UK productivity is falling. Recent figures show that hours worked rose by 0.6 per cent in the three months to April but GDP rose only 0.3 per cent in the period, implying a drop in output per worker-hour. This continues a long-term stagnation – productivity has grown only 0.2 per cent per year since late 2007, compared with 2.3 per cent per year in the previous 30 years.
This matters enormously. If productivity doesn’t grow, the economy can only expand by increasing employment or hours. And while there is more excess labour than the 3.8 per cent unemployment rate would suggest – there are 1.8m people outside the labour market who want a job and many more who would like to work more hours – there’s a limit to how fast employment can grow. This suggests that economic growth will be weak, which means weak growth in dividends.
Unless, that is, productivity growth recovers. Here, however, history sends us a depressing message. It tells us that productivity growth tends to be persistent; strong growth leads to more strong growth, and weak growth to weak. Since 1760, the correlation between productivity growth in one 10-year period and in the next has been 0.69. Much the same is true if we look at five-yearly growth or at more recent periods.
This is consistent with the fact that financial crises tend to have long-lasting adverse effects on growth. Coen Teulings and Nick Zubanov, two Dutch researchers, looked at 89 financial crises since 1974 and concluded that “the loss from a banking crisis is likely to last for a long time".
One reason for this is a scarring effect. Memories of sudden slumps in demand and of a withdrawal of credit lines deter capital spending and entrepreneurship even years later. And depressed animal spirits can be self-fulfilling; an economy that lacks dynamism is one in which entrepreneurs will be loath to start new businesses or invest in new technology.
A second reason is that the policy and institutional environments that foster or retard productivity are themselves persistent. The strong productivity growth of the 1950s led to further strong growth in the 1960s in part because the same things that encouraged investment in the 1950s (such as full employment) also did so in the 1960s. And the policy climate today is hostile to dynamism. Both the US and UK governments are prioritising nationalism – tariffs and Brexit – over economic growth.
Thirdly, some of the factors depressing productivity growth in specific industries are themselves persistent. Big pharmaceutical companies have lost productivity because patents on some high value-added drugs have expired, and since the crisis the financial sector has stagnated: more compliance officers might make the banking system more stable but they don’t make it more productive.
You might object here that new technologies such as machine learning, AI and robots will soon give us a fourth industrial revolution.
Even if such technologies are adopted, however – which is a big if – productivity growth won’t rise much for years. To see why, think about the first industrial revolution. Steam engines, the spinning jenny and the water frame transformed manufacturing. But overall productivity stagnated in the late 18th century. The reason for this was brute maths. Such technologies initially represented only a tiny fraction of the aggregate economy, which was largely agrarian and stagnant and so didn’t raise total productivity even though they did so where they were adopted. The same will be true of future new technologies.
We should, therefore, assume that overall productivity will continue to stagnate.
