in which we accidentally rediscover Kondratiev

Stumbling Chris notes that, unusually, productivity has fallen sharply in the UK during the recession, and works through various possible explanations without really hitting on anything. He wonders if real-business-cycle theorists might have a point, and the recession be triggered by a real underlying fall in productivity.

In the comments, I suggested that if there was such a thing, it was perhaps the consequence of the housing and financial bubble. This could work in two ways; firstly, productivity is a measure of output per unit of input, an efficiency in engineering terms. Typically, this is output measured by value, per man-hour. There's obviously an issue here with regard to bubbles; if the goods you produce are for some reason overvalued, like houses or collateralised debt obligations of asset-backed commercial paper, you will show unusually high productivity.

Secondly, there's a dynamic issue; part of the impact of a bubble is to draw capital and resources in general into the bubble sector and away from others. And, so far as the bubble causes a productivity illusion, this is rational in a short-term sense.

Once the inevitable happens, we'd therefore expect to see a couple of phenomena - one would be a fall in productivity in the bubble sector, as the value of output is revised down, and another might be a fall in productivity across the economy. Why? Unlike financial capital, physical capital and labour have switching costs. The end of the bubble requires the movement of resources out of the bubble sector - obviously - and into other sectors according to the new productivity picture. But this is a process that has significant costs and takes time.

You might expect a J-curve effect; it's faster to close things down and to be sacked than it is to start things up and find a job, and some things are not easy to repurpose. People need to learn different skills; capital goods may turn out to be completely useless in any other context. Come to think of it, one form of physical capital that can be very difficult indeed to reuse is a building, and we're in the guts of a recession driven by a massive property bubble.

An interesting thing here is that hugely mis-estimating productivity and therefore making very bad investment decisions is what the state is meant to be good at. The Soviet nail factory joke is exactly equivalent to the property developer who sinks jillions into newbuild dovecots or huge light-construction monsters 40 miles from town. Both of them fail because their outputs are catastrophically mispriced. Perhaps the difference is that GOSPLAN didn't get a government bailout!

Of course, capital allocation and reallocation pass through the capital market, rather than GOSPLAN. Which is why financial innovation causes bubbles. It would also explain why recessions that begin with a financial crisis tend to be the worst ones; the mechanism of reallocation itself fails, and again, propertyfail seems especially hard to fix, as well as being both unusually harmful to the public and unusually unlikely to create anything useful.