Better Mistakes in Patent Law

This Article analyzes patent mistakes — that is, mistakes made by the patent system when it decides whether a particular invention has met the patentability requirements. These mistakes are inevitable. Given resource constraints, some might even be desirable. This Article evaluates the relative costs of patent mistakes, so that we can make better ones. Three characteristics drive the costs of mistakes: their type (false positive or false negative), timing (early or late), and doctrinal basis (utility, novelty, nonobviousness, and so on). These characteristics make some mistakes more troubling than others. This Article compares the costs of making mistakes of different types, at different times, and on different doctrinal bases. These comparisons produce some surprising results — for example, under certain plausible conditions, it will be better to wrongly refuse to grant a patent than to wrongly invalidate a patent that had already been granted. The conclusions here have important implications for persistent issues in patent law, including how closely courts should scrutinize the validity of issued patents and how the Patent and Trademark Office should allocate scarce enforcement resources.