Patent Counting and the 'Top-Down' Approach to Patent Valuations: An Economic and Public Policy Appraisal of Reasonable Royalties

Abstract

In many circumstances it is helpful, and sometimes necessary, to assess (possibly even to quantify) the technological prowess of a business enterprise, either overall or with respect to particular fields of application, or possibly with respect to the firm’s relative position in an industry. In such circumstances, it is tempting to use as a measure the number of patents that has been granted to a firm. However, patent counts are an imperfect and unreliable metric. Using them may create an aura of accuracy, but it is false (scientific) accuracy for the reasons discussed in this article. In particular, the “top-down” approach to the valuation of standard-essential patents (SEPs), which relies heavily on patent counting, is a poor surrogate for the determination of the value of patented technologies. I will start with some basics. In scientific inquiry, precision refers to how close the measurement of a variable is to what is being measured. Precision is, however, independent of accuracy. Indeed, it is possible to be precise but highly inaccurate. Accuracy is, of course, more important than precision. In this paper, I will show that patent counting, while having the possibility of being precise, does not always meet that criterion in part because of ambiguities as to scope. For instance, sometimes standards are at issue with patents “reading on” or being “essential” to one or more technical standards. However, there may be ambiguities around how many patents in a given portfolio are in fact essential, versus simply declared essential by the owner or some third party. In this article, I make two suggestions. First, patent-count metrics are at best poor proxies of technological strength or value. This is not just because of inaccurate patent counts in the numerator or denominator of some index. It is also because there is at best only a weak connection between even well-specified patent indices and underlying economic value of a patent or patent portfolio. It is often the case that one will have to look downstream to the user to figure out the incremental value that the technology yields to the consumer. Second, when it comes to valuing intellectual property that “reads on” a standard, the numerical proportionality of standard-essential patents (SEPs) is a bogus measure. It is unlikely to measure the relative value of patents, let alone the value of technology. The problem is compounded because numerical proportionality requires the determination of a “total value” associated with all patents that “read on” a standard, which has typically been arrived at arbitrarily.