2 Highlight of Patent analysis The content of patent data Inventor Assignee Application/issued date IPC/UPC Reference/citation The unit of analysis Firm-year level (cross-section & time series) Patent level Firm level
3 The content of patent data Backward citation IPC UPC
14 Millar, Fern & Cardinal (AMJ, 2007) Knowledge sourcing (boundary of firm and its divisions) Intra-divisional knowledge sourcing negatively affects forward citation Extra-organizational citation (positive effect) Inter-divisional citation (positive effect) Data NBER (National Bureau of Economic Research Patent Citations Data File) MicroPatent Corporation Time frame: 1985-1996 1,644 firms 211,636 patents (observations) Unit of analysis: patent Negative binomial regression
15 Patent-based innovation performance- Patent count Impact Rosenkopf and Nerkar (2001) optical disk patents Domain impact equals the number of citations from optical disk patents (that is, citing patents that were classified in any of our initial optical disk subclasses) received by firm i’s patents granted in year t. non-optical disk patents Overall impact is the total number of citations from non-optical disk patents received by firm i’s patents granted in year t. Breakthrough innovation Phene, Fladmoe-Lindquist and Marsh (2006) Forward citations, excluding self-citations. Every original patent has an equal 10-year time window for citations. (citations received) Top 2 percent of the sample (15 original patents out of the total of 707 patents) were identified as breakthrough innovations.
16 Patent-based innovation performance- Persistent innovation Patent spell Alfranca, Rama and von Tunzelmann (Technovation, 2004) patent spells as periods of time during which the company innovates year after year without gaps in its activity. Follow-on patenting McGrath and Nerkar (SMJ, 2004) Taking out a second patent in a patent subclass that is new to the firm ( it has only one previous patent in a new technological areas that it had not patented in before).