The bogus argument for “mixing” research funds, 4

The origins of the argument for “mixing” government and private research funds can be found in the 1968 Harbridge House report. The report identifies six industry attitudes toward patenting, ranging from indifference to defensive positions to critical to business. The special special case selects this last attitude, taken from the pharmaceutical industry and directed at benefits to public health, and demands that it be applied to all industries. The bogus argument for mixing does much the same thing. Here are the five industry attitudes (six, really), with Harbridge House’s comment (with minor editing):

(1) Patents have no importance

When these firms obtain patents under government contracts, their sole purpose is recognition of technical competence within the company.

(2) Patents are of little value, compared with technical know-how

Using a new idea to enhance product performance is regarded as more important than assuring that the company owns the exclusive right to it.

(3) Patents are valuable for defensive purposes

Ownership of a patent as a prerequisite for new product development is a relatively minor factor with these firms compared with market and investment considerations associated with commercialization of the invention.

A change in government patent policy may affect some firms in this category by causing them to choose more carefully the areas in which they are willing to undertake government research. Faced with the possibility of being unable to obtain title to patents they develop, these firms may refuse to contract in research areas that would impair their operational flexibility.

(4) Patents are important in establishing proprietary positions

Firms having this attitude actively seek ownership of patents to establish and maintain proprietary positions in new technologies as well as in established product areas. Invariably, however, estimates of market potential and corporate investment requirements determine which product areas are developed.

A change in government policy from license rights to title rights [that is, the government takes title rather than accepts a license] would limit the government-sponsored R&D activity of firms in this category because of conflict with company-sponsored research activities. These companies would examine contract opportunities on an individual basis and, in many cases, might refuse to contract with the government.

(5) Patents are essential to business activities

Firms in this category regard patent rights as essential to their business activities, and are careful to avoid government claims or conflicts over ownership of inventions.

Harbridge House distinguishes two approaches by these category (5) firms:

(5A) In the first pattern, firms will assure corporate ownership of patents before initiating work on a government contract. In these situations, firms deliberately select areas of government research to match their commercial interests in order to generate product ideas with commercial possibilities.

Firms following the first pattern would be severely affected by a change in policy since their business activity is based largely on government-sponsored research that may develop commercial applications. Corporate ownership of patents is, therefore, an essential feature of the growth strategy of such firms. If title to inventions arising from government-sponsored research were to become unavailable, such firms would have to either change their mode of business or refuse to contract with the government. 

(5B) In the second pattern, firms isolate government work from their commercial operations and pursue these activities separately. The technical value of government contracts is rarely considered a valuable supplement to in-house research and development.

Those firms following the second business pattern have no proprietary expectations from government contracts. Any change in government patent policy with respect to license and title rights would have little effect on them since they have already divorced their main corporate interest from government contract work and do not regard government-sponsored R&D as a source of commercial ideas.

I have highlighted the parts of the report’s analysis that suggests that companies might not participate in federally sponsored research over patent issues. Of these five (six, really) attitudes, only the companies in category 5A are of the sort that require patents to justify the development of new products–and those firms require title to inventions and are not satisfied with merely a government license that allows them to practice but also forces them, potentially, to compete with others.

For these companies, it would not matter whether the license comes from the government or from a university–it would have to be an exclusive license free of any possible competition. That is, the exclusive license would have to have the effect of assignment, so that the company has complete control over market access to the invention. This is where the special special case shows up–the university has to do what the Public Health Service refused to do, and that is to assign the invention to a company to suit its operating model.

The other concerns of companies with regard to patenting don’t involve justification or incentives to create products. The factors that matter involve market and investment, but patents aren’t significant in those decisions. Here, then, is the effect of an arbitrary (as in “uniform”) university patent policy and practice. Universities rig to accommodate the patent requirements of the 5A companies by granting exclusive licenses that functionally are assignments, but at the same time the universities refuse to deal with companies that are fine with a non-exclusive license or who want to hold a patent as a defensive measure, available to cross-license or counter sue competitors.

It is odd, too, that universities would not adopt a variation on the category 5B companies–those that isolate their government work from other work. These companies, though they hold patents to be important, don’t do government work in areas that would affect their commercial operations. These 5B companies operate their government work as “contract research organizations”–readily giving up all rights in inventions. A number of the nonprofits that received Institutional Patent Agreement program standing operated as CROs. Most universities also operated in this fashion. They had no ownership interest in faculty inventions, and did not have any reason to assert any ownership interest–it was good enough that research was done.

WARF, MIT, Stanford, and UC changed that approach, with a combination of claims that licensing patents could generate money for research (even though consistently they testified that it did not–and WARF for a long time made its big money investing licensing income in the stock market rather than in licensing)–and that universities could do a better job engaging industry to develop inventions than could the federal government (even though the data showed that the federal government, in biomedical inventions, was doing just as well as universities, and for federally supported inventions managed by universities under the IPA program was doing nearly 5x better than universities).

So, more money for research, except that wasn’t true. And a better job than the federal government, except that wasn’t true either. These things just sounded true, and that was good enough. In bits and pieces there were things that were true–a license here, making some money there. These bits and pieces were true. But the general case wasn’t, and still isn’t.

But now we can see the shape of the argument regarding mixing of funds in companies. A company that valued patent positions would do only non-critical research for the government or would segregate its business operations from its government contracting. Otherwise, these companies–the ones for which patents mattered–would decline government work.

Now look at the Bayh-Dole policy statement (35 USC 200):

to encourage maximum participation of small business firms in federally supported research and development efforts

If some small business firms are category 4 or 5A companies, then these will not accept government research funding if they cannot obtain title to inventions they make in that research. We can see, then, the outline of what will become a bogus general argument, shifted to universities. Bayh-Dole stipulates an arbitrary, government-wide default patent rights clause adopting the “license” approach. The government gets a broad license to inventions, but has no right to title to inventions unless a contractor fails to comply or declines to elect title having obtained title. For small businesses in the other categories–indifferent, little value, CRO–title or license doesn’t matter and Bayh-Dole does nothing for them.

Thus, Bayh-Dole’s purpose is to broaden the small companies the government can fund to include those that will position themselves to use government funding as a subsidy for their own commercial product development. That’s an interesting twist on what otherwise appears to be a law directed at subvention funding–funding as grant-in-aid to develop new science without a commercial motive. One wonders if the entirety of the federal effort to support private research should be shifted to accommodate participation by small companies that view federal support as an offset to commercial development. That, of course, is the premise of the SBIR and STTR programs, but oddly–very oddly–those programs largely fail to produce commercial products. The companies that participate in SBIR and STTR appear to be happier being CRO companies than ones building commercial product portfolios. They attempt commercial products until the government funding runs out, and then they attempt new commercial products with more government funding–and the more government grants they get, the better they are able to compete for more government grants.

Rather than broaden government funding to include a class of small companies that consider patent rights essential, Bayh-Dole created instead a class of small companies that hold patent rights but rarely develop commercial products. There are, of course, small companies that intend to create commercial products–the biotech industry is full of such companies. These companies are almost always investment-backed, and the investors expect that the companies will have proprietary positions, “unfair advantages,” and potential monopolies on the products they seek to create. These companies adopt their patent positions because they work with investors who demand those positions. The effect of Bayh-Dole has been to expand federal funding to be used as a supplement by investors who require companies to value patents highly as 5A category companies. Bayh-Dole exists for speculative investors. That might be fine. But Bayh-Dole goes further. Under the special special case, generalized, Bayh-Dole makes speculative investment the purpose of all federal funding. For anyone else, Bayh-Dole doesn’t much matter.

It’s a strange federal research policy that gives advantages to speculative investment that requires monopolies when many research organizations do not require either speculative investment (investment for a high return and therefore desiring a lucrative but volatile marketplace full of uncertainty) or monopoly patent positions (they compete on other factors such as location, quality, price, brand identity, and customer service). Bayh-Dole, in essence, declares that the default outcome for federal research funding is that it serve as a subsidy to attract speculative investors, and the way to do that is by means of patent monopolies, and the way to do that is to ensure that university administrators will pass through inventions to such investors without government interference. Why not simplify things and move government research funding from universities directly to these companies? Why burden everyone with the inefficiency and financial burdens of negotiating exclusive patent agreements? Why have universities patent everything in sight “just in case” when companies know pretty much what they want and let go everything else (unless they are trolls interested in selling anything or threatening with anything)?

 

This entry was posted in Bayh-Dole, Policy, Sponsored Research. Bookmark the permalink.

One Response to The bogus argument for “mixing” research funds, 4

  1. Pingback: The bogus argument for “mixing” research funds, 3 | Research Enterprise

Comments are closed.