Wednesday, July 05, 2006

Clusters in economic development - twaddle or useful?

Ok Ok I am rehashing a previous post I made on another blog BUT!

I recently heard someone talking about 'local clusters' - as if they operate at a neighbourhood level. Get this people - clusters (where they properly exist, and aren't just the wishful thinking of a public official) don't operate on the basis of tiny geographies. They operate on the scale of regions, often nations, often areas that go beyond national boundaries. So there.

I used to work at Scottish Enterprise, and serviced a lot of teams there as a private consultant in the late 1990s. I wasn't impressed with their clusters approach. What they did:

- identified some Scottish clusters
- spent ages (and big consultancy $$$s) mapping the clusters, mapping weaknesses etc
- spent ages meeting and connecting folks in the cluster together
- did a few specific interventions
- identified some cluster gaps as R&D, invented public sector research labs to specifically service the cluster

What was achieved? after investing in cluster teams and a huge research effort (in my opinion, the research was very patchy - good for some cluster teams, bad in others) - not a lot I reckon. This is because I think the effort and approach was misplaced.

I'd like to express some of my own theories about clusters:

#1 they are no respector of government or administrative boundaries. They operate across all artificial and formal boundaries. Especially in open markets.
#2 very few (none!) examples of successful, self-contained, highly localised clusters
#3 very few clusters created out of thin air
#4 very few clusters created in less than 10 years (take note RDAs!)
#5 must have some kind of comparative advantage for cluster presence (take note again RDAs!)

OK, ok they are engines of growth in certain parts of the world, and Porter's analysis of competitiveness is compelling. But as an economic development tool I think its not been well applied, at all.

Its big use is in understanding the economy and your locality's place in it. If this leads to some intelligence interventions, then that might be a good thing. If it tells you that you are one the outer spoke of a cluster, a branch plant zone etc, that's useful. For example - it can be a great way of looking at the skills needs of your regional economy.

But that kind of analysis is very sophisticated indeed. How many RDAs do you know that intelligently create and target interventions? hardly any I'd say. The clusters approach is way too complex and sophisticated for the average RDA and RDA staff member I'd say. That's why, if you are SE, you spend £5 million on Porter's consultancy company (Monitor) to come and tell you about clusters.... heehee.

Anyhow, just a thought. I have to say that I am both sceptical and open minded about clusters, but they are not the panacea for everyone. And if I hear anyone say the "Anytown XXXX cluster" again, I will probably develop a nervous tic.

My big message - you can't have a highly localised cluster. They aren't valid. Don't try and think you can create one out of thin air. However if you have localised industry specialisms - you might want to think about how they fit into and relate to a wider industry cluster, and there might be some ways of capturing additional economic activities for your region.


Blogger Neil Craig said...

If a cluster is a number of firms all sharing production (a manufacturer of bits supplying to several manufacturers) there seems little reason for them to be located nearby - it used to be said that Swedish cars had more Scottish parts than those made in England. It may be of more use where it is a cluster of people with the same trainingwhich will attract new entrants. In which case the focus of such a cluster would be the University/College/apprenticeship scheme which taught them.

11:03 AM  
Blogger Accidental Economist (Glenn Athey) said...

My point was really that clusters transcend local administrative boundaries. They are no respecter of political boundaries or maps.

There are very few examples of successful clusters that are tightly bounded geographically.

8:05 AM  

Post a Comment

<< Home