How to Measure and Draw Causal Inferences with Patent Scope

Science quality and the value of inventions

“…Third, we measure patent scope by the length of the text of the first independent claim. This relies on evidence showing that longer descriptions of the claimed invention implies more narrow legal protection and, therefore, a lower patent value (Kuhn and Thompson, 2019). We consider separately, and when available, the length of the first independent claim in the patent grant publication at the USPTO or the EPO.…”

Visibility of Technology and Cumulative Innovation: Evidence from Trade Secrets Laws

“…Table 1 summarizes these variables across all patents in our main sample. We proxy for a patent's breadth and complexity using the number of independent claims (see Lerner, 1994;Lanjouw and Schankerman, 2004) and the length (in words) of the first claim (see Kuhn and Thompson, 2019), where shorter claims are likely broader. As an additional measure of a patent's complexity, we include the length (in characters) of the patent's detailed description text.…”

“…We follow Lerner (1994) and Lanjouw and Schankerman (2004) and measure patent breadth and scope using the number of independent claims in a patent. Kuhn and Thompson (2019), however, argue that a simple count of (independent) claims may be a poor measure for patent scope. 3 They propose the length of the first patent claim as an alternative measure for patent scope, where shorter claims are broader.…”

“…Kuhn and Thompson (2019) argue that under U.S. law the broadest claim should be listed first.55 We followGraham and Vishnubhakat (2013) in identifying patents as software patents. In our data, 66% of all software patents include a process claim, as opposed to 40% of non-software patents.…”

Knowledge Spillovers and Learning in the Workplace: Evidence from the U.S. Patent Office

“…For these purposes, we collected data from Jeffrey Kuhn and Neil Thompson on the number of words added throughout the examination process to the first claims in the patents issued during our sample. 22 Following, Kuhn and Thompson (2017), we use this word-added measure as a reflection of the degree of claim narrowing, considering that longer claims generally impose a greater number of conditions that must be met before patent infringement is found. In Panel A of Table 7, we present the results from this approach, whereas in Panel B, we present results from specifications that form the claim-narrowing variable in percentage terms-i.e., normalizing words added by the number of words in the first claim of issued patents.…”