Later today, the Senate will hold a hearing on how Congress can help prevent the issuance of poor quality patents. I recently testified at a hearing on the STRONGER Patents Act and patent quality was one of the topics discussed. In fact, in my responses to questions for the record from Sen. Tillis, I addressed exactly this topic. In particular, I identified proposals from one of today’s witnesses—Prof. Melissa Wasserman—to provide examiners with more time to conduct examination and to restructure USPTO fees to reduce USPTO reliance on post-grant fees.
The use of randomized trials for examination improvements, suggested by another witness at today’s hearing, Prof. Colleen Chien, would also help ensure that changes to examination can be shown to have positive impacts, rather than simply asserting that the changes are necessary. Prof. Chien has also suggested that there are subtle failures in the U.S. examination process that can be detected by comparison to peer patent offices such as the European Patent Office. Once detected, procedures can be implemented that will, for example, reduce U.S. examiner unwillingness or inability to rely on non-patent prior art.
These changes, combined with ensuring that the IPR process for conducting post-grant review of patent validity remains available, would help prevent the issuance of marginal patents.
My detailed response to Sen. Tillis’s question continues below.
CCIA strongly supports the goal of improving patent examination in order to ensure that the patents that issue from the USPTO fully deserve their statutory presumption of validity. While a perfect system in which invalid patents are never issued is impossible, it appears very likely that the present system could be significantly improved. One of the best proposals is also the simplest—the USPTO should provide examiners with more time to conduct examination. There is empirical evidence that the time constraints placed on examiners result in the allowance of marginal patents. This proposal does not require additional statutory authorities. In fact, the USPTO has proposed a review of examiner time allocations in its 2018-2022 Strategic Plan and could easily increase examiner allocations as part of this process. This would likely decrease the issuance of marginal patents.
Another simple proposal is also within existing statutory authorities. The USPTO’s current fee structure likely contributes to the issuance of marginal patents. In particular, the USPTO does not recover the full cost of examination using its examination fees, but instead relies on issue fees and maintenance fees to cover the full cost. Because of this, if an insufficient number of patents issue in a given timeframe, the USPTO will not receive enough fees to cover its cost of operations. Empirical evidence suggests that this may contribute to the issuance of marginal patents. The SUCCESS Act renewed the USPTO’s authority to set and adjust its patent fees. Modifying the USPTO fee structure to recover more of the cost of examination at the time of examination, rather than after issue, would be within the scope of this statutory authority, and would be likely to decrease the issuance of marginal patents.
The USPTO Director is actively investigating ways in which machine learning and artificial intelligence might be employed to increase examination efficiency. CCIA supports this investigation, but it is unclear within what time frame this may provide serious impact and it cannot be relied upon as a sole method of increasing examination quality.
Finally, CCIA again commends the USPTO on its transparency and commitment to making available data. That commitment has enabled much of the empirical research which has suggested these avenues for improvement. Maintaining the access it has historically provided will help ensure that the next generation of research into examination quality and USPTO operations can provide useful policy recommendations founded on empirical evidence.