Patent reform

The latest incarnation of patent reform would change our system from a first-to-invent system to a first-inventor-to-file system. It should be possible to understand the implications of this system, because it is a system that exists elsewhere in the world. So some bright lights might say that we should have a first-to-file system in order to coordinate with the rest of the world. Perhaps this is true . . . in this case. However, coordinating with the rest of the world may be a bad idea in general. For example, the USA has been going it alone pretty much with respect to software and business methods patents until the recent Bilski case. In the Bilski case, the Supreme Court reaffirmed the machine (no mental processes) or transformation test. It also opined that there may be other valid tests but declined to indicate what those tests might be. The bottom line is that the machine or transformation test harking back to inventions of the 19th century, appears to be the only current test. How many serious, non-trivial or non-obvious inventions in the areas of software or business methods can meet this test? Few, if any.

Let’s digress to discuss why such patents would be socially valuable. In general, economists who are pro-patent will argue that giving inventors some degree of market power for a limited time through patents will generate more invention. This is an empirical question which I would leave for others to investigate and debate. There is, in fact, another reason . . . a more powerful reason in my estimation. Without a patent, inventors will seek to protect their invention through the use of trade secrets. This is costly, uncertain, and generates uneconomic behavior: industrial espionage and reverse engineering. The use of trade secrets also deprives researchers of the very methodologies which should be studied in order to evaluate and improve the methods.

The patent office has been stung in recent years by the criticism that they have approved a raft of trivial and obvious patents in the software and business methods area. The solution to that is to stop approving trivial and obvious patents. Well, you need a test for that I suppose. The test needs to be consistent with patent law. Here is the test. If an expert in the field would need to read the claim in a patent in order to functionally reproduce it after being informed of the high-level purpose of the claim, then it is not trivial and it is not obvious. For example, I and countless others thought of the Cybersource “patent” before they did . . . but who would have the gall to try to patent such a thing (look it up). On the other hand, the Schrader auction patent application that failed to make it was profound and certainly non-obvious (look it up).

Perhaps I should try to patent this test of patentability. Alternatively, let’s include the inventions of the information age in the information age. Let’s do it through legislation. Patent reform ought to affirm that software and business methods inventions are legitimate subjects of patents. Of course, they should not be trivial or obvious. Of course, they should have a practical purpose.

For those who say that we would not be coordinating with the rest of the world if we went in this direction, I would say “good for us.” Let all those who have profound inventions in the software and business methods areas come to the USA for your patent. The market is big enough here for such patents to have an impact worldwide even though the subject matter is not patentable everywhere. That is, who wants to develop a product that rips off a USA patent and cannot be sold in the USA? Obviously, some will do so, but others will resist.