Technological innovation is a key factor – if not the key factor – in economic growth and job generation. But in today’s knowledge economy, the capacity to innovate and generate new high-tech industries and jobs is much more concentrated and clustered in some places than others.
Why? That’s a question that has long vexed economists and policy-makers.
A new study focusing on the key concept of "innovation productivity" offers a new take on this question. The authors — my University of Toronto colleagues Ajay Agrawal and Alberto Galasso, Boston University’s Iain M. Cockburn, and Georgia Institute of Technology’s Alexander Oettl — previewed their research on the VoxEU blog in late December.
The authors refine a basic measure of innovation based on the number of patents, which they call "innovation productivity." To measure the productivity of the local tech economy, the study looks at two significant fields — computers and communications. They look at the number of inventors working in each metro area, and then compare this to the number of citation-weighted patents that the local economy produces per inventor — in effect, checking to see if there is an economy of scale in the innovation economy. They use this to provide insight into why some metros are more productive at turning their talent into both overall innovation and commercially relevant ones. Then, they extend their analysis to look at the underlying factors creating strong innovative economies and making some places far more productive than others.
The graph below, from the VoxEU post, looks at this "innovation productivity metric" — plotting the number of important innovations per inventor against the overall number of inventors working in these fields. It thus compares the industry’s potential — based on the number of inventors — to how significant its output was.
Note how far Boston is over to the right of the graph. Huge, even unparalleled number of inventors are employed in its tech firms and universities. But they saw little economy of scale, and didn’t produce proportionally more important innovations than their counterparts in Dallas, Atlanta, or San Diego. In fact, inventors working in Seattle, Washington, D.C., and Portland, Oregon were far more productive than those in Boston.
What can explain these differences? And what can places do to ensure that their existing tech workers are being as innovative and productive as possible?
The study finds that the key differentiating factor lies in the way that inventors are organized and employed. On the one hand, some cities and metros have focused on bolstering and concentrating the research and development capabilities of their universities and large companies — the so-called "anchor institutions."
On the other hand, more recently, a number of places have focused on developing more robust entrepreneurial and start-up ecosystems where smaller firms can thrive. As Boulder-based entrepreneur Brad Feld has noted, the "inclusive" and "non-hierarchical" nature of a loosely connected start-up community is a key to building a strong local innovation economy.
These approaches have pros and cons, according to the study. Large companies, universities, and other anchor institutions help spread the fixed costs of research (think buildings, suppliers, etc.). And by massing large numbers of research and development workers together in large organizations, they give the capacity to capitalize on new inventions. The authors argue that the presence of just one large-scale firm can be a huge advantage in providing the institutional heft behind a fledgling innovative economy.
But these big companies tend to focus on things that relate to their existing products and may overlook inventions in other areas. Regions with large numbers of smaller, more entrepreneurial companies can gain an edge in capitalizing on these innovations that larger companies might pass over. The story of Steve Jobs' trip to Xerox's Palo Alto Research Center — walking away with a whole host of ideas Xerox hadn't brought to market — is perhaps the most legendary example of this.
Take a look at Portland, Oregon, and Rochester, New York, on the graph above. Both are at the same place on the horizontal axis, with about 1,000 inventors apiece. But Portland had twice the number of quality-adjusted patents per inventor. One explanation, it turns out, could be the number of smaller tech firms. Though both metros had similar numbers of large labs, Portland registered a whopping five times as many smaller labs as Rochester. The study concludes that having a diverse firm base — with at least one large lab and many smaller labs — can increase innovation by 17 percent.
• • • • •
So what does this mean for cities and metros that want to improve their climate for innovation and encourage more 'productivity' from their existing pool of inventors? As one Silicon Valley entrepreneur and observer once noted a touch cheekily, the key to creating the next Valley would be as simple as "take one great research university. Add venture capital. Shake vigorously."
The reality, of course, is far more complicated, as the authors explain:
Policies focused exclusively on attracting ‘anchor tenants’ or cultivating new ventures may not be as effective as promoting firm size diversity, leading to an appropriate mix of large and small firms. In other words, our results suggest that there is no universal ‘best’ policy…. A region with large firms but few young entrepreneurial firms may benefit more from policies designed to cultivate new ventures rather than to attract more large firms, whereas regions without local large firms may benefit most from attracting these.
The study essentially argues that there is no one-size-fits-all solution to creating an innovative tech economy. Though civic boosters love to tout single-solution policies — by focusing on attracting one major tech firm, or by bolstering their start-up ecology — a mix of both approaches may be far more effective and prudent.
Top Image: Thomas Alva Edison, that most famous inventor, working on his dictation machine. Courtesy Library of Congress via Wikimedia Commons.