A repeated argument regarding inventions made with federal support was that the public would benefit from these inventions only if companies invested substantial amounts of private capital in developing the inventions as commercial products. Without commercial development at private expense, the public would not benefit from these inventions. They would “sit on the shelf.” And if these inventions sat on the shelf, America would lose its leadership in technological innovation.
The argument went further. The cost to develop commercial products was greater by far than the cost to do the research that led to any particular invention. For any company to commit that sort of money to commercial development, the company had to have a patent monopoly to ensure that it could recover its vast investment and make a profit. Without a monopoly, no investment would be made, no product would be developed, the public would not benefit, and America would lose its leadership position.
As writers repeatedly put it, as if stating a general truth: “What is available to all will be developed by none.” Thus, the federal government’s practice of releasing freely the inventions it owned (“dedication”) or licensing the inventions non-exclusively, royalty-free where the government obtained patents destroyed the opportunity for a patent monopoly and with it the reason to invest in developing the government’s inventions.
A corollary of this argument has been that researchers hosted by universities working with government support also should not be allowed to release their inventions and discoveries to the public domain or convey those inventions to the federal government to be dedicated to the public. Such actions also destroy the interest that any company might have in an invention that could be protected by a patent monopoly and thus become a desirable foundation for a commercial product that might benefit the public. Such actions prevent the patent system from securing for commercial firms the protection they need to justify making new products at great risk and expense. And if commercial firms don’t make products, the research is all but wasted, and certainly the value of patents on inventions made in that research is squandered.
A second corollary of this argument has been that inventors on their own lack the resources and sophistication to undertake the patenting of their inventions. They hardly even know what might be patentable, let alone how to select a patent attorney, devise a strategy for patent prosecution, identify potential licensees, select one and negotiate an exclusive patent license, knowing the details of the company’s business well enough to secure a fair royalty, require diligence, and be able to control the relationship if the company underpays or drags out the development or produces product with inconsistent quality. In one sense, it is true that inventors generally cannot do all these things.
Inventors then must rely on skilled professionals to do the patenting, marketing, negotiation, and contract management. As well they should, because their skill lies in discovery, not law or business, and the public most benefits if they keep their focus on doing research and allow others to take over the complex effort to translate inventions into patents into exclusive licenses into commercialized product into public benefit into American technological leadership into confirming genius federal policy that everyone can take credit for playing their role in turning into a roaring success.
This argument is pernicious. In part it’s bluff. In part it’s only partly true. But it isn’t true in general, and isn’t true across a vast sweep of history, and doesn’t even make logical sense. But the argument is not made to state a general truth about discovery or invention or the public benefit from publicly supported research. The argument is political, made generally by company lawyers and lobbying groups for university administrators. As such, we might expect these folks to “put the best face” on a position they want to hold. Whatever works. There won’t be any introspection about other possibilities, no qualifications that the argument goes only so far. In politics, that’s for someone else to do–“do your best to stop us.”
Thus, we have Bayh-Dole, a federal law created by folks who, for a brief lucky moment, did not get stopped. What they wanted became law. We might then ask, what then did they want? Given their repeated assurance that Bayh-Dole has been a wild success, obviously they wanted was a pipeline that pushed all new drug candidates discovered at universities with federal support to pharmaceutical companies as patent monopolies. They wanted that pipeline to be broader than just NIH funding. Any government funding. They wanted no accountability, impossible enforcement, and everything kept a government secret. They wanted monopoly pricing. They pretty much got everything they wanted. University administrators played the willing village loons to hand it to them, even when Bayh-Dole didn’t exactly connect all the legal dots.
Despite the clear evidence that the argument for patent monopolies has carried the day in federal contracting for research, it is worth looking at what else is possible–even what is true. Just because company lawyers worked their best argument to policy makers, and university administrators were willing to be complicit for the crumbs on offer does not mean that anyone else has to accept the argument, or try to improve it, or even to bother to rebut it. But here at Research Enterprise, we tilt at our windmills, bird killers that they are.
It is true that if inventions “sit on a shelf,” then they aren’t being used. And it is true that inventors, like most of us, could hardly be expected to be adept at patenting and licensing inventions. And it is also true that universities have more resources available to them than do inventors personally. But it’s not true that inventions that aren’t held as patent monopolies are never used or developed. Nor is it true that inventions must be developed into commercial products for the public to benefit from them. Nor is it true that inventors must master the complexities of licensing for product commercialization in order to encourage the use of their inventions. Nor is it true that companies invest in developing new products only when they can first secure a patent monopoly that excludes all others.
Advocates for Bayh-Dole argue for the use of the patent system and in particular for the importance of offering exclusive licenses to inventions made with federal support. But Bayh-Dole has been a dismal failure. One might have seen it coming, given the Harbridge House report (1968) that found that licensing of inventions performed more poorly than ownership of inventions. From the outset, Bayh-Dole’s emphasis on exclusive licensing started on the wrong foundation for using the patent system to set up a monopoly interest in inventions. Really, who should care whether the federal government or university patent brokers do a better job with exclusive licensing, when exclusive licensing itself is a generally lousy way to encourage development of an invention?
It is true that in the wake of Bayh-Dole, university administrators have built a substantial apparatus to take possession of all university inventions (patentable or not) and push these inventions into a system of patenting (at substantial expense), placing them behind a licensing paywall, where the vast majority are never licensed, and when licensed rarely are developed as commercial products, and even if eventually product is sold, in almost all cases not in sufficient volume to serve much public benefit or make a speck of economic difference.
Every university has its cheese shop, and has it stocked with cheese, even, but it cannot find its way to sell its cheese. This failure is called the “valley of death” and the cause of the failure is attributed to the lack of available funding to develop each invention, even with patent “protection” to the point that private risk capital finds the invention sufficiently attractive to undertake producing a commercial product. There must be “bridge” funding or “seed” funding or “pre-seed” funding to develop each invention to the point that major investors or companies take an interest. University administrators then create paper companies and place the inventions in these companies with exclusive licenses, and then use state and government funding (and sometimes their own funds, suitably laundered) to continue development of an invention within this company shell.
While this effort produces the appearance of innovation and economic potential just ripe to happen, along with all the anticipatory press releases, human interest stories, and inventor-entrepreneur of the year awards, it does not do much at all to improve the commercialization outcomes. University administrators might induce a state government to allocate money for another research building, or persuade investors to buy into university bonds, but the public doesn’t benefit from the use of a new product, nor does the activity contribute to American technology leadership.
Despite the limitations, this argument is seductive, especially if it sweeps by as rhetoric that no one bothers to check. It sounds true, especially if said with conviction, repeatedly.
Let me illustrate the argument to show some of what it leaves out and how it bends its logic. Imagine if you will that a university faculty member makes a proposal for research that a federal agency agrees to support. We can represent the funding over time as a blue rectangle. There it is, perhaps a year of funding for $100,000, say. Nothing much. Orders of magnitude below what a big company could even afford to think about. Kinda like discussing the employment of an additional senior office assistant in a company that employs 20,000 people. Now there’s OMG an invention made in this research, occupying some portion of that $100,000–likely not all of it. The cost to the university to identify that invention, document it, take ownership of it, and file one or more initial patent applications can easily run $20,000–and the expense in marketing the invention and negotiating with a licensee will cost the university even more, say $10,000 in time and legal review. Put the patent and administration cost in red.
The university considers its expenditure in red to be “an investment” in the development of the invention. It is in its way thought of as “seed funding” made in anticipation that someone, an exclusive licensee, will be happy that such an investment was made on its behalf. This $30,000 the university looks to recover from any licensing deal–the patent work as a cash reimbursement and the administrative and legal time as an expense to be deducted from royalty payments, which are almost always structured to include a “license issue fee” that covers much of this cost upfront. These charges, then, form the minimum amount of money that a licensee must have to gain rights to the invention.
This is the problem that university administrators have set for themselves. Somehow the invention, the patent application, and the university’s own sunk administrative and legal costs must be sufficiently attractive to induce a company to make a huge development investment, frequently taking a number of years and running many millions of dollars, at some substantial risk that the invention works as claimed, the patent will issue with meaningful claims, that other inventions won’t make the licensed invention obsolete, and that there will be a meaningful market for the invention once a commercial product has been developed. The minimum the university needs to see in the deal is the recovery of its costs. No little wonder that a “valley of death” springs into existence whenever a university decides on this management path:
Now look at it from the company perspective. The pharmaceutical industry claims the development cost is somewhere north of $800 million per drug. Who knows how they come to this number, but it is clear that they intend it to be a really big number. Here’s a depiction of the argument, so far, without attempting anything exact with regard to scale:
The research funding is nothing. The university’s invention costs are a blip. The development investment is substantial, but if the “size of the prize” is huge, and the likelihood of getting a big slice of that prize is decent, then perhaps it makes sense to dive in and take the license, make the development investment, and pay the university its royalty–that 1% or so out of the return.
All this at least seems remotely plausible. The government’s grant funding is just a tiny start to things. The invention is “early stage”–there are years of company development ahead before there will be a product, millions more spent getting things to work, to be manufactured properly, to have the features customers will want, to put product into a distribution channel, train sales reps and dealers, set up for product support, mitigate liability, and market the hell out of it. This is the corporate view presented in the debates on government invention policy. One chosen company will spend a huge amount to produce a product, for which there had better be an even huger return on investment.
The argument can spread itself out. One successful product must pay for many failed efforts. So a company needs more than to recover only its costs of development and pay the university its droppings–the company has to earn enough with its winners to pay for many products in development, most of which will fail for some reason–the technology doesn’t work, there’s blocking IP, the product is too complicated, the manufacturing will be too expensive, the market is changing, someone already has product there, the company has changed direction, the key personnel have left–whatever–there are scores of reasons why a product development effort fails.
This spread out argument makes the patent monopoly even more important–the candidate company prepared to make a development investment can’t promise anything but absolutely must have a monopoly position for the life of the patent to even consider making a risk-filled investment in the hope of recovering not merely that investment plus a reasonable profit, but with a profit sufficient to pay for a bunch of other failures. One might see that if the university licensed to a company that had only a few products in development and made it a point of bringing most of those to market, the need for a huge, huge profit would be less. But the argument that’s made doesn’t go there–it is an argument made on behalf of corporations that take in many things and fail with most, and must have an operating model that permits them to continue to fail with most things–and in doing so preclude universities from licensing inventions to companies with less audacious goals.
Next, let’s look at how this plays out for the pharmaceutical industry.