By Robert Hickson 09/09/2015


Following on from my last post about robots, evolution, and metaphors, the Journal of Economic Perspectives has, for a non-economist, a surprisingly good pair of other articles in the same issue.

Both Mokyr et al. and Autor provide good historical perspectives on how economists and others have a poor track record of predicting future jobs, and that the current concerns about technologies displacing human workers is nothing new.

Autor notes that discussions of the workforce impact of technologies focuses, not unnaturally, on what is lost rather than how such innovation may have longer-term social and economic benefits.

Obviously that’s not much comfort to those losing their jobs right now. But it’s not necessarily bad from a societal perspective. Autor states that “societal adjustments to earlier waves of technological advancement were neither rapid, automatic, nor cheap.” Over the last few centuries the nature of work has never been static for long, and working conditions and pay have generally improved.

Middle skill peril

There is no doubt, though, that what Autor calls “middle skill” jobs (such as clerical work, labouring, and sales) are disappearing, while demand for both lower and higher skilled jobs (food service and cleaning jobs at the lower end, and technicians and “professionals” at the upper) are increasing, or at least declining more slowly. With some of the higher skill jobs more at risk than the lower ones.

Robots and other automated systems won’t be taking all our jobs any time soon. But there is the risk that we could get a polarization of employment; lots of lower paid jobs, and fewer higher paying jobs.

However, Autor also helpfully points out that a “job” usually involves lots of different tasks. So, just because one or two of those tasks may be automated doesn’t mean that the job as a whole can be taken off a human. Human car assemblers now work alongside robots. So, too do people at Amazon warehouses, though they often don’t do so well as auto workers.

What is needed is not fatalism that jobs are disappearing, but a commitment to invest in developing the “human capital”, so people have the right skills to complement new technologies .

We know what currently are good employment opportunities, and retrospective analysis can identify which types of degrees may give less favourable employment opportunities.

But not everyone in the future will want to be a data analyst or engineer. Social skills, particularly if combined with good mathematical skills, may be where some of the top jobs of the future will be.