PANEL I
WHY CLUSTERS MATTER:
INNOVATION CLUSTERS AND ECONOMIC GROWTH
Moderator: William Kittredge Economic Development Administration Department of Commerce
Cluster Development: A Path to Growth
Maryann Feldman University of North Carolina
Dr. Feldman began by noting the strong interest in cluster development she had found in other countries, and joked lightly about how late this interest had come to her own country. “It’s my pleasure to be with you today,” she said. “This is very similar to a talk I’ve been asked to give in India, China, Europe, Japan, and Korea. It’s wonderful to now be giving it in the U.S.”
She began her discussion by giving “the academic viewpoint on cluster development,” and signaled her agreement with previous speakers on the fundamental point that “all growth is local and grounded in place,” taking the focus from nations to sub-national units.” She attributed this shift to a local focus on understanding innovation as a “cognitive and contextual process,” predicated on face-to-face interactions, serendipity and chance encounters and their outcomes.
One kind of outcome, she said, is that people, when dealing with something new, naturally develop a shared common meaning and a language to describe it. And the characteristics of knowledge are such that it grows when it is shared among people. These network effects lead to increasing returns and greater productivity from economic activity. “When we look at firms in that context, we realize why they benefit from strategic location. This doesn’t mean firms that are attracted to a jurisdiction for a while and then move out. It means firms that have deep roots and deep social connections. Often these firms are homegrown. One of the great paradoxes of globalization is that even as we are all more interconnected, what really matters is place and location specific context. Clusters form both here and in other nations because humans are
physical and social beings.” The best position is to be locally grounded but globally connected.
Clusters Reflect the Qualities of the Place
In this light, it is logical that cluster formation reflects the local qualities of the place where it forms. And it follows that it is seldom possible to imitate a cluster formed elsewhere. In the words of Robert Metcalfe, a pioneer of the Internet: “Silicon Valley is probably the only place on earth not trying to copy Silicon Valley.”1 How, then, do clusters come into existence? That is, how do regions change from being inert—with little innovation, little entrepreneurship, slow economic growth—to being active places? She said that a central finding of scholars who study clusters is that they are not “economic development sausage machines,” where the right ingredients added at one end produce the desired product at the other. This logic leads to a persistent creation myth that calls for lining up a research university, some venture capital, and some entrepreneurs, and then “turn a crank” to produce a cluster with good economic growth.
Studies of famous clusters revealed no evidence of a single creation formula; on the contrary, they suggest the opposite. Hollywood, California, for example, is such a successful cluster of film industry activities that the name refers to both the place and the industry. This success, according to a study by Allen Scott of the University of California, Los Angeles, drew on a unique blend of causes. The movie business, he showed, depends on its own form of clustering and a coincidence of new ways of organizing the film industry.2 Similarly, Silicon Valley had its own particular characteristics as a cluster region that far transcended the presence of major universities, entrepreneurs, and venture capital firms. The task of creating a cluster by conscious intent is further complicated by the fact that many appear to be the products of historical accidents or serendipity. She concluded that the consensus in the literature is that social processes are the most important determinants of cluster development.
These social processes, more than location or physical attributes, combine with or produce a vision of some new way of doing something.
_______________________
1Robert Metcalfe, an early pioneer in developing the Internet, has also been an entrepreneur, publisher, and columnist, and is currently a venture capitalist with Polaris Venture Partners in Waltham, Massachusetts.
2Allen J. Scott, On Hollywood: The Place, the Industry, Princeton, NJ: Princeton University Press, 2004. Scott attributes much of Hollywood’s success to its physical density and the proximity of many specialized but complementary skills. This clustering is essential, he writes, because “the relations between firms cannot be planned over extended periods of time so that useful inter-firm contacts need to be constantly programmed and reprogrammed.”
For example, Hollywood superseded New York as the nation’s movie capital because of their new innovative ways of making a film. The old way was a theatrical model, in which the action, including not only human speaking and actions but also train crashes and other simulations, were filmed on an old-fashioned stage. The resulting film was sold by the foot, like a commodity. The new social consensus about making films was to add value to the product by using a studio and a continuous script. And this was more easily done on the huge outdoor Hollywood lots than on the Broadway stage, quickly attracting the rest of the film business to the West.
Metrics Should Include the Quality of Community Life
In planning how to measure the outcomes of clusters, she suggested we should not limit metrics to innovation or job formation. For example, in a recent report from the Milken Foundation reported an innovation index of 100 for Silicon Valley; the next closest place scored 503. “But we don’t all want to live in Silicon Valley,” she observed. “When we think about metrics, we need to think about economic outcomes more broadly, about community prosperity. Prosperity is a wonderful word that means sharing in the distribution of the outcomes.” This community prosperity includes the broad question of what kind of world we want to live in, including all relevant economic and social objectives. She discouraged planners from thinking of Silicon Valley as the cluster archetype; creating another such cluster, she said, was unrealistic, undesirable, and unsustainable. There is a paradox in such thinking because “the essence of corporate strategy is doing something unique that is not easily replicated.” To prove this point, she cited the current example of the biotechnology industry. “When you look around the U.S., everyone is trying to capture a biotech industry, even though a few places have already moved so far ahead.”
What, then, is a city or region to do? Feldman returned to her point that designing an economic development strategy “may be the ultimate local innovation.” In Asia, the government is able to dictate from the top down that a cluster will be established in a certain location. She put this in contrast with the West. “In a free-market capitalist economy,” she said, “that won’t work. The clusters we have in the U.S. and Western Europe are complex, self-organizing, and composed of a broad patchwork or ecology of people and institutions.” For such clusters, the role of the government is to provide appropriate incentives. Feldman
_______________________
3Ross C. DeVol, Kevin Klowden, Armen Bedroussian, and Benjamin Yeo, North America’s High-Tech Economy: The Geography of Knowledge-Based Industries, June 2, 2009.
cited Nassim Nicholas Taylor’s 2007 book The Black Swan4 and its premise that people tend to be limited by their own experiences in thinking about what outcomes might be possible or feasible. “What Black Swan teaches us,” she said, “is that it’s difficult to anticipate new technology or future economic change.” The role of the government in promoting innovation is therefore more challenging than that of a private company. While a private company seeks primarily to grow and to earn a profit, a government must consider many outcomes of cluster activity, including the quality of life in the community. “An economic development strategy that will work,” she said, “has to be predicated on a deep understanding of the location.”
Ideas Do Not Stop at Borders
Feldman then referred to the concept of “coherent geographical systems” as a framework for organizing economic activity. The defining feature of a coherent system is one that catalyzes the flow of ideas. Political boundaries are not useful frameworks for clusters, because ideas do not stop at borders. Instead, it is more helpful to think of political units as divisions of the same economic entity. These units can collaborate in economic activities for the greater good, rather then competing for the benefit of one fraction of the population. She urged planners to consider each place’s position in an urban hierarchy, for example, in deciding what kinds of collaboration are feasible.
Another topic she discussed under the concept of coherent systems was the danger of concentrating resources. “Do we want to live in a world where everyone is in a city?” she asked. “We need to consider synergies between urban and rural areas. We have wonderful lessons from 30 years of state and local technological development initiatives. We need to learn from these laboratories.”
She addressed Susan Crawford’s question about the role of Government, responding that “private activity is just the tip of the iceberg.” Capacities, incentives, and institutions drive innovation. These are traditionally the responsibility of government. “We have seen the growth of new quasi-governmental entities, such as public-private partnerships and nonprofit organizations, to become catalysts in this process. And it is important to remember that economic development policy is increasingly about social policy and enabling development potential.”
A New Language for Clusters: Resiliency and Diversity
She urged that the debate be recast to replace competitiveness and a “winner-take-all, casino economy” with a new language of biology. By
_______________________
4Nassim Nicholas Taleb, The Black Swan: The Impact of the Highly Improbable, New York: Random House, 2007.
this, she did not mean “survival of the fittest,” but other lessons of evolutionary biology such as resiliency and diversity. “You would never tell an investor to put all his money in one stock or one sector. You would advise diversification. The same lesson needs to be applied to the economy.” Resiliency is needed to survive external shocks; diversity is needed to ensure that enough members survive changing conditions.
She also suggested several revisions to our understanding of how academia, industry, and government contribute to economic development. With respect to the academic contribution, “A dirty little secret of universities is that very few university offices of technology transfer make money. But universities still provide fundamental resources that are critical in establishing clusters.” Similarly, she said, many people disdain the lifestyle of entrepreneurship as mundane. But the day-to-day “hunting and gathering” activities of small firms can be essential to the success of clusters. Finally, government is often characterized as an agency of collective action, she said, but government does not function remotely or unresponsively. “Government action is and should be subject to citizen preferences.”
Focusing on Local Capacities
To be viable in the global economy, clusters must act locally and practically. The literature suggests that clusters do best when they focus on local capacity, expertise, and problems that are well understood. When a product or idea is developed and found to be viable locally, it can often be readily scaled for global markets.
A cluster should also focus on education. Its most valuable programs may be strong partnerships between universities, community colleges, and high schools. She offered a reminder that what the United States does best is to support broad-based, universal education and produce well-educated university graduates.
Dr. Feldman concluded by suggesting that knowledge and economic development, as reinforced by clusters, will be essential to the United States as other nations increase their own standards of living. She reminded her audience that innovation must be regarded not only for its economic power, but also in light of its highest purposes, such as providing the means to reduce the causes of disease and human suffering. Too often, she suggested, innovation can become an end in itself, unless we think about the ultimate purposes of new knowledge. “It is time for a new golden age,” she said. “Instead of treating knowledge as a commodity, we must value it as something that enhances human potential. Rather than regarding people as skilled inputs to production, we must think about them more broadly for their human potential. These are the outcomes we want from all of our clusters, and from social and economic development.”
Stimulating Regional Economies
Andrew Reamer The Brookings Institution
Mr. Reamer noted that a year previously he, Karen Mills, and Elizabeth Reynolds had collaborated and coauthored a paper on clusters and competitiveness as part of a “Blueprint for American Prosperity” series that Brookings was assembling in anticipation of a new administration.5 He said that he would report on the major ideas of that piece and offer some updating.
Clusters, he said, have a number of policy-relevant characteristics. All clusters are unique, differing in purpose, scope, composition, trajectories of development, and adjustment to external circumstances. Experience suggests, he said, that three factors are critical to cluster success:
- collaboration, which he defined as relationships;
- the skills and abilities of people in the workforce and the people at the head of the organizations in the cluster; and
- the cluster’s organizational capacities for innovation—traditionally thought of as products and processes, but including how to do business and how to relate to one another.
An Industry Can Have Only a Handful of Clusters
Despite long-time experience with clusters, he said, economic development is accompanied by “a lot of magical thinking.” He concurred with earlier speakers who said that “every place wants to be Silicon Valley.” Despite the fallacy of this wish, economic development agencies continue their efforts to do so, using any number of “magic bullets.” In the 1980s, they trusted that a business incubator was the key to success. “Today clusters have that danger,” he said. “They’re the next magic bullet.” The reality, he said, is that any industry can have only a handful of clusters. For biotech, San Francisco, San Diego, and Boston host more than half of the entrepreneurial activity, and it is very difficult for smaller areas to follow. “The more successful a cluster,” he said, “the more likely it will pull in firms from small places, because the best thinkers want to be where the action is.”
As an example, he mentioned a biotech initiative created by the University of Pittsburgh; the initiative had “some moderate success in building new firms,” but when firms reached a certain size, they left.
_______________________
5Karen G. Mills, Elisabeth B. Reynolds, and Andrew Reamer, “Clusters and Competitiveness: A New Federal Role for Stimulating Regional Economies,” op. cit.
Similarly, even the primary clusters fluctuate in relative strength. He said that a venture capital firm that had been in Boston for three decades had chosen the week before to move to San Francisco—one sign, he advised, that Boston is challenged in sustaining its prominence as a center of technology and venture capital. “The economy is in a perpetual state of transition,” he said, “so cluster dominance cannot be taken for granted.”
Facilitating Natural Collaborations
In a sense, he said, “we all grew up with clusters” of particular industrial activity: a cluster of tire companies in Akron, insurance in Hartford, cars in Detroit, glass in Toledo. Today, with improved information technology and transportations, new kinds of clusters are emerging based on functions, such as transportation in Memphis and Louisville. Part of the key to cluster success, he said, is a paradox: they develop competitiveness through a variety of collaborative activities. The objective of cluster initiatives is to facilitate those natural collaborative tendencies through joint efforts in market development, education and training, R&D, incorporation of new techniques, networking within the cluster and region, new business development, and marketing that attract firms and workers.
In the universe of cluster initiatives, he said, there are both wide differences and some common features. Among the differences are size: Some are tiny, some have hundreds of participants. Some follow economic boundaries, others political boundaries. They may be initiated by a city or a state. They may be a subsidiary of another organization, like a chamber of commerce, or a standalone effort that resembles a trade association. Among their common features: successful cluster initiatives tend to be ad hoc enterprises, formed from the bottom up. Most are led by industry, with government and nonprofit involvement. They occur across the full array of industry sectors, occurring in mature as well as new industries. A typical cluster initiative is supported by a dedicated organization that works hard to promote it and sustain collaboration.
The factors guiding successful cluster initiatives are reasonably well understood. Successful cluster initiatives are almost always industry-led and inclusive. Despite the natural tension between competing and collaborating at the same time, dynamic cluster initiatives manage to bring everybody to the discussion, including supply chain firms, educational institutions, and intellectual property experts. Successful cluster initiatives develop roadmaps to help understand where they are, where they want to be, and how they plan to get there. Successful initiatives have found a way to be financially self-sustaining.
Cluster initiative success, however, can be blocked by a number of barriers. Among these are:
- Market failures.
- The “public good and free rider problem.” Clusters initiatives require a lot of effort, and there is a tendency to hold back and “let someone else do it.”
- Lack of trust among competitors or across different cultures.
- Lack of knowledge. How to start, who should be involved, how to make it successful.
- Weak financial resources.
- Competition from other priorities.
State and local public-purpose organizations, he said, take active roles in starting cluster initiatives. Sometimes these are governments, sometimes ad hoc organizations (e.g., the Massachusetts Medical Device Council). Other places, such as Oregon and South Carolina, have set out to stimulate economic development across a number of clusters, which requires a broader program.
Such state-wide experiences, he said, suggests a role for the U.S. federal government as well. Around the world, he said, national governments have played a central role in cluster formation, which is not the case in the United States. Of 31 European Union nations, 26 have a cluster development program at the national level, as do Japan and Korea. Further, the EU even operates a European Cluster Observatory, which maps clusters across the continent.
National and sub-national programs, he said, play different and complementary roles. Sub-national programs are “on site” and can easily promote site synergies. A national program can provide regional groups with information, such as knowledge about best practices and financial resources. Knowledge is a public good that needs to be created only once and can then be shared. A federal program can also transcend political boundaries and stimulate nationwide coverage.
The U.S. Government’s Absence from Cluster Formation
He said that the federal government’s absence from processes of cluster formation is partly a function of historical trends. Since the 1940s, national economic policy has focused on managing the economic cycle—issues of growth, recession, and inflation. Economic policy has not focused on the institutional foundations of the economy, which seemed to be stable. The economy was largely manufacturing based and dominated by a relative handful of major corporations. They were based in well-established regional clusters where they were thought to be immune from foreign competition. In the 1950s and 1960s, the federal government took several steps to enhance economic activity by creating new structures, such as the National Science Foundation, Federal
Highway Administration, and Small Business Administration. In the 1960s, the government focused on equity, trying to lift certain groups of workers (e.g., the Employment and Training Administration) and regions (e.g., the Economic Development Administration). Both approaches used structural, prescriptive, top-down policies.
In the 1980s, the economy was shaken when it lost 12 percent of its manufacturing jobs. National economic policy, he said, attempted to respond with steps that were ad hoc, siloed, and uncoordinated, such as creation of the Technology Administration, Manufacturing Extension Program, and the Workforce Investment Act. He said that the jointly authored paper mentioned at the beginning of his talk had tallied some 250 programs, budgeted at $77 billion that had been created to “try to staunch the bleeding, to move forward with some effectiveness in regional economy policy.”
Time for the Government to “Enter This Space”
However, he said, the national economy continues to be “macrofocused” and to lack a competitiveness strategy. It has failed to recognize that “national competitiveness is a function of regional competitiveness,” and that “regional competitiveness in turn is largely a function of cluster competitiveness.” Hence, he concluded, we have had no federal policy on clusters. He recommended that it is time for the federal government “to enter this space for purpose of a stronger national economy,” and that government participation would require an approach to managing the economic cycle and economic structure in more integrated ways. He suggested several principles for this approach:
- First, federal programs should be flexible, bottom-up, and collaboration-oriented, rather than top-down, prescriptive, and input-focused. In an analogy from physics, he said that federal policy should shift away from “Newtonian physics—the world as a machine” toward quantum physics, “a function of millions of independent decision makers and much uncertainty.” An effective role for government is to catalyze the decision makers and to increase the probability of effective action at the “bottom.”
- Second, the government should use a kit of diverse tools, including information, knowledge, and grants to promote effective clusters. Information, which is inexpensive, should be the first tool used, grants the last.
- Third, a federal effort should be funded at a level appropriate to need. Strategically, he said, this should include a way to bring some of the energy of “those other 250 programs,” which are already funded, into the process.
He concluded by saying that the federal effort should build and rely on the capacity of state and regional organizations so they themselves catalyze competitiveness locally. Federal policy should also link,
leverage, and align existing federal programs that support regional economic development. “This,” he said, “creates synergies among what is. This it not a small challenge. It is going to take nothing less than a new federal culture.”
