. Insurers know that some men can be very safe, and that some women can be very dangerous. But they’re not going to spend reams of money investigating every individual they insure to assess their driving competence – it’s quicker and cheaper to make a judgement based on statistics, and the payoff is higher.
OK, if you’re not convinced by this argument, let’s take the game further. If gender disparity in insurance is no longer allowed, what about age disparity? Should younger drivers be forced to pay higher premiums simply because they are younger? Of course, most people would say – being younger means you have less driving experience, and are therefore more likely to have an accident. Precisely the right sort of broad statistical judgement that insurers make – while the person in the street bases their opinion on anecdotal evidence, the insurers use more rigorous statistics to confirm it. Some young drivers are excellent and safe, but you can’t check ’em all, again because it costs too much to be profitable.
This statistical judgement is identical to the gender case above. It uses the same approach to find a broad correlation between the properties of a human being, and the frequency of potential insurance claims from human beings that possess that property. This is not discrimination based on misguided, unwarranted prejudice, it’s an actuarial, evidence-based calculation that insurers are justified to make.
Bottom line: there are much bigger cases of unwarranted discrimination out there, like the gender pay gap, which has no statistical casus belli, which we should be working really hard to fix. Why are we being forced to focus on such irrelevant issues?