There seems to be a popular belief that what it takes to create an industry cluster in bioscience or whatever is to pair research with commerce. That is, to find an academic institution doing cutting edge research, and connect it with venture capital and entrepreneurs to start companies to commercialize it. Soon enough, you have a “cluster” of businesses that takes off like a rocket. This is the perceived Silicon Valley model, and no company epitomizes it more than Google, which was started by two Stanford students to commercialize their graduate research.
But is this true? There are many top flight research universities in this country, but few major startup clusters. When major research institutions fail to generate commercial spinoffs, this is often blamed on a lack of venture capital. But is that really the case, or is something else at work?
Anyone interested in this matter simply must read AnnaLee Saxenian’s seminal book, “Regional Advantage: Culture and Competition in Silicon Valley and Route 128“. A social scientist at UC Berkeley, Saxenian lived and worked in both Silicon Valley and Boston’s Route 128 technology corridor. She wondered why Route 128, which started out with far more of a technology business and economic base than Silicon Valley, eventually lost ground to become a clear number two. She sees this resulting from the different social structures that exist in the various areas.
According to Saxenian, Route 128 suffered from a culture that was oriented towards a traditional maturing industry, not a rapidly changing one like technology. This included more deliberative decision making; vertical integration and self-sufficiency; hierarchical, centralized command structures; focus on economies of scale; a high friction job market; geographically dispersed locations; and low levels of cooperation and sparse networks between firms in the region. In other words, all the standard traits of a typical large corporation. While she doesn’t dwell on this point, it also comes across that Boston, probably due to its New England locale with all the history there, was a much more closed society. The social network and hierarchy was more fixed (the phrase never appeared in the book, but I couldn’t help but think Boston Brahmin) and the process of establishing trust and credibility much slower than California. While famous as one of the bluest states, Massachusetts is socially conservative in many ways, and highly risk averse. This is the land of the suit and tie, and the difference between that environment and California casual was more than just a surface thing.
Silicon Valley, of course, was just the opposite. It adopted social structures that were very focused around innovation and time to market. It was open, with rapid, decentralized decision making. Firms quickly specialized, focusing on their core competency, and established close links with suppliers to fill in the rest of the value chain. These links were often such that it was not clear where one company ended and the other began. The clearly functioned on high degrees of trust. Even direct competitors often talk to exchange ideas and help each other solve problems. Here’s a quote:
Competitors consulted one another with a frequency unheard of in other areas of the country. According to one executive: ‘I have people call me quite frequently and say, “Hey, have you ever run into this one?” and you say “Yeah, about seven or eight years ago. Why don’t you try this, that or the other thing.” We all get calls like that.’ (33)
Clearly this is quite unique. I’m not even sure if it’s all legal, but hey, it works for them.
The job market in Silicon Valley is extremely fluid, with people constantly changing jobs, starting companies, etc. It is expected that you won’t stay that long with any given employer. Route 128 operated on the “company man” model and to leave was to show disloyalty, often resulting in ostracism. Since Silicon Valley was a new country with almost all immigrants of one type or another, family history and credentials meant little. What mattered was whether you could perform.
Now of course it was almost entirely men, originally white men, who set this up. The tech industry is famous for being one of the most gender imbalanced. What I found particularly interesting was that many of the founders had Midwestern roots. Again quoting:
This collective identity was strengthened by the homogeneity of Silicon Valley’s founders. Virtually all were white men; most were in their early 20’s. Many had studied engineering at Stanford or MIT, and most had no industrial experience. None had roots in the region; a surprising number of the community’s major figures had grown up in small towns in the Midwest and shared a distrust for established East Coast institutions and attitudes. They repeatedly expressed their opposition to ‘established’ or ‘old-line’ industry, and the ‘Eastern establishment.’ (30, emphasis added)
The many examples of engineers with humble origins who became millionaires by starting successful companies had no parallel in the more stable social structures of the East. Jerry Sanders, founder of Advanced Micros Devices … grew up in south Chicago, the son of a traffic light repairman. (38, emphasis added)
To digress for a moment, remember how I said the contrarian, ornery Hoosier/Midwestern attitude is, in the right context, a huge strength, not a weakness. This shows that in action. These guys didn’t toe the conventional wisdom line. Instead they created a whole new business model. I’ve got to believe the Midwest mindset played a huge role in making this possible. The unfortunate thing is that they had to leave the Midwest to do it. Imagine if they’d stayed home and made it happen around one of the great engineering schools there? Alas, to this day Midwesterners often have to leave to turn things into reality. Famously, Marc Andreessen had to leave Illinois to start Netscape, and in fact had U of I actively hampering him all the way. If the Midwest cracks the code on this piece alone, it would be a huge step in the right direction.
(By the way, for a wonderful look at how these Midwesterners invented Silicon Valley and the “elder days” of semiconductor business, see Tom Wolfe’s 1983 Esquire essay, “The Tinkerings of Robert Noyce“.)
These extremely fluid job markets, open social institutions, high trust customer and supplier interactions, and competitor information exchange create an environment of so-called “dense networks”. In a period of rapid change and innovation, these networks, by efficiently distributing information and dispersing risk, create an environment with very rapid speed to market and high levels of adaptability. A traditional Route 128 do it all yourself model simply can’t keep up with the power of this vast network.
It was this network, more than anything, that created the Silicon Valley we know today. The “cluster” we see in Silicon Valley is not an artifact of spatial co-location. It comes from the network. According to Saxenian:
Spatial clustering alone does not create mutually beneficial interdependencies. An industrial system many be geographically agglomerated and yet have limited capacity for adaptation. This is overwhelmingly a function of organizational structure, not of technology or firm size … The current difficulties of Route 128 are to a great extent a product of its history. The region’s technology firms inherited a business model and social and institutional setting from an earlier industrial area. (161-162).
Sound familiar? It describes the Midwest perfectly. What I find interesting is how Saxenian illustrates her thesis not using a struggling Midwestern burg as a case study, but rather Boston’s Route 128, the second largest technology hub in America, home to possibly the greatest collection of universities in the country, with massive access to capital, etc. If this town had its problems, how much more so places without those advantages? It certainly shows the scale of the challenge in building industry clusters.
Obviously, changing the social structure, culture, and institutions of a region is difficult to do. Even positive articles highlight the scale of the challenge. I’ll refer a recent article on Milwaukee startups that I linked that quotes a local businessman saying, with some pride I gather, “Milwaukee is a one-strike-you’re-out town.” That’s not a good thing. Silicon Valley shows that failure and risk taking are good. The way to innovate is to figure out how to try lots of things and to fail quickly and cheaply. If you are overly concerned that you’ll be permanently ruined if your business goes bankrupt, you’re not that likely to take a chance.
It reminds me of a discussion I once had with a friend from Germany. He told me, “We’re the children of the people who stayed” and bemoaned the highly conservative outlook of his countrymen. He noted the extreme reluctance to take risks because in Germany, if you go bankrupt, you’re stigmatized for life. Obvious some of that carried over to heavily German Milwaukee.
I should note that one should not over-internalize Saxenian’s case studies into some sort of cookbook solution. Every city and region needs to find its own unique path to success based on its own culture, institutions, history, etc.
I would be remiss I did not point out a few areas where I was skeptical of the Silicon Valley model. One intriguing factoid from the book was that in 1962 federal government purchases, principally defense related, accounted for over half of Route 128’s sales. Indeed, the area got its start in technology through defense related research during World War II. Could it be that dependency on government contracts is really what caused the dysfunctional culture there? Government largesse encourages rent seeking behavior at the expense of building a competitive business.
Also, Saxenian highlights how the non-business social networks in Boston substitute for the type of technology networks in Silicon Valley. But is this a bad thing? The books paints a portrait of Silicon Valley as a bunch of geeky guys who toil away long hours on tech projects and even talk about technology at the bar when they do go out. It’s like a community of idiot savants. Some might say “get a life!”
What’s more, there is some research that suggests dense networks themselves aren’t a recipe for success. In an thought provoking paper called “Why the Garden Club Couldn’t Save Youngstown” Sean Safford contrasts the experiences of Youngstown and Allentown, both small steelmaking cities. Despite similar dense networks, Youngstown failed while Allentown fared much better. His conclusion that was the dense networks in Youngstown only reinforced an already closed leadership circle who were economically aligned, while Allentown’s served to bridge otherwise non-overlapping groups.
Perhaps to a great extent, the key attribute is less the networks themselves, than the ability of outsiders and new thinking to penetrate them. Silicon Valley’s social structure was open, Route 128’s wasn’t exactly closed, but there were barriers to entry. In a globalized world of ever faster change, the ability to respond and adapt, to process new ideas and react to rapidly shifting global forces, is critical. This puts a bit premium on dense social networks that are also open and flexible.
This is somewhat the thesis also of Richard Florida. He has a somewhat different spin, saying that the economy is now powered by the creative class, and they want to live in places that are open, tolerant, etc. This is his “three T’s” model: talent, technology, and tolerance. The last appears not to be so much valuable in its own right, but for what it says about the openness of social networks. Thus a large number of gays in a community isn’t what drives economic growth per se. Rather, a thriving gay community is a signaling mechanism that lets people know that diverse ideas and people are welcome.
I think we all know places where the social network is impenetrable. This isn’t necessarily a function of size, prosperity level, etc. I mentioned the Boston old money, social register concept. In any number of southern cities, who your daddy is, or what sorority you went to in college is a huge determinant of your place in a social hierarchy. If you don’t come from the right family, the right schools, etc., you can forget it.
Perhaps this explains my Cincinnati conundrum. Here’s a city with better assets than almost any in America, but it is one of the all time relative decline stories in US history and to this day is on a moderately stagnant, slow growth path. Why is that? There was an intriguing study I saw recently called Who Rules Cincinnati? [dead link] This is by an independent researcher named Dan La Botz, who I get the impression is some sort of hard core left activist, so keep that in mind. Nevertheless, he uses a similar approach to the Garden Club study to track social networks in the city, coming to the conclusion that officers of seven major corporations basically run Cincinnati, mostly to that city’s detriment. Another person I know offered the interesting insight that when he meets someone in a bar in Cincinnati, the first question they ask him is where he went to high school. This both indicates a highly inbred culture as evidenced by the assumption one must have gone to high school in Cincinnati, and shows that the school you attended is an important social marker. (It perhaps also shows a lack of regard for higher education).
It could be that the Midwestern cities that have the best potential for future growth are those with the most open social networks, as well as exhibiting other of the characteristics Saxenian cites. I think this would be fertile ground for social science research. It also makes me wonder if perhaps that goes part of the way to explaining the relative success of the Midwest’s larger state capitals. State capitals constantly have people traveling and doing business there from all corners of the state. This flow in and out might potentially prevent a social structure from completely congealing into a small, impenetrable elite. I sense another potential dissertation topic here.
The key takeaway is not to focus on purely the institutional infrastructure (universities, venture capital funds, labor force, etc.) when trying to set out an economic strategy. The local culture, norms, and social practices, and in particular the density and openness of the social networks is critical. Clearly, as anyone who has found themselves mired in a corporate or governmental bureaucratic organization, changing a culture is an extremely difficult thing to do. But it is something that clearly warrants an examination.
This post originally ran on July 13, 2008.