I am quite sympathetic to the deadweight loss arguments in this paper, but placing the blame on the trolls is a bit misguided. They’re responding to incentives, even if we don’t like them. I wish the paper were more oriented around the law and the practices of the patent office.

Also, again while sympathetic, this paper was clearly researched after the authors had taken a position on the issue (notice the book mention). They started with a position, and went looking for data ex post facto.

They make no distinction between more and less legitimate patents. If a company is infringing on what many of us might consider a legitimate patent, then yes they will suffer financially. The paper elides over such a distinction.

Using stock price as a proxy for future foregone profits is clever, but it’s a heck of an assumption. I don’t think it rises to the level of empiricism. It’s interesting for sure, but best understood as a statistical artifact.