On manufacturing productivity

Some people criticized my last post on the grounds that we’d expect productivity growth to have slowed down in recent years because the economy has shifted from manufacturing – where productivity gains are easier to achieve or measure – to services, where they are less easy to achieve or measure.

My chart helps to speak to this objection. It shows productivity growth for manufacturing only. I’m taking Bank of England data since 1948, updated for last year from the ONS.

This shows that productivity growth did accelerate in the 80s: ten-year growth peaked in 1981-91. You might read that as evidence that the Thatcher reforms did raise productivity. However, in recent years productivity has slowed. In the last 20 years, it’s grown by 2.4% per year, which is almost a percentage point less than it grew in the 1953-73 peak.

If neoliberalism did raise productivity growth, it did so only briefly.

I’ll not accept the objection that productivity growth slowed recently because of the financial crisis. True, it did, through several channels. But the crisis was not an accident. Crises are an inherent feature of capitalism. Claiming that productivity would have done well but for the crisis is like a football manager claiming that his side played well except for the six goals they conceded.

What’s more, the productivity slowdown recently is surprising because two things should have raised it.

One is offshoring – both outsourcing and the displacement of low-wage manufacturing in the UK by cheaper imports. This should have raised productivity in the same way that dropping a team’s worst batsmen would raise the side’s batting average. You might detect this effect in the 00s, but not more recently.

The other is robots. There’s not much evidence in these data that robots and digitization are raising productivity.

These data, I think, leave the debate open. They are consistent with my claim that neoliberalism is now depressing productivity. But they’re also consistent with the notion that perhaps neoliberal reforms did temporarily raise productivity, but that effect has now been outweighed by secular stagnation effects: less innovation, diminishing returns and suchlike.

I’m surprised and disappointed that we don’t hear more free market pessimism – the idea that free markets are a good thing but their effect on growth is offset by the dead hand of diminishing returns. This was David Ricardo’s view, and it seems to me a reasonable one now.

Comments

[Ian Dykes]

Manufacturing jobs are boring and require a certain mindset. Sending the jobs abroad means those workers either a) do more intelligent jobs worse; or 2) are permanently unemployable. A graduate will not do a manufacturing job at same rate of productivity as a lower educated worker as the graduate will have attitude issues. Hence dropping productivity from educated workforce.

[Martin Holterman]

Isn't this a case of the Jevons paradox in reverse? Productivity improves, so the related products get cheaper, so more consumption (and therefore weight in the productivity metrics) switches to other products?

[Dave Timoney]

The improvement of productivity in the 80s was in part simply a return to trend following the negative impact of the oil-price shocks of the 70s, though higher business churn (i.e. exit/entry) following the carnage of the early-80s will also have played a significant part in this.

Technology (digitisation) did provide a boost to manufacturing in the 1990s, though again this was amplified by the churn occasioned by the recession early in the decade. It's possible that the current low impact of technology on manufacturing productivity growth reflects the limited company churn since 2008 - i.e. extend and pretend and weak wage growth pressure are bigger problems than a lack of innovation.

PS (and in response to Ian Dykes), manufacturing jobs are not intrinsically boring. People like making things and solving problems. In comparison, service industry jobs are often tedious and more alienating (in the sense of not seeing the fruits of your labour). While some have a high anxiety/satisfaction spectrum because of direct human interaction, a lot involve screen-based admin and old=fashioned paper shuffling.

[tired of armchair "experts"]

"Manufacturing jobs are not intrinsically boring."

Clearly From Arse To Elbow has never worked in manufacturing.