Academic and non-academic scholars alike have tried to pinpoint where the centers of innovation are. Whenever the question arises, the usual suspects -- Silicon Valley, Boston and, often, Austin, Texas and the Research Triangle in North Carolina -- come up high on whatever list is created. Those rankings usually are based on a concentration of high-tech industry, venture capital investment, world-class universities and patents per capita, under the assumption that these indicators are a good proxy for innovation.
But what happens if we modify the question by asking: Where are the fast-growing firms geographically concentrated? Even without including the term "innovation," this question actually gets us closer to detect where innovations are located. Why?
Innovations, according to Joseph Schumpeter, a pioneer in the theory of innovations and entrepreneurship, are defined as something new that produces commercial value. By definition, firms with fast-growing revenue have established commercial value. The annual Inc. 500 list provides an excellent dataset for analyzing the geography of fast-growing firms. Each year since 1982, the list has ranked private firms with fastest-growing revenues in the nation. It contains firms' locations at the street level, allowing us to very specifically trace fast-growing firms' geography over the last 30 years, though we focus here only on the last decade.
In the first-ever geographic study of the Inc. 500 companies that my Kauffman Foundation colleagues and I just completed, we report two noteworthy and surprising findings. First, among large metropolitan areas, Washington, D.C., has the highest concentration of Inc. firms, both in absolute number and per capita (normalized by population). Additionally, more than 46 percent of Inc. firms in the D.C. area operate in government services. This D.C.-area cluster has emerged in the last two decades, regardless of which political party's administration was in power. It demonstrates, ironically, that government can play a large role in the growth of private firms.
The second surprise is that clusters of high-growth companies are located outside of the usual innovation suspects. The top 20 metros that house these fast-growing firms include Salt Lake City, Utah, (2nd), Indianapolis, Ind., (6th), Buffalo, N.Y., (11th), Baltimore. Md., (15th), Nashville, Tenn., (18th), Philadelphia, Pa., (19th), and Louisville, Ky. (20th). The so-called Rust Belt Regions may experience population declines, but firms in those regions can still produce a number of innovations!
We did not stop at just producing the ranking of metros. It is more important to understand what regional factors are associated with the cluster of Inc. firms. It turns out that the usual proxies -- the presence of venture capital investment, patents per capita, high-quality research universities and federal R&D spending in Small Business Innovation Research, a research fund disbursed for small businesses -- are not associated with a concentration of Inc. firms. Instead, our analysis found that the presence of a highly skilled labor force is the only important regional factor associated with high numbers of Inc. firms.
These findings have several implications. First, cities don't have to join the race to become the next high-tech center or Silicon Valley. After all, innovations from fast-growing companies -- which include several other industry segments in addition to high tech -- generate wealth and employment, and policymakers should consider more about what are sources of growth for these companies and how regional development models function in those understudied areas. Second, government seems to play an important role -- not in the sense of providing R&D funds, creating large research universities or establishing public venture funds, but by training a highly skilled labor force. Last, while federal government employment has stayed relatively constant since the Clinton years, the spending per GDP has increased steadily since 2000. The U.S. government generally is not associated with industrial policy. However, government outsourcing creates de facto industrial policy. It therefore is important to revisit government's role and spillover effect in the development of clusters of fast-growing companies.