Building a Better Bridge: Improving Patent Assignee-Firm Links

David W. Dreisigmeyer, Nathan Goldschlag, Marina Krylova, Wei Ouyang, Elisabeth Ruth Perlman

2018 · 12 citations · 0 references

Concepts

Abstract

In this paper we describe the creation of the Business Dynamics Statistics of Patenting Firms (BDS-PF) patent assignee-FIRMID crosswalk. A number of efforts have been made to link patent assignees, the businesses to which patents are granted, to Cesnsus Bureau business microdata (e.g., Kerr and Fu (2008) and Balasubramanian and Sivadasan (2010)). The coverage and quality of these links are limited by the lack of detailed information about patent assignees found in the USPTO patent data. The BDS-PF crosswalk overcomes these limitations by leveraging additional information about inventors to generate more and higher quality patent assignee-FIRMID links. The match methodology extends and improves the triangulation strategy first introduced by Graham et al. (forthcoming). At its core, the triangulation methodology leverages fuzzy matches of both patent assignees and patent inventors, in combination with job-level data, to disambiguate and validate matches.