Introducing Commercial Utility

This is the third and final post in a series discussing my forthcoming article Reinventing Usefulness. The other two are here and here. Those posts introduced usefulness and its forgotten history. This post suggests that patents should not issue unless there is evidence that the invention is commercially useful. After the jump, I'll briefly describe what I mean, discuss some practical concerns, and briefly summarize some of the benefits and costs of the proposal. There are many nuances to the proposal that I don't address here, but that are discussed in the article.

Update: John Duffy has written a nice review of this article at Jotwell for those interested.

Defining Commercial Usefulness

The proposed test would find commercial utility present with sufficient evidence to convince a person with skill in the art that a) there is a market for the invention, and that b) the invention can be manufactured at a cost sufficient to fulfill market demand. The test would be applied in a manner similar to that of operable and practical utility, such that expectations justify utility even if the expectations prove incorrect in hindsight. Furthermore, no invention is complete without commercial utility. Thus, if a person with skill in the art would not expect the invention to be manufacturable in commercial quantities as of the filing date,  then the patent would not issue even if the patentee proves skeptics wrong. 

However, because it is based on expectations, commercial usefulness does not require actual commercial production, which might conflict with rules that penalize sales of patented inventions before filing. Likewise, a patent expected to succeed would not be void simply because a product incorporating the patent flops in the market.

The first factor seems straightforward; in order to be commercially useful, some group of people must want to purchase it.  The second factor constrains the result somewhat, only allowing a finding of commercial utility where there is evidence that near-term market demand can be satisfied. The factor bars inventions that will remain so expensive to reproduce for such a long period of time that many who might want to purchase the invention are unable to obtain it.

The doctrinal test is theoretically elegant. It draws on the intersection of supply and demand that leads to a market clearing price. In that sense, commercial utility comes closest to the economic definition of utility, here measured by consumer surplus. Consumer surplus is the difference between the amount consumers are willing to pay for a good and the price they actually pay. The first prong of the test represents the amount consumers are willing to pay, and the second prong represents the price that suppliers are willing to accept.

However, positive consumer surplus cannot replace the doctrinal test suggested here for a couple of reasons. First, supply is traditionally based on marginal cost, but new inventions may require high fixed costs that preclude commercial sales at any price. Second, there will almost always be some consumer surplus so long as one person is willing to pay more than the marginal cost of a good. While the commercial utility test only requires de minimis utility, it must also require both sufficient profit expectation to justify recovery of fixed costs and market demand of more than a handful of people. Otherwise, the test would be rendered illusory.

Practical Concerns

Despite the apparent elegance of the two prong standard, it may prove difficult to apply in practice. First, determining what constitutes a market is difficult. There will almost always be one person willing to pay high costs for new products, and the patentee will always argue that she is willing to supply the product at that high price. Thus, evidence must be objective but it may still be difficult to judge. However, difficulty in applying a standard should not stand in the way of important policy. Obviousness, for example, is notoriously difficult to apply, yet it is one of the more important quality standards available.

Showing consumer demand is further complicated by the train paradox. Proverbial railroad officials must decide if there should be a noon train. To find out, they visit the platform at noon. Surprisingly, there are no customers waiting at noon, and the officials conclude that there is no demand for a noon train. The paradox, of course, is that passengers will not visit the train station if there is no train scheduled.
Similarly, consumers may not know enough about a new product to know that they would be willing to pay for it. This is especially true of some of the most important inventions that change the way we look at the world, such as polypropylene or even the television. 

These practical difficulties in showing demand need not be a detriment to the test. For example, market research can generate important information, and patents currently do not protect investments in such research. The commercial usefulness requirement treats inventions – especially pioneering ones – as incomplete until the threshold is met. Thus, inventors must research and develop inventions to the point that consumers would demand them prior to filing for a patent. This would allow inventors to capture otherwise unprotected value in developing a new market by extending the patent term. 

Furthermore, because the invention is not considered complete until commercial usefulness is known, such market research would be considered pre-completion “experimental use” in order to avoid patent invalidation due to public disclosure of the invention prior to seeking a patent.  By allowing patents despite traditionally invalidating market research, companies will have more of an incentive to conduct such important commercialization activities.

Benefits and Costs

The benefits of a commercial utility requirement are manifold. First, costly and uncertain commercialization efforts would be protected by the resulting patent. Such efforts would not be patentable in themselves, but they would be necessary to obtain a patent. As a result, they would be considered “experimental use,” and thus would not invalidate the patent when done before the patent filing.

Second, the requirement would limit patentability to those inventions more likely to increase consumer welfare by satisfying a public demand.

Third, patent disclosure would be improved, because patentees would know more about the invention and would face more difficulty withholding valuable know-how from patent applications. Inventors are required to disclose the best mode of practicing the invention.  With a commercial utility requirement, the best mode would now include commercial practice of the invention.

Fourth, patent claims would be more concrete, because only commercially useful claims would be patentable. It will be more difficult to show manufacturability and consumer demand for abstract claims; markets do not demand amorphous descriptions – they demand products. This might spur future innovation developing commercializable embodiments of the abstract concepts taught by the patent and minimize the first inventor’s ability to block future products that were not within its initial development.  

However, a commercial utility requirement would create several offsetting costs. For one, commercial utility is difficult to observe, which increases administrative costs.  Such costs should be offset by some reduction in the number of patent applications, presumably because fewer people would seek protection for commercially useless inventions.

Then again, fewer applications are not necessarily a good thing; because most patent applications are published, more filings expand knowledge in the public domain even if no patent results.  These patents might increase public knowledge even if they do not promise commercial benefits.

More important, the standard would thwart the filing of patents that are not yet commercially valuable, but might be in the future.  Of course, this delay is the purpose of the commercial utility requirement, but delay may be costly nonetheless.

The magnitude of the cost of delaying (and potentially eliminating) future-commercializable patent filings depends on a variety of factors. One factor is the extent to which delay increases the cost of inventing and patenting, which reduces the ex ante incentive to invent.  Determining whether this results in a net social cost is a difficult question, because the reduced incentive may primarily eliminate worthless inventions.  It may be optimal to reduce investment in particular types of inventive activity.

Another factor is the extent of independent inventor patenting. Independent inventors often lack commercialization capacity, and might not be able to show commercial utility. If independent inventors contribute important knowledge through patents,  then making such patents unattainable may decrease the benefit of that knowledge. This cost is exacerbated to the extent that small companies must have patent filings to obtain capital investments;  without such funding the number of innovating firms will decrease. If, however, small and independent inventors provide little additional inventive benefits and instead large companies simultaneously develop most inventions,  then discouraging independent invention may be an acceptable cost.

A final factor in the cost of delayed patents is the effect on concurrent races to invent. Delayed patent filings may extend duplicative patent races. If simultaneous research is considered inefficient,  then delaying the patent even further will extend the inefficiency. Further, a commercial utility requirement would exacerbate duplicative costs because the commercialization adds more uncertainty and cost to the process, and only one person reaps benefits from expending such costs.

On the other hand, patent race concerns are overrated for three reasons. First, many races would occur in secret anyway, and to the extent the races are simultaneous, at least eighteen months of time will pass before one inventor learns of a patent filing by another,   if they ever do. Second, in many cases there is no race (though the inventor does not know it), such that benefits are maximized if the law provides an incentive to commercialize. Third, the race may lead to different beneficial solutions, which is good for competition and innovation.

While I believe that the benefits of a new requirement outweigh the costs, I suspect the jury will be out on the question. My goal in proposing this new standard was to take a new and critical look at what it is we demand from the patent system. I believe this proposal achieves this goal if no other.