Prof. Doug Lichtman, UCLA School of Law
December 1, 2009
The United States Supreme Court recently heard oral argument in a patent case called In Re Bilski. At issue is the question of under what conditions an innovative process falls within the subject matter of patent law. On one view, some processes should be excluded from the patent system – for instance, because they involve no physical real-world step. On another view, every process should be welcome in the patent regime, at least so long as the process otherwise satisfies patent law’s requirements of usefulness, novelty, non-obviousness, and the like.
The Bilski case has generated an enormous wealth of commentary, including eighty-odd amicus briefs, dozens of thoughtful articles, and hundreds of blog posts, CLE seminars, and the like. Here, however, I want to add four thoughts; three on the substance of the Bilski case, and one on the Supreme Court oral argument itself.
First, Section 101 of the Patent Act tells us that the patent system is open to anyone who “invents or discovers any new and useful process, machine, manufacture, or composition of matter.” Yet, as we sit here today, it is clear that no one is ready to read that language literally. That is, literally, the word “process” would sweep in almost any set of steps conceived by man; and, were we starting from scratch, I think there would be a good argument to say that process should be read in exactly that broad fashion. The rest of patent law, after all, could do the hard work of weeding out inventions that don’t fit, and, at that, weeding them out for specific reasons (like obviousness) rather than weeding them out based on tangentially related characteristics (like “does it transform matter?”) or potentially overbroad categorizations (like “is it a business method?”).
Again, however, that ship has sailed, in large part because the Supreme Court long ago made clear that – whatever it means – the word “process” does not include laws of nature, abstract ideas, or mental processes. The Bilski fight is therefore not a fight about whether “process” is a complete catch-all. It’s not.
The fight is instead about whether the word should be narrowed further still; with the petitioner basically wanting to stop at those three exclusions and leave subject matter flexible beyond that, and the respondent either wanting to add the machine-or-transformation test as yet another exclusionary mechanism, or wanting to use the machine-or-transformation test as a way of identifying which inventions are “abstract” and hence excluded under the old cases.
Second, although I’m not sure that I would use subject matter exclusions to do it, I do think that patent law needs to exclude all sorts of inventions even if they do meet patent law’s other existing requirements. That is, there are a lot of inventions that should not be patent eligible even if they are new, non-obvious, useful, fully enabled by the disclosure in the application, and so on. Some inventions are just too broad, for instance, such that a patent right would enormously disrupt down-stream use and innovation. Other inventions are adequately rewarded by first-mover advantages, or adequately funded through non-patent mechanisms; for those, there is no need for society to suffer the costs of patent protection.
I have no strong view that additional filtering along these lines should be done through interpretations of Section 101’s subject matter rule as opposed to (say) interpretations of Section 112’s enablement requirement or the addition of new eligibility thresholds or even the addition of new rules about patent remedies. But, big picture, I’m sympathetic to the work the Supreme Court is trying to do in these subject matter cases – which I understand to be an effort to roll up the tent a little bit and trim back on the scope of the patent regime.
Third, in light of all that, the Federal Circuit’s machine-or-transformation test looks to me like a bad idea. If the goal is to knock out inventions that are too broad, or to exclude inventions that are adequately encouraged by non-patent rewards, or any such similar thing, the Federal Circuit’s test simply doesn’t resonate. Just because an invention is tied to a machine or involves a physical transformation doesn’t remotely imply that the invention will be comfortably narrow. Conversely, just because an invention is not tied in those ways does not remotely imply that the invention will be too broad.
Add to that the obvious problem of defining what it means to “transform” something, and the obvious difficulty when it comes to defining how tightly an invention must be associated with a machine in order to qualify, and you have a disappointing mess. The machine-or-transformation test is uncertain, hard to implement, and, even if implemented, doesn’t seem like it would serve as a good proxy for any of the public policy concerns that motivate the fight.
That said, I do understand why the Federal Circuit thought it needed to follow this path. The Supreme Court itself planted these seeds. But, ideally, the Court will now use the Bilski case to right the ship, tossing the machine-or-transformation language and in its place offering a clearer explanation for why some inventions must be excluded from the patent regime. After that, sure, maybe a test will emerge that can serve as a proxy for the relevant policy concern.
And indeed it’s not crazy to imagine that some subject matter categories, like business method patents, might match those policy concerns close enough that the category itself can be safely excluded. But, for now, I think it would be a huge step forward were the Court simply to articulate the touchstone for why we are excluding abstract ideas, or why we might want to exclude non-transforming processes; and, armed with that explanation, the Federal Circuit might then be better situated to craft detailed rules that would champion the relevant concern.
Lastly, aside from the substance, I do think it worth pausing a moment to think about whether the overall judicial process here worked. Patent law, after all, has an unusual structure in that all patent appeals are focused at the Federal Circuit rather than being distributed by geography across the various numbered circuits. The intuition behind that design was that patent law would benefit from having a specialized appellate body; thus there is always a bit of trepidation when the unspecialized Supreme Court steps in.
In this case, though, I think the Court thus far has shown that it can easily keep pace. The oral argument at the Court, it turns out, didn’t sound all that different from the oral argument that was conducted months earlier in front of the Federal Circuit. And, in the Supreme Court’s back-and-forth, I do think the Court relatively quickly identified and pushed hard on each side’s biggest weakness. That is not to suggest that the argument didn’t have its share of cringe-worthy moments; but, overall, I think it’s fair to say that the generalist Supreme Court looked good in this interaction, and certainly did not seem at a strong disadvantage as compared to the specialized Federal Circuit.
Whether that tells us anything more broadly, though, I’m not sure. For one thing, the subject-matter question might be a lot more accessible to a generalist court than are other serious patent problems. Enablement, for instance, raises complexities that might be hard for non-patent judges to fully appreciate. Moreover, the Federal Circuit’s decision in Bilski might not fully represent what that court can do, given that the Federal Circuit in this case was already working within constraints set out by previous Supreme Court decisions.
Put differently, it might have been that the Federal Circuit’s expertise would have been more significantly on display had the Federal Circuit decided Bilski on a blank slate. It didn’t, though, and indeed one of the most striking things about the Federal Circuit’s decision was how much it abandoned that court’s own precedents and instead quoted, relied on, and parsed what were at that point clearly dated prior Supreme Court decisions. So the comparison here might not be representative; the Supreme Court might have been working on an issue particularly well-suited to a generalist bench, and the Federal Circuit might have been playing with one hand tied behind its back anyway.