On May 18, George Magnus had tweeted that Chris Giles’ article on the age of the robots was a good response to the Luddites and that they had to repent! Strong words. I saw that tweet only two days ago and I made a mental note to read that Chris Giles’ article, anticipating some solid arguments as to why we should all rejoice and join in the robotic revolution or robotic revelry, robotically without questioning it. In the end, the article was nothing of that sort. Magnus’ endorsement is a puzzle.
Look at two of Giles’ key paragraphs:
The first is that the digital technology has not yet matured but is on the cusp of profoundly changing employment patterns and productivity. In this world, many jobs will become automated, providing much by way of productivity gains and economic growth. This would enable everyone to be better off so long as public policy tools were devised to ensure a fair distribution of the spoils.
The alternative is that productivity growth remains close to zero, employment rates stay high and the main driver of increased economic activity comes from forcing people to work longer. A lack of productivity growth would bring profound challenges to public finances, requiring ever-increasing prosperity to fund both accumulated government debt and the liabilities incurred by an ageing population. [Link]
Both the paragraphs make elementary errors. The first para assumes that automation would lead to productivity gains. Yes but in the sectors that are automated. But, let us proceed. Let us assume that there is economy-wide productivity gains. A very wrong assumption because economy-wide productivity gains would depend on whether the workers displaced by automation migrate to sectors with low productivity or high productivity and boost output further. If it were the former, overall productivity can drop. Then, there is the other fallout, on employment and hours worked. Even if productivity gains were positive, hours worked could decline and more than offset productivity gains, leading to a decline in growth. Mr. Giles makes an elementary error. Other things do not remain constant. If productivity gains accrued to the economy, hours worked would most likely not stay constant, especially if robotic technology diffused throughout the economy.
Mr. Giles thinks that he has anticipated this argument by putting in a caveat that there should be a fair distribution of spoils. Oh, yes. That is easy. Haven’t the elites been doing it so well in the last thirty years?
Then, there is his second paragraph. His alternative is that technology does not diffuse enough and that productivity gains are zero and growth comes through jobs and hours worked.
The problem with his logic is that these are not two alternative scenarios. They could be just one scenario. There are no neat boundaries between his optimistic technology diffusion scenario wherein everyone lives happily ever after as long as the spoils are fairly distributed and his pessimistic scenario wherein workers toil away for pittance. This can very well happen as part of his optimistic scenario.
Far better to accept that there are large unknowns than to paint a simplistic scenario and to spread false cheer.