The use of the patent system for federal research results, 6: The seeming middle ground

While the FSA policy makes what appears to be nice gestures–royalty-free licensing or at least licensing without unreasonable restrictions and without excessive royalties–there’s little here to provide guidance so far. The policy continues, looking at the prong in which the agency director makes a determination whether to allow the grantee to retain rights:

If he finds that there is adequate assurance that the invention will either be effectively dedicated to the public, or that any patent which may be obtained thereunder will be generally available for royalty-free and nonexclusive licensing, the effectuation of these results may be left to the grantee.

Here, Possibility 1 is first indicated–nonnprofit open access. We have it rather clearly that “royalty-free” is within the scope of not “excessive” royalty, and that “nonexclusive” meets the expectation for not “unreasonable” restrictions. That is, if the grantee hasn’t gotten the right to retain title to inventions upfront, then an agency head can allow the grantee to retain title anyway, provided the grantee intends to make the invention available nonexclusively, and royalty free. Either the government does this, or the grantee for some reason does this. Doesn’t much matter.

The compromise word in all of this is “generally”–“will be generally available for royalty-free and nonexclusive licensing.” That is, the default is royalty-free, nonexclusive, but “in some cases” patent exclusion practices may be “necessary” to “foster an adequate commercial development to make a new invention widely available.” This compromise opens up 110-1 to exclusive licensing, suppression of competition (and collaboration), and patent speculation. Of course, that’s not put forward expressly, but such things can be construed as within the scope of “generally.”

Another way of looking at FSA 110-1, however, is that it draws a distinction between utilization and a ramp up for mass production and widespread distribution. In this distinction, making and using a given invention locally in various forms is still available to all, and it is the right to make and sell in a special form required for mass production and national distribution that comes into play under a contractor’s patent right. Even here, the right to make and sell is not absolute–it’s the right to make and sell in mass produced, nationally distributed form where that form is necessary to mass production and distribution, and to make that special form takes lots of money and involves some risk that that special form is not possible or economically viable. The distinction then in 110-1 is between two forms of making and selling–for local use, DIY, low production volumes that can meet specific needs, perhaps in custom ways, and for national use, mass produced and distributed in a version that might meet a general consumer’s needs. The “necessary” bit then in the policy covers not the “development” of an invention into any useable form, but the “development” of a form of the invention suited to mass production and widespread distribution where no such effort has been made.

In effect, then, a patent holder might, in 110-1’s view, be allowed to license a mass-produced instantiation of an invention that takes money and risk to create, and this instantiation (or class of products) then requires a special license endorsement, so that companies that failed to take a free non-exclusive license when they could have, or got that license and dinked around locally having a good time but didn’t get around to developing a mass-produced form suited to mass distribution, then can be made to wait while companies (one, or more than one–even better) that are willing to make the investment at risk to develop a mass-production version of the invention have the right to do so without a bunch of free-riding previously unambitious companies tagging along without paying their way or bringing anything to the effort. It’s the mass-market version of an invention–where that version requires expensive or difficult development for mass production and widespread public use–that’s the subject of necessary exclusive or “co-exclusive” patent licensing. The players that want to operate nationally get this added license, but that does not affect local making and using and selling, does not affect research and scientific uses, does not affect public access for DIY implementations.

You won’t see this distinction laid out in  discussions of federal patent policy, not FSA 110-1, not of the Institutional Patent Agreements, not of Bayh-Dole. In those discussions, it’s always about exclusive patent rights, without distinctions. Policy and patent folks find it difficult to recognize a non-exclusive, open access program running quite happily underneath a occasional, conditional, limited exclusive licensing program for national scale up investments producing extensively or with some difficulty “developed” versions of a given invention. And what they don’t get is that a happy non-exclusive open access program may serve as the basis for justifying a later scale up. The open access program demonstrates utility, establishes ad hoc standards, explores variations and applications, and mixes and matches technology into cumulative platforms and libraries. Call it “social proof” that people are using the technology. Seeing others use a technology to their advantage is a pretty good marketing pitch, a network effect. If people don’t adopt because they don’t want an unbuttoned (as it were) product, or they want hundreds or thousands or more and no one can figure a way of scaling up production with just the right features for all those potential users, then maybe a limited patent right ought to come into play to leverage scale up with private funds. Maybe only one company is interested in doing that, or maybe only a handful. No reason to cut off the local use to develop a national version. Do both. But that’s not at all university patent practice, even if at one point that distinction was implicit in federal policy. Arguably, that distinction still operates in Bayh-Dole, though it is spread around in pieces like the scarecrow in The Wizard of Oz after the flying monkeys have had their way with it.

FSA 110-1 contemplates, without making it express, a non-exclusive license to make, use, and sell, with a time-limited conditional clause that can be invoked by a patent holder if a licensee has not, within that time (say, three years), undertaken to mass produce and sell nationally (to The Public, not just to some bit of it) product or services based on the invention. When such a conditional clause is invoked, the license of non-exclusive licensees is affected only as to the scope of their right to sell–they can keep selling as they have been, if at all, in the markets they have got, but they cannot (for some time, if the patent holder grants one company an exclusive license) mass produce and mass distribute. A patent holder might also offer non-exclusive licensees the right to upgrade their sell right to national mass production and distribution, conditioned on timely exploitation of such a right (and no doubt paying some fee or a royalty on sales).

Again, this division of rights allows anyone to make, use, and sell for a local market and only restricts scale up for licensees that have shown no interest or capability in such scale up. To restrict licenses to make and sell only in a product version that must be developed for mass production and widespread distribution is what’s implied in the “necessary” in 110-1’s policy. It would not be “necessary” to suppress all making, using, and selling of an invention in favor of mass production and distribution if no meaningful changes have to be made to the invention for such a thing to come about. In such a circumstance, the problem is not development of the invention but having production capacity, sales force, and distribution resources in place. These are normal business issues, not ones to solve by denying everyone assess to an invention while some company fumbles to get such things in place in favor of other companies that already can mass produce and distribute–even if the fumbly company and its investors are willing to pay to exclude all others.

This distinction in the two different scopes for making and selling is at the core of the “in some cases” and “generally” qualifiers in 110-1. But this implicit distinction gets lost in the FSA drafting, overwhelmed by a range of other patent practices that do not depend on “in some cases” or “generally” or “necessary” and come into play because they are not expressly forbidden. These are practices that, were they expressly authorized, many people would reject them as stupid, exploitative, counter to the use of public funds, and immoral with regard to matters of public health. You might make good money investing in companies that sell babies, but we (sigh, generally) refuse to live that way.

This entry was posted in Open Source, Patents, Policy and tagged , , , . Bookmark the permalink.