More radical consensus – the innovation strike

I was recently pointed to NESTA’s report on innovation in the British economy and our recovery, or otherwise, from the Great Recession. This is a pretty good example of the first two functions in the post on thinktanks – it’s assessment and comparison first, with ideas arising from them – but that’s not really my point.

It covers the usual topics (clusters, networks, and intellectual property), but I was pleased to see that it is sceptical of patents and copyrights in general. For a counter-example, check out the Acemoglu & Robinson paper Noah Smith savages in a recent blog post, and especially Smith’s comments thread. Quite apart from the open source criticism of patents, they are anything but an internationally comparable metric because the definition of what can be patented, the criteria for getting a patent, the enforceability of the patent, and the value of the patent monopoly once you have it vary dramatically between jurisdictions. Obviously, patenting your invention or “invention” in the US is more valuable than patenting it in Sierra Leone. You can patent software and even “business methods” like clicking a link to buy something in the States, and anyone who follows the tech industry sooner or later has occasion to think that in fact you can patent anything and get away with it.

NESTA’s analysts took note of the huge literature on industry clustering, and also that efforts to deliberately create clusters have a horrible record of failure. On the other hand, they say, public policy often does have a major effect on clusters once they emerge.

This is an interesting and possibly profound point; the caricature of innovation is the eureka moment of invention, but the notion of invention is a questionable one. Inventions are very often collective rather than individual, emerge from incremental development, and it is quite common for the event of invention that actually matters to be the second or third, or to be an invention in something parallel.

You remember James Watt, not Newcomen or Savery. Frank Whittle’s jet engine wasn’t actually the first attempt at a gas turbine for aircraft, although it was the first that worked at all well, and it might never have worked properly if Rolls-Royce hadn’t got involved and put in the capital and expertise to set up a dozen of them on parallel test and hammer on the hundreds of hours of development running and continuous improvement they needed to work out the kinks and arrive at something useful. Packet-switching was independently invented at least three times in the USA, France, and Britain round about the same time, and possibly also in the Soviet Union, but the degree of scientific exchange between the people involved was such that it is probably a category error to try to identify the inventor, especially as people from NPL and INRIA ended up contributing quite a lot to the work on ARPANET.

If there’s more to the innovation process than just inventions, this implies that public policy has more to work with, without trying to guess what is going to get invented in advance.

An important point in the NESTA report is the role of people and skills. If there is a network or a cluster or something of that sort, this usually means in practice that the people involved are closely associated and all know each other, and that it is easy to find people with recondite skills (and conversely, that acquiring those skills is worth doing because it’s easy to find a job). This surely implies that the dichotomy between “stimulus now” and “long term development” is a false one.

One of the reasons things like Silicon Valley, the south-west German automotive cluster, or the Midlands industrial network that nurtured Power Jets Ltd. happen is that both formal, trained skills and informal, tacit knowledge build up over time through practice. In Whittle’s memoir, he mentions repeatedly that he solved technical problems by challenging craftsmen metalworkers to do things that hadn’t been done before. Very often, the toolroom was able to respond. Similarly, faced with problems machining turbine blades, Power Jets turned to the British Shoe Machinery Company of Leicester, just down the road, for advice on production engineering and designing specialised machine tools. (In fact, there’s a really good book or at least a blog post on the topic of Power Jets considered as a funky tech startup, but I think Jakob Whitfield may be writing it.)

In that sense, we really can’t afford more unemployment. If it takes 10,000 hours to make a craftsman…

A lot of this debate ends up being about institutional changes, and the caricature of these is always, at the moment, the idea of a public sector bank of some sort devoted to small and medium businesses. They make a variety of common-sense suggestions on how to organise such a thing, but the interesting bit was their analysis of the problem. This is where NESTA’s report gets surprisingly radical, arguing that the rate of investment in innovation has been low for years in the UK, plunged (of course) in the crisis, and hasn’t recovered. This is fascinatingly close to the Marxist concept of a capital strike (they even call it an innovation strike), where the capitalists stop investing and therefore bring about a permanent state of crisis until they get what they want. In the light of the revised estimates of the output gap, it’s also very much on point.

Putting them together, you might say that unemployment is the symptom of the innovation strike. It’s another example of the radical consensus, even if NESTA then swerves away into suggesting that we use the LTE spectrum auction to capitalise a new institution – rarely has the skin of one bear been sold so many times, and I for one doubt the takings will be that huge.

Another really interesting point is that the services economy spends very little on innovation. In some ways this is inevitable – much of it is affected by the cost-disease, and what people consider to be great service tends to be the opposite of the standardised and routinised processes that drive productivity in manufacturing. But I do wonder if this is assuming the conclusion, and possibly doing so for reasons that have a lot to do with class. If you aim to have disposable, zero-training workers you by definition give up on the craft ethos and you eliminate the accumulation of informal knowledge as a source of strength.

This was roughly what I was driving at with the Politics of the Call Centre series – if you declare part of the workforce a disposable cost base, you’re also writing off any chance of doing better.

Obviously, doing better in the services sector is very relevant indeed to the public sector, as an enormous provider of services. But do we see any spillover from public-sector innovation to the wider economy here? We see about as much of it as we see spillover in the other direction. Here’s a bit of provocative wanktanking for you: now that there’s a Service Design movement, why isn’t there a Council on Industrial Service Design and why aren’t public service organisations expected to get its kitemark?