The Economic Outlook and the Prospects for Automation

In reconsidering the hugely influential 2013 Carl Benedikt Frey and Michael Osborne study The Future of Employment it is worth remembering the contrast between what the study's authors were supposed to have said according to its publicity, and what the actual study said. The press (and those commentators who took their understanding of the study from it, which included a great many "experts" who probably could have understood the report but were too lazy to read it) took it to mean that half of the country would be out of work by the 2030s. What the report really said was that, according to the authors' method of evaluating the matter (which undeniably contained a great many highly uncertain assumptions that can now seem optimistic, entailed unavoidable simplifications in its task assessment that seem wide open to question, and produced a number of projections that frankly struck me as odd) half of the country's jobs had a two-thirds chance of proving to be automatable in principle over an unspecified period of time that may well be a decade or two--a very different and much more modest claim indeed. After all, even if they were right about the in-principle feasibility of that automation being there, the perceived or actual cost-effectiveness of automating particular tasks, the problems of fitting a task into the larger functioning of an industrial or commercial enterprise, customer attitudes, the inclination to let others take the lead and see how they do, the hesitation to invest in the newest and latest when one already has a working set-up, etc., etc., etc. would slow down the practical engineering and implementation of the possibility.

Moreover, the broader macroeconomic situation would count for a lot. A situation of slow economic growth and frequent and severe recession, high unemployment and underemployment, slight and ineffectual annoyance to employers from government regulation or organized labor, mean a "disciplined," low-wage work force--workers being that much cheaper and easier to employ, and automation having to be that much more cost-effective to be an attractive option than if growth is brisk and steady, the job market tight, standards with regard to labor conditions stringent, workers assertive, wages high. Likewise a situation where it is easier and more attractive to try and make money speculating on assets like stock or real estate is one where investors have that much less incentive to put their money toward some R & D project promising a better robot.

Which of those scenarios sounds more like the economy we have known these past decades? And, all other things being equal, expect in the decades to come?

Considering the obvious answers to those question I find myself wondering how much further we might have come in different circumstances--and suspecting that, all other things again being equal, we will progress much less swiftly in this important area than we might be able to do in other circumstances.