My wife and I own a well known Electric Vehicle that monitors our driving in Southern California. The car company knows how many miles we drive and the car company knows that Dora is a safe driver based on her average speed and the braking she engages in and the fact that she doesn't engage in stop and go driving. While I have a driver's license, I do not drive.
Six months ago, I asked Dora; "Why doesn't Tesla sell car insurance? We would get a better deal from Tesla because it knows that you are a great driver." The Wall Street Journal reports that GM is now selling car insurance. As usual, I am ahead of my time as I can see the future!
An interesting self selection issue will now arise. If you are the boss of a stand alone insurance company, how should you price insurance to the select sample of Tesla and GM car owners who chose not to buy insurance from those companies? Do you see that these are the risky drivers? As stand alone insurers raise rates on these folks, will a death spiral emerge? I think the answer is no because all drivers must be insured. The "sick" drivers will no longer be quoted low rate and they will pay more for their bad driving. If driver "quality" is a choice, then these drivers will have an incentive to improve their driving and our roads will become safer due to the rise of Big Data and the risk pricing that GM and Tesla and other new insurers can engage in.
Death to the pooling equilibrium! In the case of health insurance, I understand why a pooling equilibrium offers cross-subsidies to sick people. In the case of driving insurance, the pooling equilibrium hurts society because we want bad drivers to have an incentive to invest in becoming better drivers. In the case of health insurance, yes we want people at risk of getting very sick to engage in healthy habits but this "transformation" is likely to be more challenging (given genetics) than transforming a bad driver into a safer driver.