Yelderman on Accuracy in the Patent System

When the USPTO is faced with a patent application of uncertain validity, which way should it err? How do the costs of erroneous patent grants compare with the costs of erroneous patent denials? Steve Yelderman provides an insightful new take on these questions in The Value of Accuracy in the Patent System (forthcoming in the University of Chicago Law Review). In short, he concludes that erroneous patent grants are more harmful than previously assumed, but only when "the unsatisfied patentability requirement is one that seeks to influence a mutually exclusive choice"—i.e., whether it incentivizes invention over non-invention, disclosure over non-disclosure, licensing over reinvention, etc.

Under the conventional account, if one assumes that the legal criteria of patentability are set correctly (an important caveat!), then social welfare is maximized by awarding patents to patentable inventions (true positives) and denying patents to unpatentable inventions (true negatives). Erroneous denials (false negatives) create a cost of decreased ex ante incentives for other inventors (I_FN), while erroneous grants (false positives) and correct grants create ex post costs like deadweight loss (C). If the probability that a given application is patentable is q, then it should only be granted at the threshold q > C/I_FN. (Note that it doesn't matter whether I_FN is included as a cost for false negatives or a benefit for true positives; you get to the same result.) If the ex post costs of patents are low relative to their incentive benefits, they should be granted even for inventions that are unlikely to meet patentability standards, and as the relative weight of these factors reverses, patents should only be awarded when it seems quite likely that they are deserved.

Yelderman's key insight is that this analysis ignores another important cost of erroneous grants: false positives also reduce ex ante incentives because they often narrow the expected difference between inventing and not inventing. This framework can be illustrated as follows, with I_FP shown in blue:

	Invention Patentable (prob. q)	Invention Not Patentable (1 – q)
Award Patent	– Imposes Ex Post Costs (C)      (True Positive)	– Imposes Ex Post Costs (C) – Reduces Ex Ante Incentives (IFP)      (False Positive)
Deny Patent	– Reduces Ex Ante Incentives (IFN) * No Ex Post Costs      (False Negative)	* No Ex Post Costs      (True Negative)

When this additional cost I_FP is added in, patents should be awarded only if q > (C + I_FP)/(I_FN + I_FP). Yelderman notes that this "reveals an intuitive principle that is often overlooked by traditional analysis of errors in the patent system: It would be important to examine and deny patent applications even if individual patents imposed no ex post costs on the public at all." Recognizing the incentive costs of false positives has the effect of driving up the minimum probability of validity q that should be required before a patent is awarded. And if one thinks that the incentive harm of a false positive is at least as large as the incentive harm of a false negative (I_FP ≥ I_FN)—perhaps because they are more observable—then even if C = 0, the minimum probability of patentability q to justify a grant is at least 50%, independent of the unknown marginal costs and benefits of patent protection. This "provid[es] a basis for increasing scrutiny of patent rights that does not depend on disputed empirical priors."

Of course, this conclusion depends on the premise that false positives decrease ex ante incentives. As Yelderman explains, this is only likely to be true for some patentability requirements. Part III of his article explores the extent to which different patentability requirements are designed to influence a mutually exclusive choice, and here, we can't completely get away from disputed empirical priors. For example, lax enforcement of the nonobviousness requirement will only reduce ex ante incentives if it encourages inventors to pursue obvious projects over nonobvious ones—but are there real constraints in how many R&D projects will be pursued, or does it make more sense to consider each decision to invest in R&D in isolation? Yelderman concedes this uncertainty, but he makes a good argument that his two-sided incentives model generally should apply to enforcement of the disclosure requirements, and to novelty—at least when the novelty issue is based on accessible prior art.

Even readers who do not agree with all the conclusions in Part III will likely conclude that the "mutually exclusive choice" factor should at least be given more attention. And I expect most scholars of the law-and-economics of the patent system will find something in here to intrigue them. Yelderman's article has certainly improved my thinking about the costs of errors in patent grants and denials, and for that, I think it is well worth a read.