Hi. Let me know if I’m misunderstanding how insurance works here:

The inputs below show the core insurance model variables with an assumption that no profit is being made. Obviously where the profit exists is in the fact that the likelihood of an accident would be less than the 10% shown here (for conventional vehicles/traffic), which is what makes the insurance industry a viable gamble — theory of large numbers implies that if you insure enough people, you’ll get a result that comes close to the probability of the <10% accident rate, and as such, will be able to operate a profitable business.

On the left, the monthly accident rate is 10%, the average claim is $1,000, and people are paying $100/mo in order to get to a nice happy break-even point (again, we’ll omit/assume it’s obvious where the profit could be built in.)

On the right, the monthly accident rate is 1% because of the safer vehicle technology, the average claim is still $1,000 (because a wrecked car is a wrecked car), and people are paying $10/mo for the model to break even.

Does this all look correct?

If so, please explain the following:

1. How is everyone on the right column paying $10/mo when the “1%” who get in accidents are now the same 1% all the time? It’s no longer a random event that “could happen to anyone” — the 99 customers in self-driving cars are not getting in accidents (certainly not at-fault accidents, to say the least). So, following the wisdom of any good fiscal conservative, is it not obvious that the 99 responsible drivers/cars would pay closer to a tenth of that cost, while the person who keeps getting in accidents would pay 10x? If that were the case, the one conventional driver would still be paying the same $100/mo he paid in the left column when all cars were conventional, so nothing for him has changed (of course, it has indeed changed, since everyone else in society has drastically improved their financial leverage… but we can leave that alone for now.)
2. If the accident rate drops to 1/10th of what it was with conventional cars, and premiums follow suit, then how does the insurance company maintain enough revenue volume to operate a successful model? Geico scales its staff and infrastructure down to 1/10th its current operation (as any multi-national corporation would do, right? Just happily wither away until there was nothing left?), but it still has to insure a large enough swath of customers to grab that 1% who is accident prone — or, target them directly and forget about self-driving car customers. So you’ve got 1/10th the claims adjusters, paperwork processors, customer support reps, and connections to body shops. What happens when one of those 1% drivers in Wyoming gets in an accident, and your closest claims adjuster is in Colorado? You fly him out there? With what money? Certainly not with the 1/10th of revenue you have. You need to drastically increase the cost of that driver’s premium to offset your expenses if you want to continue operating a viable business model.

Did I miss anything?