So over the past few days I’ve found myself in two debates over the impact of automation on jobs with the New Statesman’s Alex Hern and the inimitable Frances Coppola.
In both I wanted to discuss the demand impact of automation, which has thus far been largely ignored in the “robots ate my job” discussion. Basically my point is that the consequences of high unemployment on demand would ultimately be enough to offset the case for automation. Or looked at another way, businesses might choose to retain employees even if the cost per unit is higher than with an automated alternative.
For me, this has important implications. In the short term the fear of demand destruction is likely to do little to dissuade individual businesses from trying to increase productivity through automation – so some of the fears about job destruction are at least theoretically justified.
In the longer term, however, the costs of rising unemployment should help to erode profit margins faster than automation can maintain them. As persistent mass unemployment would be intolerable for both sides – either government transfer payments would have to be significantly increased or circumstances would risk profound social upheaval.
One response to this could be a temporary expansion in the public sector as the private sector adapts to technology “shocks”. There are still plenty of long-term productivity gains that can be achieved through large-scale infrastructure improvements and meanwhile it would allow new types of industry to form.
Alternatively, the private sector may seek to stagger investment in new technology in order to avert a demand shock. To an extent this appears already to be happening, particularly in the energy sector, and has been abetted by protectionist measures by governments (usually at the behest of lobby groups). As Frances points out, this type of labour hoarding comes very close to the Luddite argument.
It is fair to say that both Alex and Frances were sceptical of the demand picture I was painting. I concede that the reality could be much more disruptive and far less smooth a process than it suggests, but I still believe it could ultimately become an important factor – if it’s not already so.
And yet even if you accept the thesis is does not seem a sufficient explanation for what is actually happening in UK labour markets. Figures out from the ONS earlier this week showed that unemployment fell again between September and November, even as early GDP estimates suggest growth was pretty much flat over 2012.
Ben Broadbent, External Member of the Monetary Policy Committee, gave a speech last September on productivity and the allocation of resources. In it he laid out the current debate on the causes of the surprising resilience of Britain’s labour market:
“One explanation for this is labour hoarding. Because it is costly both to fire and to re-hire employees firms may have hung on to staff in an expectation that demand will pick up. Another is that underlying productivity growth has slowed. In that case, weak output may not be the result purely of weak demand. It may instead reflect an independent hit to supply, one that explains why employment and inflation have been relatively high.
As extremes, these two hypotheses can have very different implications. If the first is true, then without a strong pick-up in economic activity, employment and inflation are likely to decline. Finding themselves in the position of a cartoon character that has run off the edge of a cliff, but is not yet aware of the fact, firms will at some point have to face reality and trim their under-employed workforce. Cost growth and inflation will then fall. If, instead, supply is (and remains) the problem, we could continue to see below-par output growth without any such implications for employment or inflation.”
Broadbent’s claim is that there is an element of both at play, as the fundamental problem is a misallocation of capital. While labour has proven itself flexible in the aftermath of the crisis, ironically capital has proven sticky.
This has been caused, he contends, by a combination of uneven demand between sectors and an impaired financial system unable to efficiently reallocate capital - hence the growing number of stories about “zombie companies” kept going simply to service debt. Although this would have a negative impact on productivity and output in the short run, it should also mean that there are good reasons to expect these to ultimately recover.
However, I think here there is also an issue that Broadbent does not address. In its efforts to avert a systemic collapse in financial services the Bank of England, along with the other major central banks, has in effect given an open-ended commitment to support asset prices.
In doing so they may be exacerbating the causes of misallocation of capital, in particular through a mispricing of risk. If central bank action is helping keep unproductive areas of the economy on life-support then default risk is effectively taken off the table for investors, encouraging unproductive investment.
Unfortunately this has left successful companies unable to gain access to capital, which is holding back investment in making their businesses more efficient (eg investing is new technology). Employing more workers to meet demand is a partial solution to this, but the current trend could quickly run out of steam if the underlying economy continues to stagnate and real wages continue to fall.
As such while it is understandable that central bankers have focused on the problems within the financial sector to prevent a system-wide collapse, it is not a costless strategy. The problems of “zombie” loans and the rebuilding of shattered confidence between institutions are likely to take many more years to resolve themselves. In the meantime Britain’s economic potential is being squandered.