How can we make businesses more innovative? That’s easy isn’t it? We just group them together into clusters (preferably in science park developments) and it will happen… won’t it? The trouble with this cluster theory of innovation is that it confuses cause and effect. When we see a successful cluster like Silicon Valley it’s tempting to asssume that the clustering made these firms successful. Jumping to the conclusion that clustering is a cause of innovation success has triggered a lot of government expenditure to create these ‘hot spots’, but what if clustering is an outcome of succssully innovating firms? Are clusters really an innovation ‘red herring’?
The trouble is that the evidence for cluster theory is really shakey. In a previous post I mentioned the cargo cult analogy for initiatives like clusters and the weight of evidence against clusters as a precursor to innovation. A recent study of 1600 Norwegian firms concluded that:
The results indicate that firm innovation in urban Norway is mainly driven by global pipelines, rather than local interaction. The most innovative – both in terms of basic product innovation and radical product and process innovation – firms are those with a greater diversity of international partners. Local and even national interaction seems to be irrelevant for innovation.
The full paper is definitely worth reading. It is robust research and if you don’t enjoy the stats then skip to the conclusions. The findings make sense to me. If we are thinking of innovation as an open ecosytem then why should all the significant people and firms be in one city. The world is a big place. Australia, for example, is 2% of the global economy. Why would any firm want to priortise connections in one city that is just a drop in the international knowledge pool?
The really interesting part of the study is the relationship between managerial mindsets and the deliberate formation of international linkages.
…what this study has demonstrated is that the attitudes of individual managers play an important part in the innovative capacity of their firms. Open-minded managers without excessive regional orientations are often in charge of firms which develop a greater number of international interactions of the sort that promote greater innovation.
This has implications for developing innovative industries. For most of the 20th century, Sweden was an economic backwater with low standards of living. One of the really smart things that the Swedish goverment did was to sponsor students and young professionals to work and study overseas. In developing connections through this program, they also encouraged firms to be more innovative.
I’m currently reviewing a research paper based on a sample of Canadian firms that also finds a link between innovation and internationalisation. In a survey of firms in Brisbane, Australia (the Brisbane Innovation Scorecard) we also found a close relationship between innovative firms and connection to international markets. At least one study shows that international activity can preceed innovation rather than the usual assumption that innovative firms grow to compete in international markets.
Travel broadens the mind…. and makes you more innovative too.
Hi John,
Enjoyed the post – and very good points.
Coming from an analytics background, I am going to suggest a different interpretation :-).
I think there is nothing inherently wrong with the cluster theory of innovation. What is misleading is the assumption that the clusters should be based on geographical proximity. In modeling, when one does cluster analysis or factor analysis – the variables chosen play a key role in what types of clusters emerge or what type of factors are extracted.
I think the case is similar here. It is true that Silicon Valley might appear to be a “one city” cluster. But look deeper and one will see a global composition within that “city” — unlike in many cases where there is a lack of such diversity.
Bottom line, you are absolutely correct in your conclusion that there needs to be diversity and internationalism to bring in fresh perspectives and generate new ideas. If you have a rich playing field within your city walls, that is good – if not, seek out from outside your borders.
Best,
Ned
Hi Ned:
that’s a sound proposition. I can’t fault the logic. It depends how diverse the cluster is and how connected it is. I still think that the result that international connections matter more than local stands up but you make a really nice refining point here.
John
I hope you at least considered the obvious (?) title for this blog post – “clusterf*cked”?
Interesting stuff. There’s an insane amount of boosterism/Florida citing/cargo-culting in Manchester, UK around this whole innovation cluster thing.
btw, saw this post and thought you guys might be interested
http://keithsawyer.wordpress.com/2011/04/05/collaboration-drives-human-evolution/
Hi Dwight
thanks for that link. It’s a really interesting perspective on social networks.
Cluster effed didn’t come into mind as a title for the blog but I must admit that I’ve used the term in talking about some shonky work on innovation clusters (although not in polite company… the journals don’t like that sort of thing).
I’m not against clusters but I think that their benefits are often overstated and the cargo-cult phenomenon driven by desperate governments and witch-doctor consultants is a very real problem.
Here’s a story… Some time ago I worked in Tasmania. As a small island, it has always faced economic challenges and the solution has never been simple. In 1996 the Australian goverment decided to privatise the national telco but it needed the support of an indpendent Tasmanian senator to get the legislation through. In true realpolitik style, the senator named his price at just under $200 million for a development package for Tasmania and he got it..
One the govenment got the money, they faced the problem of what to do with it. Cluster thinking prevailed and consultants were hired and think-tanks formed. After some time, the decision was made to create a bioinformatics cluster.
I knew about the bioinformatics initiative before the press conference because someone from the Premier’s department had called me to ask what it was. There was a general opinion that it had something to do with medicine and computers but exactly what was involved was anyone’s guess.
At the press conference it soon become very obvious that the then Vice Chancellor of the University of Tasmania didn’t know what a bioinformatics business looked like and neither did the goverment representatives.
Some years later an investigation revealed that the intitive had created no jobs (except for the consulants) and the $200 million had mostly been wasted.