Recent Research: Local Factors Influencing Tech Commercialization

What are the factors of commercial success? As they say in real estate: location, location, location.

So what makes a good location for commercializing innovation? Innovative ideas clearly thrive where R&D spending flows and local patent activity exists. But, do R&D dollars and level of patents also indicate locations for tech transfer?

Not necessarily. A recent working paper in applied economics finds a more complex web of relationships at work.

Dr. Joshua Rosenbloom of University of Kansas seeks to identify the factors in common for the areas with greatest commercial innovation in The Geography of Innovation Commercialization in the United States During the 1990s. Dr. Rosenbloom analyzes three measures of commercialization: public grants (SBIR and STTR), venture capital investments, and initial public offerings (IPOs) within the 50 largest U.S. metro areas. He also uses econometric models to identify the factors that influence those measures of commercial innovation, or tech transfer.

Controlling for variation due to population, Rosenbloom found that patent activity and the number of doctorates awarded (defined as scientific and engineering capacity) had the greatest influence over his three measures of commercialization.

In his analysis, five metro areas account for at least 40 percent of the three commercialization activities in the U.S. in descending order: San Francisco, Boston, Denver, San Diego and Austin. The extent of commercialization within these regions is highly unequal, with Austin achieving less than half (43 percent) the level of commercialization in San Francisco.

However, patents and doctorates cannot explain all of the regional variation and high concentration of tech transfer in a few areas. For instance, Rosenbloom notes, "Columbus, Detroit, Cincinnati and Minneapolis appear to be doing a relatively poor job of converting patents into commercial innovations."

Using a causal model and assuming a linear relationship, Rosenbloom finds the number of doctorates has a highly significant impact on the commercialization measures. "For instance a 10 percent increase in [doctorates] would increase SBIR/STTR grants by 5 to 6 percent, venture capital funds by 6.5 to 7.5 percent and IPOs by 1.7 to 2.9 percent." When factoring in both doctorates and patenting, only the level of doctorates remains a significant indicator of commercialization. However, IPOs appear to be more connected to the level of patenting than doctorates.

The model suggests the relationship between SBIR and IPOs "appears relatively weak," a finding that may surprise many involved with the federal set-aside for small tech businesses, given recent debate on the role of venture capitalists in SBIR-eligible firms.

Rosenbloom offers several observations that may guide or influence local and regional tech-based economic development policies:

• High rates of innovation commercialization do not necessarily follow from high rates of idea generation. The commercialization process may need to be supported for its economic benefits to be realized locally.
• Increasing the university science and engineering capacity and the resulting number of doctorates promotes innovation, particularly through more SBIR awards and greater venture capital investments.
• "The unexplained differences in the different measures of commercialization are highly correlated," Rosenbloom says, pointing to strong external economies centralized around the regions with both high innovative research and tech transfer.

The Geography of Innovation Commercialization in the United States During the 1990s is available at:

http://www.ku.edu/~bgju/2005Papers/200502Rosenbloom.pdf. Statistics for each included metro area are provided in tables 1 and 3 of the study.

[Editor's Note: It is important to note that Dr. Rosenbloom's method of measuring commercialization is relatively narrow and, as a result, his findings should be viewed in that context. One can argue, in particular, that using SBIR grants as a measure of commercialization is a stretch, however. Dr. Rosenbloom faces the problem we all do -- the lack of data to measure what we want to measure. Despite these limitations, we've written about the paper because we believe it provides food for thought for the TBED community.]