To see why, consider a set of statistics constantly used by our politicians. They regularly tell us that Germans and the French are 36 per cent and 30 per cent more productive than we Brits are in terms of output per hour. According to Labour politician Chuka Umunna, “it takes on average a British worker until Friday to do what equivalent workers in Germany and France will complete by the end of Thursday afternoon”.
Read that sentence again. The key word is “equivalent”. There are at least three reasons why these figures might be misleading.
First, countries with restrictive policies (high minimum wages or heavy redistribution, for example) may have inflated measured average productivity simply by virtue of low‐skilled workers not having jobs. The UK unemployment rate is 4.7 per cent and we have high levels of inward migration. In France, the unemployment rate is 10.5 per cent. At least part of the difference between these countries, both looking at output per hour worked or indeed output per worker, can then be explained by the composition of the workforce.
Another crucial difference is the number of hours worked. The OECD calculates that in 2015 the average Brit worked 1,674 hours, compared to 1,482 hours in France and 1,371 hours in Germany. Given workers tend to be more productive at the start of the day than at the end, we’d also imagine that those working longer hours would tend to be less productive by the end of the week. Their average productivity would be lower because of working more hours, making deviations of GDP per hour worked with countries with shorter average working weeks starker still.
That Brits work longer hours on average is one reason why the productivity differential falls to 11 and 15 per cent with Germany and France when we look at output per worker rather than output per hour worked. Provided choices are freely made, though, it isn’t a “problem” that Brits decide to work longer but less productively — it’s a choice for which they are compensated.
A final reason why measured average productivity figures might differ between countries relates to the top‐end of the labour market. If many higher‐skilled older people and women tend to freely choose to spend less time in formal employment, then that again can drag down measured productivity, even though people might be “better off” in the sense of being able to fulfil their own preferences.
None of this is to say that productivity does not matter. It clearly does. Nor has anything I’ve written here implied that the recent productivity slowdown in the UK has been caused by the changing composition of the workforce (indeed, there is some evidence that, purely in terms of skill level, recent trends should be putting upward pressure on productivity levels). What they do show is that targeting some aggregate average measure of productivity with policy is fraught with danger.
Rather than targeting “boosting productivity”, politicians should instead seek to tackle problems directly. If high house prices restrict labour mobility and job matching, liberalise planning laws. If there are policies which actively deter women and old people from staying in the labour force, change them. But do not look at the outcomes of people’s free decisions to move, work longer hours or exit the formal labour market and assume that a remaining measure is necessarily a problem.
Low labour productivity can be a symptom of a problem, and recent trends suggest it may well be, but this need not be the case.