Worthless Patents

In my internet travels, I came across a useful article by Kimberly A. Moore titled “Worthless Patents.” At the time she wrote the article, she was a law professor at George Mason. Presently she is a judge with the Court of Appeals for the Federal Circuit, and served as one of the three judges to hear the Stanford v Roche case as it made its way through the appeal process. By way of background, Kimberly Moore has sterling tech credentials, including a multiple degrees from MIT.

For the “Worthless Patents” article Moore pulled all the US patents issued in 1991 and asked how many of them were allowed to expire at one of the three maintenance fee renewal terms. The number is crazy wild–53% (51,949 of 96,713) were allowed to expire before full term. Moore then points out some interesting properties of the expired patents relative to unexpired ones. The expired patents have fewer inventors, fewer claims, shorter prosecution times, fewer related applications. It’s not clear what the variability might be, or whether these relationships would hold if one examined multiple years, but the observations themselves are intriguing. 

Moore then looks at the characteristics of the patent owners. Here is her Table 2:


The thing that leaps out is the US government lapse rate–nearly three-quarters of government-owned patents issuing in 1991 were allowed to expire before their full term, and over half were allowed to lapse by 8 years. It happens that 8 years was a key term for exclusivity in the old Institutional Patent Agreements. It would appear that the US government was, at least in the period 1991-2003, continuing without fanfare the approach of placing inventions made with federal support into the public domain after holding the invention for a brief period to ascertain its importance.

I am not persuaded, though, that such patents are “worthless”–but rather that a public policy of placing inventions into the public domain is valuable. It’s worth some discussion, at least, as to how one characterizes value, both of patents and of the underlying inventions. Note that the first objective of the Bayh-Dole Act is to use the patent system to promote the use of federally supported inventions. The focus is on use of inventions, not money-seeking; the focus is on use of inventions, for which patenting is a directed means, not itself an end.

Moore turns then to looking at what classes of patented technology are most likely to be maintained to expiration of their patent term. Here is Moore’s Figure 3:


For patents issued in 1991, semiconductors are the thing, and biotech and drugs are not so much. Again, it’s difficult to tell if this is just a variation in one year’s worth of patenting, or something of a trend. Even then, an opportunity for semiconductor investment might not coincide with that for biotech. I expect there are a lot of DNA patents after the Myriad decision for which maintenance fees are not going to be paid.

What would be really interesting to see is data for university-based inventions and those made under Bayh-Dole’s standard patent rights clause as subject inventions.  How long do universities hold patents? How many of university-based inventions do the inventors get to manage personally or through non-university agents? How many of the patents on these inventions are abandoned before the expiration of the patent term? How many of the patents in each of these categories cover subject inventions? What happens when a university and/or inventors allow the government to take title to these inventions? Does the government routinely file patent applications? Does it let patents lapse at 8 years? Are most of these patents “worthless”? Inquiring minds want to know. Indeed, how could anyone think to discuss a national research innovation policy without having this information available? Yet here we are, with all sorts of opinions, and the information is not being reported. From the patent office, we can get information on the patents that issue, presuming that the government rights statement is generally accurate (I have my doubts, given what I have seen of university practice), but we cannot see the activity ahead of patenting–the management of the inventions themselves.

One last point. Even though some–indeed–many patents are “worthless,” it does not mean that the underlying inventions are worthless. That is, inventions have many more roles in a research and innovation economy than simply making money for patent owners. University officials, of all folks, ought to not only understand this point, but ought to be the most vocal advocates for it. Yet, again, they don’t and aren’t. I am waiting for the first colleges and universities to turn away from the unicorn and rainbow chase for windfall profits from monopoly patent positions and get back to the fundamental business of making connections between the university research community and community practice, industry practice, government practice. Those connections are only rarely ones involving royalty-bearing patent licenses.  There is absolutely no good reason to dedicate the entire output of faculty research scholarship to serving up a platter of undercooked patents to a speculative investment community on the proposition that the only thing to be done is to find a way to make money on the pretense of commercial products for public benefit.

Yes, we are happy with commercial products, and with standards, and with libraries of stuff readily available for use, and with stuff that illustrates a research direction and little more. And commercial products may have a public benefit, and they may not. Making money selling product is not necessarily a public good.  Certainly simply trading on the idea of making money for the university–which is what most university patent licensing is these days–is not a clear public good, no matter what the university says the money is needed for (and few universities actually say anything about how they use what money they do bring in from patent licensing–it goes into an administrative slosh bucket unless it’s a whole lot, and then it builds a building for more research–never does it go to offset, say, tuition). Worse, trading on the idea of making money from monopoly patent positions because the reputation of the university suggests that the invention is more reliable than others, or “broader,” or that doing the license is wholesome–how is that a public good, when it generally means that the university has put the public’s trust and goodwill on the line simply to try to coax, lure, or bully companies into taking licenses rather than be sued, and if not companies, then investors, and if not investors, then entrepreneurs, and if not entrepreneurs, then turn faculty into entrepreneurs, and if not them, then grad students, anyone. How is that a public good?

Not only are most patents worthless in university hands, but the fact of the patent + institutional control makes a lot of inventions worthless, too. Invention value is not something intrinsic that can be discovered by careful reading of an invention disclosure report. Invention value happens based on the social conditions under which a new insight is put into play, including who leads the effort, who helps them or doesn’t, and what they hope to accomplish.

We need, then, a study like the one that Professor Moore has done, but we need more years, and we need to see where universities stand in all of this. Then we can start to uncover the difference between a patent made worthless, and an invention made worthless by patenting. A recent article in Nature suggests that more than 95% of university-held patents go unlicensed. That’s about the same rate that advocates for the Bayh-Dole Act railed about in criticizing the federal government’s handling of inventions. So, where’s the outrage now?


This entry was posted in Bayh-Dole, Commons, History, Metrics, Policy. Bookmark the permalink.