Tort-uous Economic Logic

Sasha's recent post on the subject of tort liability prompted to Mark to make a thoughtful reply. I'm too lazy to summarize the whole interchange, so I'll just let you follow the links. However, I had to respond to the following passage in Mark's post:

"[Sasha] says that ideally an entrepreneur would get all the benefits and pay all the costs of his innovation, and argues that tort damages are part of 'all the costs.' But -- putting aside the external-benefit point, which Sasha makes -- the only entrepreneur who gets all the benefits of his innovation is the hypothetical price-discriminating monopolist. In the more usual case, there are consumers' surpluses. (Being forbidden to buy a vaccine at the market price would leave the consumer worse off.)
"So if we stick manufacturers with all the costs, but can't secure for them all the benefits, then the overall level of market activity will be suboptimally low."

I was about to reply by saying that Mark is looking at total costs and benefits, whereas marginal costs and benefits are the relevant parameters. A firm doesn't have to be a perfect price discriminator to face the full *marginal* benefit of his production decision, because the valuation of the marginal consumer is generally reflected in the maket price (unless this is a good whose production creates external benefits).

But Mark is one smart guy, and he seems to know something about economics, so I figured he probably wasn't making that mistake. Then I realized he might be talking about a different problem: external costs that are fixed costs, incurred as a one-time result of innovation (the word he used in his post) or business start-up. If so, then his argument goes through, albeit as a special case. The problem is not that any existing firm underproduces, but that there are not enough firms producing in the first place. The threat of liability prevents them from opening, even though the total benefits of the activity (as measured by consumer surplus) exceed the total costs.

I was going to post that instead, and then I realized Mark might be talking about yet another issue: the general theory of the second-best, which implies (among many other things) that some amount of monopoly power can be desirable in the presence of externalities, and vice versa, because the underproduction attributable to monopoly power can counteract the overproduction attributable to negative externalities. As a result, imposing liability for negative externalities can result in underproduction.

So now I'm really unsure what exact point Mark was making. I guess I'll email him, point him to this post, and ask him to let me know. For now, I will just make the general (and vague) observation that all of this analysis is conducted without any reference to the Coase Theorem, whose main lesson is that external costs are not *necessarily* ignored by decision-makers; whether they are considered adequately will depend on both transaction costs and the institutional structure.

UPDATE: Mark informs me that my interpretation #2 (external costs that are fixed) was what he meant.