Page 22 -

patents, although we will demonstrate later that universities may, on average, produce
more patents with technical breakthroughs.

Patenting in biofuels has been steadily increasing over the time period, and this increase
appears to have come mainly from large firms, which maintain a dominant role in
patenting activity. This raises the question of whether independent inventors are
inventing poor quality technology or whether other issues, such as commercialization
challenges or a lack of social networks that diffuse knowledge, are the cause.

3.1.3 Funding Sources
We do not observe broad variability over time in the role of government funding in
producing inventive activity. Broadly speaking, in solar, both DOE and the National
Aeronautics and Space Administration (NASA) historically account for about one-third
of the patents with government rights associated with them. A similar trend is observed
in wind and biofuels. The Department of Defense (DOD) and National Science
Foundation (NSF) both play a small role in funding patents (< %5 each). It would be
interesting to attempt to correlate government technology funding (e.g., Anadon et al.
[11]) with patent output. With sufficient data resolution, bibliometric scholars could
make some significant evaluations of the effectiveness of funding agencies on stimulating
innovation and the efficiency of research and development organizations in converting
public funding into public good [65, 66].

3.1.4 Social Features
We also performed preliminary analysis of social networks of patenting. Unfortunately,
our keyword-driven search approach creates artificial boundaries around social networks
of invention, and for this reason, we do not calculate a unique social influence variable
for each inventor since their true social influence should be derived from their entire
inventive career, which is typically much broader than this dataset would indicate (Figure
5). In this figure, an NREL innovator is identified (circled), and their energy patents
identified in our database are mapped (A). However, that analytical method may
underestimate the actual network of collaborators that the inventor has previously worked
with (B) and the total network of their inventive sphere (C). The inventor’s inventive
career likely has spanned other disciplines besides energy, and identifying formative
collaborations that lead to energy innovations is essential. Each node is an inventor
connected by lines to their collaborators (distance is a network plotting artifact and does
not indicate closeness), colors indicate different patent assignees, and square nodes and
red lines indicate energy patents versus black lines an circles for non-energy patents.

15