Why Treating Patents As Property Is A Bad Idea

Say That Again

from the yet-again dept

We’ve pointed out in the past why it doesn’t make much sense to treat “intellectual property” as “regular property,” since it ignores some very important differences between the two. James Bessen and Michael Meurer, who wrote the recent book Patent Failure have always taken a slightly different approach. While they agree with us that the patent system tends to do more harm than good (and they’ve got a ton of research to back that up), they claim that the problem is that patents don’t act enough like property. They say the problem is that the “fuzzy boundaries” around patents mean that there aren’t clear rules or “fences” as with real property. So, their prescription is to look for ways to treat patents more like real property.

Eric Goldman points us to a recent paper by well known law professor and patent system expert Mark Lemley on why it’s a good thing that patents aren’t treated more like property. Lemley is mostly a patent system supporter, but (unlike some around here) he appears to recognize that the system could be improved, and seems open to evidence concerning where the patent system does more harm than good.

This paper makes some important distinctions between patents and real property, and notes that it’s probably for the best that in the real world patents actually are not treated like property, because it would slow down innovation if, before anyone invented anything, they had to secure a ton of agreements with patent holders:

It is currently very much in vogue to talk about patent rights as a form of property, and in particular to draw analogies to real property. So let’s engage in a thought experiment: what if we took the analogy seriously and actually behaved with patents as we do with real property? Product manufacturers would have to stop ignoring patents. No venture capitalist or bank (or shareholder, should Intel fund the project internally) would give Intel the money to build a new manufacturing plant (or “fab”) unless it could demonstrate that it had conducted an exhaustive search for patents it might infringe in manufacturing its chips and had obtained irrevocable or at least long-term licenses to use any patent that anyone might conceivably later assert against the chips or the manufacturing plant. Intel, in turn, would look to a group of “patent insurance” firms that would spring up and that would conduct the search and determine what patents needed to be licensed. Unless and until all of this had happened, Intel could not start construction of its fab, much less make or sell chips produced by that fab. If there were significant disagreement over whether a party legitimately owned patent rights, perhaps Intel could bring a declaratory judgment action to try to clarify those rights, but it would hold construction in abeyance until it got an answer. And since there is no experimental use defense to patent infringement, scientists at both universities and corporations would have to conduct a similar search and wait to get permission from all possible interested parties before they began their research, lest they infringe a patent in the lab.

Would this world be desirable? I’m skeptical. Let’s begin with the benefits of such a world. Patent owners would get paid early and often. Patent litigation would decrease, or maybe even disappear entirely, because anyone who wanted to make a product would find the patent owner and enter into a deal up front, or else not make the product. And patent owners who compete in the marketplace, and rely on the patent to preserve exclusivity, would not face competition during the often-protracted period during which the patent is being litigated.

At the same time, these benefits would come at significant cost. First, both research and the manufacture of products would be regularly delayed for years and perhaps decades as potential defendants identified and cleared rights….

Second, a real-property patent system would replace competition with central coordination in a significant number of cases. So far we have assumed that the patent owners will be willing to license their patents. But that is likely not to be true in many cases. Patent owners who compete in the marketplace want exclusivity, and there is no license price an equally efficient competitor will be willing to pay that will compensate for the loss of monopoly rights. Even patent owners who do not compete in the marketplace may find it more lucrative to grant an exclusive rather than a nonexclusive license to someone who does make a product, for the same reasons. Nor will a competing company be particularly sympathetic to efforts by outsiders to engage in research on the invention if the effect of that research will be to design around or improve that core invention. The effect of a real-property or title-search system is to replace competition in the shadow of a patent while it is being litigated with single-firm markets whenever the patentee participates in the market, either directly or by proxy. Researchers who could not obtain a license would direct their scientific efforts into different fields, and potential competitors would do the same, meaning that the owner of a core patent could control who, if anyone, worked on a particular technology. If you believe, as I do, that the evidence suggests that competition is often a better spur to innovation than monopoly, removing that contingent competition is a potentially significant cost.

Third, and perhaps most important, a significant percentage — maybe as many as three-fourths — of these patents turn out to be either invalid or not infringed. It is this probabilistic nature that most critically distinguishes patents from real property. Under the current system in patent-ignoring industries, consumers benefit from competition during the time before those patents are invalidated or held not to be infringed. Under a real-property patent system, the owners of invalid patents can capture supracompetitive profits during the time before their patents are invalidated, profits made at the expense of consumers and that they will never have to disgorge. That extra profit, in turn, would create significant incentives to obtain and enforce dubious patents….

Finally, people usually build a house on a single plot of land, while as I have noted, there may be hundreds or thousands of rights that must be aggregated to build a multi-component product. As Carl Shapiro and I have argued elsewhere, this fact exacerbates the patent holdup problem and leads to systematic overpayments by manufacturing companies, because individual patent owners won?t discount the royalty they charge to account for the complementary rights owned by others.

He goes into a lot more detail beyond that, and basically suggests that the current system of companies simply ignoring patents until later is probably better than going to one where patents are fully treated like property.

However, he then suggests a bunch of changes to the patent system that could (he believes) create a more reasonable middle ground: (1) more resources to the patent office to get through more patents faster (2) preventing patent applicants from delaying the process through things like continuations (3) requiring rapid publication of all patent apps (4) allowing peer-review and post-grant opposition to better establish what a patent really covers (5) allow independent invention as a defense against infringement (6) change the rules on “willful infringement” (which currently encourage people not to look at any patents by tripling damages if they find out you even looked at a patent you’re later found infringing on) (7) change patent remedy rules to end situations where patent holders can hold up the production of useful innovations (8) require companies to do patent searches to make sure they are not infringing (this, combined with #6 almost flips current willful infringement rules on their head) and (9) require the publication of any patent license terms.

It’s quite a list… At this point, this post is long enough not to get into whether or not this is a good or bad proposal (suffice it to say, I think some of it makes sense, while other parts are troubling — and I’d bet that patent system supporters will say the ones I think are good are bad, and the ones I think are bad are good), but wanted to post it up here for discussion.

Filed Under: mark lemley, patents, property