Want to Improve Patent Quality? Tweak the Incentives for Low-Quality Patents First

One of the biggest issues facing patent prosecution today is the notion of "patent quality." Everyone wants "quality" patents, but no one can agree on how such patents may be defined, nor can anyone agree on the best way to achieve an appropriate level of quality across the board. While there are many proposals to "fix" the process, it may be that, unless more fundamental issues are addressed, many of the existing proposals will be doomed from the start.

Improving patent quality is generally viewed as an administrative problem. In other words, by improving matters such as funding levels, regulatory, judicial and bureaucratic processes, improvements in patent quality are likely to follow. However, this approach largely ignores the point that the modern patent system affirmatively encourages low patent quality - the incentives at work are such that we cannot reasonably expect anything other than very large numbers of low-quality patents

R. Polk Wagner, professor at the University of Pennsylvania Law school, takes a look at these incentives in a new paper titled "Understanding Patent-Quality Mechanisms." In the paper, Wagner suggests that only by understanding the mechanisms of patent quality - the incentive structure that no only discourages "good" patent behavior but also encourages "bad" patent behavior - will we make any real progress in improving the current situation.

Wagner identifies a series of powerful incentives that lead to lower quality patents:

(1) incentives that encourage patentees to draft patent applications that effectively obscure the true scope of the invention and its relationship to the prior art;

(2) incentives that lead the administrative agencies (the USPTO and other patent offices) to conduct relatively ineffective examinations of many patents; and

(3) incentives that compel modern innovative firms to adopt a high-volume, low-quality patenting strategy.

With regard to the first incentive, Wagner properly points out that patent applicants must necessarily be futurists when an application is written. A properly-written application must take into account the potential market(s), technological advancements and alternatives that may (or may not) materialize many years in the future. This timing-based incentive creates strong interests in deferring a careful analysis of the patent (especially with regard to claim scope) for as long as possible, or at least retaining as broad a range of possibilities for as long as one can. This inherently leads to a number of problems:

First, a patentee will almost certainly seek substantial vagueness, thus gaining flexibility to effectively alter the scope and description of the patent according to changing circumstances. Second, because of legal rules which penalize detailed descriptions, the less description the patentee can provide the better she will be. Both of these strategies—creating vagueness and providing a lack of description—have important costs to the patent system. At best, they make it much harder to evaluate the scope and validity of a patent, and make it much more likely that mistakes will be made. In many cases, they allow patentees to exploit the dual-stage-analysis process noted above to obtain a patent under one understanding of the language (e.g., a narrow understanding) and later assert that same patent in a way that broadens the scope of coverage. And, in all cases, they yield patents that are substantially less likely to comport with the statutory standards of validity—that is, low-quality patents.

These problems can be exacerbated by the rules themselves: although 35 U.S.C. § 112 requires clear and adequate disclosure as well as clear and distinct claim language, the USPTO rejects fewer patent applications for § 112 disclosure problems than for prior art problems and virtually never provides a detailed analysis of claim language, meaning that serious § 112 analysis is left for litigation. Also, the statutory presumption of validity encourages patentees to pursue a flexible view of their patent: narrow during prosecution, when the risk of rejection because of prior art is higher, and broader during litigation, when patent scope is of paramount importance. Moreover, the Federal Circuit's treatment of claim construction as a process that interprets claims "in a contextual, holistic manner, without any established process or framework" virtually guarantees that the final analysis of claim scope will only occur after appeal to the Federal Circuit.

With regard to administrative incentives, the PTO has openly taken the position that higher throughput has been a perpetual goal of the office: deny more patents more quickly, and grant more patents more quickly. While the PTO's allowance rate has dropped over the past few years, the number of patents issued has risen as well.

Most importantly, Wagner argues that rational patentees adopt a high-volume, low-quality patenting strategy because it maximizes the possibility of gaining an advantage from the current patent system. Such a strategy can be advantageous in the modern innovation economy for a number of reasons: having many patents, even if their quality is low, can (1) provide much-needed marketplace power in a world where individual patents become increasingly less certain in scope and validity; (2) hedge against the difficulties in predicting the future by casting a broader net - with many patents in a particular field, less emphasis is placed on the need for any individual patent to endure into the future; and (3) hedge against changes in the law itself.

Using this strategy, patentees are often encouraged to seek more patents than necessary, but expend fewer resources on each patent. Also many patentees file for defensive purposes, or file patents only as signals of innovative acumen - in such cases, the bare minimum required to squeak through the PTO would be sufficient. Thus, reliance on the PTO to help improve quality would be misplaced:

The basic problem with viewing patent quality as an administrative issue is that, given the incentives, there is relatively little that the USPTO (or any patent office) can do to change them. First, as long as the balance of interests tips in favor of high-volume, low-quality patenting behavior, we should expect to see those sorts of patents irrespective of the administrative process used . . . There are real limits to how much quality the USPTO can “add” to a filed application: it can reject claims or request revisions, but the power of language remains in the patentee’s hands. In patent prosecution, it will often be a case of “garbage-in, garbage-out”: if the filed application is low quality, the chances are that any granted patent will be low quality as well.

Wagner's paper goes on to examine various proposals (increasing examiner headcount, improving search tools, public access to prosecution, post grant review) and weighs the pros and cons of each proposal. The paper also proposes some alternate proposals:

(1) encouraging early clarity by having the PTO conduct meaningful claim construction analysis with the applicant during examination,

(2) impose further "penalties" on significant amendments made during prosecution,

(3) weakening the presumption of validity,

(4) set yearly quotas or tradable patent application rights, and

(5) impose a PTO "quality panel" that randomly selects patents to evaluate their validity, and apply penalties for invalid patents on that basis.

Of course, the controversial nature of these proposals are not lost on Wagner, and he acknowledges "that patent quality is far from an easy problem to fix, and in many cases the cures may be worse than the problem itself . . . I ultimately conclude that there is no easy answer or simple fix. But at least if we understand the problem fully—here, the mechanisms that
underlie low-quality patents—we can begin to address it, and, more importantly, avoid making changes that will only make matters worse."

- Read/download R. Polk Wagner, "Understanding Patent Quality Mechanisms" (link)

- See also, R. Polk Wagner and Gideon Parchomovsky, "Patent Portfolios" (link)