I'm in NYC for the IP Scholars Conference (IPSC) at Cardozo Law, and one of the (many) article presentations I'm really looking forward to is Patent Invalidity Versus Noninfringement (forthcoming in the Cornell Law Review), by Roger Ford (Bigelow Fellow at Chicago, @rford). Ford argues that patent defendants are more likely to argue noninfringement than invalidity, and that this imbalance exacerbates the problem of bad patents. This article is well worth a look, both because it is a fun read (peppered with examples from actual patents) and, more importantly, because the basic insight seems right.
Ford explains that patent defendants face at least three tradeoffs that often cause them to focus on either invalidity or noninfringement, rather than contesting both issues: (1) the choice between broad claim constructions to support invalidity arguments or narrow constructions to support noninfringement arguments; (2) the desire for a simple trial narrative (rather than "Our product is fundamentally different from this thing the plaintiff invented—which, by the way, wasn't actually new, but had been invented before by this other inventor, or would have been obvious to any idiot in the field."); and (3) resource constraints. Given these tradeoffs, defendants might choose to focus on whichever defense has stronger merits, but Ford notes three asymmetries that will tend to push the choice toward noninfringement: (1) noninfringement has a lower standard of proof; (2) defendants often have a comparative informational advantage on noninfringement (because they know about their own products) and a disadvantage on invalidity (because they may have less experience with the patents at issue); and (3) defendants do not internalize all the benefits of success on invalidity (such as the ability of their competitors to freely enter the market).
As Ford notes, many others have recognized that success on invalidity is a public good, including Gideon Parchomovsky and Alex Stein in another excellent forthcoming article, Intellectual Property Defenses (Columbia Law Review). They explain that invalidity is a "general" defense, the "inverse of a right in rem" in that it "negates the [patent] right vis-a-vis the rest of the world," whereas noninfringement is an "individualized" defense that "is limited to the case at bar." (Their taxonomy of defenses, which includes all types of IP, also includes a third category of "class" defenses such as fair use in copyright and trademark.) But Ford goes further than prior literature in analyzing other asymmetries between patent invalidity and noninfringement, and in crisply explaining why a patent defendant might choose just one.
So if the imbalance between noninfringement and invalidity is a problem, what should we do about it? Ford suggests three solutions: (1) eliminating the elevated burden of proof for invalidity; (2) bifurcating invalidity and noninfringement; and (3) allowing defendants who succeed on invalidity to bring an action for an accounting against industry competitors for the expected damages the competitor would have paid if sued for infringing the patent. Other scholars have proposed bounties for those who successfully invalidate patents, but Ford is (to my knowledge) the first to suggest that the reward be paid by industry competitors. Parchomovsky and Stein have a different proposal targeting competitors: they would allow defendants challenging validity to implead other potential defendants, and competitors that declined this invitation either would be precluded from challenging the validity of those patents in future suits or would have to share the cost of raising the defense if it is successful. Ford argues that this might not "go far enough to make invalidity arguments attractive," while Parchomovsky and Stein counter that their impleading solution, which enlists the help of other defendants early on, would be more help than Ford's accounting action to defendants who face liquidity constraints in mounting a successful defense.
The asymmetry between invalidity and noninfringement defenses—and what to do about it—seems ripe for empirical evaluation. To start, it would be useful to measure the extent of the problem. Basic statistics on how often defendants actually contest invalidity and noninfringement are probably readily available, but it will take more effort to determine how these data relate to any asymmetries in the underlying merits. Empirical evidence should also be used to evaluate proposed solutions—and there are many types of evidence that could be useful. For example, work by Dave Schwarz and Chris Seaman with mock juries suggests that the standard of proof for patent invalidity matters, and also that Microsoft-v.-i4i may in fact have mitigated the asymmetry Ford worries about. Comparing the bifurcated patent litigation system in Germany with similar non-bifurcated systems may help determine whether bifurcation alleviates the problem Ford describes (and how it fares along other dimensions). Data on whether patent defendants in fact face significant liquidity constraints could be used to evaluate Parchomovsky and Stein's argument that this factor favors impleading other defendants rather than suing them later. Careful interviews with attorneys about patent trial strategy could also be used to understand the root causes of the asymmetry. Finally, as I note in Patent Experimentalism (the paper I'm presenting at IPSC this week), the most rigorous evaluation of procedural changes to patent litigation (such as any of these proposals) would be to divide appellate patent jurisdiction and vary rules across courts, along the lines suggested by Craig Nard and John Duffy.
In sum, the asymmetry between noninfringement and invalidity seems like a problem worth worrying about, and I hope that these great contributions by Ford and by Parchomovsky and Stein spark discussion about the extent of this problem and the most effective solutions.