1/ Have been reading about tech clusters (in the broadest sense of technology, not just IT), and a couple of notions come to mind. Main one: seems like the rate of cluster formation has slowed down enormously over past 2 decades and I can't think of a good reason why.
2/ Think about it: from 70s to 90s you had clusters popping up all over the place in out-of-the-way areas, both in US and Europe (Nice, Austin, K-W, etc). I can't think of one which has really developed since 2000. Dynamic since then has been to reinforce existing clusters.
3/ Or worse, relocate parts of clusters to places where costs are stupidly high (eg Toronto) which makes zero economic sense. No one seems to be starting new industries in out of the way places anymore and this must contribute to general lack of economic growth/dynamism
4/ A few possible answers: a) an increase in anti-competitive behaviour and increasing monopsony effects, b) a change in "innovator" behaviour, in that they refuse to go to out-of-the-way places for cultural reasons (lets call this the "Florida effect")
5/ But there's another, weirder, possibility. New clusters usually take off when new industries are being born. Maybe the rate of new *industry* creation is down in the last couple of decades (the "IT ate everything" effect). I don;t know how to measure/test this though.
6/ Discuss.
7/ OK wait, there's a corollary maybe. Most clusters form because someone has worked out some new way to serve consumer demands or feed government procurement. None, so far as I know, come from government waving a magic wand and saying "let there be clusters". But govts still try
8/ One way they try is to increase research $ to universities, because research unis are at core of a number of N. American clusters (though not many Euro or Asian ones). But if secret to clusters is that it happen in completely new industries, there are implications for policy.
9/ And the main one is: whatever you give them money for should on no account be in fields related to even moderately mature industries. It needs to be in fields with almost non-existent industries and almost no practical short-term applications.
10/ That is, if you want to have a genuinely self-sustaining innovation cluster. If you just want to attract some foreign capital to do a little piece work in areas dominated by others, we can just carry on doing the stuff we're doing.
