I answered a Quora question a while back. I thought I’d repost it here. The question was:
Just over 30 years old, the Bayh-Dole Act has set the path for most research universities and commercialization, yet few seem to really generate substantial revenue through its frameworks.
Gerald Barnett, IP specialist with major universities, 20 years dealing with BD
It’s a difficult question to answer. Here’s why. First, Bayh-Dole gutted public oversight that previously came with federally funded research. Reports on the status of inventions made with federal support claimed by universities (“subject inventions”) were made optional. A federal agency had to expressly ask for reports. And even then, Bayh-Dole exempts the reports from public disclosure law, so while the federal government might have the data on how all those subject inventions are doing, the public does not have access to that data, and universities don’t publish it. One might be able to get status data from public universities under state public disclosure laws, but expect a costly fight to do it.
So there are no real data available.
Second, Bayh-Dole does not mandate either patenting or commercialization. A university has to decide that it wants to patent a subject invention and then has to acquire rights from the inventors. Only then is it obligated under Bayh-Dole’s standard patent rights clause to file a patent application and to promote the “practical application” of the invention. Practical application is defined to mean (i) the invention is being used so that it is demonstrable that the benefits of the invention are (ii) available to the public (iii) on reasonable terms. Success would mean that each of these points in the definition is met for lots and lots of subject inventions. But we don’t know—but I would expect that if there *were* lots and lots of subject inventions meeting the definition of practical application, universities would be crowing about it. And they aren’t.
For all that, patents, licenses, and startups don’t have anything much to do with practical application. Some inventions can be used without any need for patents at all—TCP/IP, for instance, or most methods. Others can be used without any need for a company to make a product version. Many disease assays are this way—any good lab medicine operation can do its own tests without buying a kit from a company. And even if patents help—for instance, in defining and policing an industry standard—there’s no requirement that they must be licensed exclusively or generate huge royalties.
So the standard of success for Bayh-Dole has nothing directly to do with either commercial products or royalties—even if university licensing offices announce these as *their* goals.
Third, the list of objectives for Bayh-Dole that Congress placed in patent law (see 35 USC 200) is difficult to evaluate. How do we determine how well universities are collaborating with industry or whether university patent positions are having an adverse effect on free competition or future research? Happiness surveys, perhaps, but it’s going to be a soft answer spun from however the questions are asked. The only segment of industry that appears happy with Bayh-Dole is pharma.
Fourth, Bayh-Dole has never been properly implemented by universities, who refuse to comply with key parts of the standard patent rights clause. Instead, they have exploited the lack of public oversight to turn Bayh-Dole into a tool to claim the inventive personal property of everyone associated with university work—faculty, staff, students, visitors, collaborators, volunteers. Their patent policies have become garbled messes with overreaching demands for ownership and implied charges of misconduct for anyone who resists. Advocates for Bayh-Dole have taken pride that they pushed inventors to the bottom of the pecking order for rights. Is the growth of inventor loathing a success?
Fifth, the existence of Bayh-Dole and its standard patent rights clause does not mean that Bayh-Dole has been a net positive. Inventions may come into use despite university administrators insisting on having their thumb in every opportunity. So even counting up the “success stories” does not mean Bayh-Dole and its bureaucracy contributed—it could be that the success would have been even greater, or sooner, had the university stayed out of the patenting (which it can do under Bayh-Dole). Coincidence is not causation.
Although Bayh-Dole was passed on the strength of the network of external invention management agents that were doing most of the patenting and licensing before Bayh-Dole, university patent administrators set about to dismantle this network and replace it with overloaded, bottlenecked, institution-first bureaucracies. In this sense, Bayh-Dole has been an absolute disaster, destroying what we had that was working (somewhat) and replacing it with institutional bureaucracies that don’t work well together, raise the overhead of negotiating contracts, and don’t do a particularly good job with all the stuff they’ve claimed.
In short, we don’t know, we still get by, and university bureaucracies have gotten bigger, and try to put a glossy spin on what they do
For another take, consider Bayh-Dole, the franken-sausage god.
Universities claimed a licensing success rate of 25% to 30% in the run-up to Bayh-Dole. IPAs had a success rate of 4% in 1981. University of California reported internally that they had a 0.5% success rate. At Stanford, 287 out of 6,400 inventions in 36 years earned enough ($100K+ cumulatively) to suggest they might have involved, even briefly, commercial products–4.5%.