> Coase's theorem doesn't apply when the parties can't transact with each other to shift the burden.
That's true, but it's irrelevant to the point murbard2 was making. He's not arguing that the person who gets hit should bear the liability; he's arguing that it doesn't necessarily have to be Uber that bears the liability, at least not as an initial condition; it could be either the individual driver or the passenger, since all of those parties can transact to shift the burden. (Notice that he didn't list "the person who gets hit" as an option.) Since you were arguing that Uber should bear the liability (instead of just arguing that the person who gets hit should not), his point is a perfectly justified response to yours.
> You can make an argument that consumers are irrational and are unlikely to really look into whether or not the cars / drivers are insured, but that's a different argument than externalization
This part makes me think that he did not consider the scenario of a driver hitting someone else. It is precisely the problem of externalization. If the initial liability isn't on the company, it will be on a party (the driver or passenger) that won't actually have the assets to pay in the case of an accident. In that case, it is de facto externalized onto the poor sap who gets hit.
> If the initial liability isn't on the company, it will be on a party (the driver or passenger) that won't actually have the assets to pay in the case of an accident.
Why not? In the hypothetical Coasian scenario where the company doesn't have liability but the driver and/or passenger will negotiate with the company to change the terms to compensate, the driver and/or passenger know that they are liable and the company isn't. (If they didn't know, they wouldn't know to renegotiate the terms.)
What you appear to be assuming is that the driver and passenger will not know, i.e., that they will be blindsided by the fact that they are liable, the first time the vehicle gets in an accident. But even if that's the case, why is that the company's fault? Isn't the driver responsible for knowing what he is and is not liable for when he takes paying passengers? Even if he wasn't taking paying passengers, he'd still have to have insurance, so it's not like the concept of him being liable when his car hits someone is new to him. (The passenger has more of an excuse here, since he's not driving and would not be expected to bear responsibility under just about any legal standard currently in play.)
Because even though they might be technically liable, they'd never actually have to pay, assuming you don't have a law requiring them to carry commercial insurance. If you do have such a law, then we're not talking about anything Coasian--we have just picked the driver to carry the cost of accidents.
> even though they might be technically liable, they'd never actually have to pay, assuming you don't have a law requiring them to carry commercial insurance
Um, I had assumed that "liable" meant "liable", not "technically liable but not actually liable". We're not comparing scenarios where the law is the same but we arbitrarily shift the "liable" label from one party to another. We're comparing scenarios where:
(a) the law says the company is legally liable when any driver they contract hits someone while driving a passenger under that contract, and has to carry insurance accordingly, vs.
(b) the law says the driver is legally liable when they hit someone, and has to carry insurance accordingly--commercial insurance for when they're carrying passengers under contract, and ordinary insurance for when they're driving for personal reasons.
The Coasian point is simply that under legal regime (b), as compared to (a), the drivers will negotiate for higher wages or other additional compensation to offset the increased cost of their insurance.
> If you do have such a law, then we're not talking about anything Coasian--we have just picked the driver to carry the cost of accidents.
Sigh. This is just case (b) above, and the Coasian point about the comparison with case (a) does apply.
That's true, but it's irrelevant to the point murbard2 was making. He's not arguing that the person who gets hit should bear the liability; he's arguing that it doesn't necessarily have to be Uber that bears the liability, at least not as an initial condition; it could be either the individual driver or the passenger, since all of those parties can transact to shift the burden. (Notice that he didn't list "the person who gets hit" as an option.) Since you were arguing that Uber should bear the liability (instead of just arguing that the person who gets hit should not), his point is a perfectly justified response to yours.