Using patent technology codes to study technological change

Abstract

Much work on technological change agrees that the recombination of new and existing technological capabilities is one of the principal sources of technological novelty. Patented inventions can be seen as bundles of distinct technologies brought together to accomplish a specific outcome – and this is how the US Patent Office defines inventions. The technologies constituting inventions are identified by the US Patent Office through an elaborate system of technology codes. A combinatorial perspective on invention, emblematic of approaches to technological change informed by evolutionary economics and complexity science, is inherent in the use of technology codes to summarize what is technologically novel about a patented invention. The technology codes represent a set of consistent definitions of technologies and their components spanning 220 years of inventive activity, and are an underutilized data resource for identifying distinct technological capabilities, defining technology spaces, marking the arrival of technological novelty, measuring technological complexity, and empirically grounding the study of technological change. The present discussion provides an introduction to the use of patent technology codes as well as some basic empirics. Our results highlight the highly discriminating nature of the codes and their usefulness in characterizing the type of processes by which technological capabilities generate novelty.

Keywords:

• patents
• technology codes
• technological change
• technological novelty

JEL Classification :

• O30
• C81
• N7

Acknowledgements

We benefitted from comments by Luis Bettencourt, Fabrizio Lillo, John Miller, Bruce Kogut, William Kerr, and Tihamer von Ghyczy. The work presented in this paper was largely done during a 3-week research stay at the Santa Fe Institute during the summer of 2010 – we gratefully acknowledge the Institute's hospitality.