Not every disparity is discrimination, car insurance edition

Here’s a Pew report from back in February that I recently came across:  “What? Women Pay More Than Men for Auto Insurance? (Yup.)”  Here are some of the key bits:

It’s a widespread belief that men pay more for automobile insurance than women. But that’s only true for young adults.

Several studies in 2018 and 2017 revealed that women over 25, particularly those between 40 and 60, often pay more than men — not less — for auto insurance, all other rating criteria being equal. . . .

In an interview, [former California Insurance Commissioner Dave] Jones said it’s fair for insurance companies to set premiums based on a driver’s accident history, number of speeding tickets and other factors that are under the driver’s control. But using gender is unfair because a person has no control over that, he said.

What’s going on?  The report indicates that there was no seeming consistency across insurers.  It cites a 2017 Consumer Federation of America study, in which, among 60 year olds, differentials ranged from a premiums 4% higher for men at Liberty Mutual to a 12% higher for women at Geico.  For 40 year olds, the differences among 6 insurers studied were -1%, 0%, 5%, 8% and 16% higher premiums for women.  And, oddly, for 20 year olds, premiums were higher for men, ranging from a 5% to a 16% difference, except at Geico, again, where women had 6% higher premiums.

Is this discrimination?  Are women being charged more than men, for no particular reason except that rate-setters want to give advantages to men because of the patriarchy?  The Pew article suggests, in the midst of a broader discussion around rate-setting processes, that there’s something nefarious going on:

But a professor at University of Minnesota Law School, Daniel Schwarcz, said if companies are not allowed to use “outdated stereotypes based on generalities” about men and women, the insurers will have to consider “more directly” such measures as the actual number of miles driven, the number of years customers have been driving and where they live.

Really?  A lawyer is asserting that actuaries develop rate models which add in some sort of factor based on the “women are bad drivers” stereotype their fathers or grandfathers might have believed, and that’s worth referencing in an article put forth by an organization as respected as Pew?

It should go without saying that actuaries price insurance premiums based on the totality of the data available to them.  Price a rate too high for a given rate class (age, sex, residence, driving history, etc.) and you lose a sale.  Price a rate too low and you lose money on that sale.  Especially now when customers are able to comparison-shop far more easily, insurance companies’ actuarial departments want to get this right.  At the same time, I suppose, if a company has identified a demographic which it believes to be particularly susceptible to marketing and less likely to comparison-shop, they might focus more on marketing to that group and worry less about the competitiveness of their prices.  (Heck, are women less likely to comparison-shop insurance and more likely to choose a brand that gives them warm fuzzies?)  And all of the above is true even with the disparities in pricing among various insurers, simply because each of them will have different pricing models, and will have different claims experience even for the same demographic group (and an objectively similar demographic group could differ if different insurance companies attract different types of customers) — complexities of business operations which these companies are under no obligation to disclose to the general public any more than KFC must provide its secret recipe.

But the 2017 Consumer Federation of America study the Pew study references takes its claims even further, writing

The inconsistent pricing decisions of these insurance companies illustrates CFA’s concern that tying auto insurance rates to factors that a customer cannot control and have nothing to do with their driving safety record – such as one’s biological sex – leads to unfair discrimination and indefensible claims of actuarial soundness. . . .

“Every state but New Hampshire requires drivers, regardless of their sex, to buy auto insurance, so regulators and lawmakers have a special obligation to make sure coverage is priced fairly,” said CFA insurance consultant Douglas Heller, who conducted the study . . . . “What we have found is that insurance companies punish female drivers with perfect records more often than men, and far more often than we expected. We also found that the insurance companies’ use of sex as a rating factor does not seem to reveal much in the way of a consistent risk assessment, and regulators should reconsider allowing companies to continue using it at all.”

But let’s back up:  should insurance rates be only about those characteristics which customers can “control” — in this case, driving history and maybe residence?  Based on this rationale, there shouldn’t be any differentiation by age, either, but no one suggests that 16 year olds and 36 year olds should have the same rates, because it is generally acknowledged that teens, as new drivers, are less skilled, even though these are both items that cannot be “controlled.”  (And, quite honestly, I find it believable that men, once they outgrow their impulsive years, might be more likely to be better drivers; in terms of external factors, women might be more likely to be distracted by kids they’re transporting somewhere, and, besides, it is not out of the question that men could have a better awareness of their environment, spatial awareness, following distance, reaction time, whatever, in the way that there are simply differences between men and women.)

What if a state mandated that, since it’s unfair to charge young adults more for insurance when they can’t “help it” that they’re inexperienced drivers, insurers couldn’t differentiate but had to wait until a driver got into an accident or got a ticket?  (Yes, I know, exactly the demographic with the most political power would be disadvantaging itself, so it’s purely hypothetical.)  Would it be fair to say that any subsidies young people receive would be evened out by paying relatively more than otherwise when they get older?  After all, they’ll earn more then, too, on average.

But intuitively we know this is not actually fair. It is true that all drivers are required to have basic levels of insurance, but there is discretion in terms of the deductible amount and whether, in addition to state minimums, one elects collision/comprehensive insurance.  Plus, of course, drivers purchase the cars they do in part knowing that insurance premiums vary among cars (due to the age and cost of the car plus relative repair expenses and risks of theft).  What happens if these optional coverages become subsidized for some groups?

Now, that being said, it would genuinely be interesting to see what’s driving the rate disparities (no pun intended).  But suggesting directly or indirectly that insurance companies are anti-woman isn’t helpful.

 

Image: https://pixabay.com/photos/traffic-highway-car-driving-road-966701/