Patent Quality: Towards a Systematic Framework for Analysis and Measurement

How should we define and measure patent quality?

We see measuring patent quality as essential for evidence-based policy and for understanding the impact of technological innovations. Patent quality is highly variable, affecting how different stakeholders may perceive the same patent. There is an increasing reliance on patent indicators to assess innovation's significance, but prior studies reveal inconsistencies in these measures. Understanding these inconsistencies helps refine empirical methods and informs policy discussions about patenting strategies and innovation funding.

We select a cohort of 634,003 patents granted by the United States Patent and Trademark Office from 2001 to 2004 and evaluate three commonly used post-grant outcomes: stock market reactions to grant events, patent renewals, and forward citations. We employ statistical analyses to assess the overlap and relationship between these outcomes based on intrinsic patent characteristics.

We find that the three patent quality measures show significant inconsistency regarding which patents are deemed high quality. Specifically, there is an important lack of agreement among the measures, with forward citations showing the strongest correlation with renewal decisions. Furthermore, we show that the relationship between patent characteristics and ex post outcomes differs across technological fields, emphasizing the need for contextualized interpretations of patent quality indicators.

• A cohort of 634,003 patents granted by the USPTO between 2001 and 2004, including indicators such as stock market reactions, patent renewals, and forward citations.

• We assume that different metrics used to evaluate patent quality capture distinct dimensions of quality.
• We assume that patent characteristics established at grant are associated with post-grant outcomes.