Economic possibilities for our future robot overlords

Following up on Erik’s post about working hours in France, in 1929 John Maynard Keynes published what eventually became a famous essay, entitled “Economic Possibilities For Our Grandchildren,” in which he tried to predict what “the progressive countries” (what would now be called the developed world) would look like in 2030.

The essay makes two big predictions:

(1) By 2030 the developed world would be in per capita terms four to eight times wealthier than it was a century earlier.

(2) This explosion of wealth would produce a tremendous reduction of hours worked, as people chose leisure over yet more income.

The first prediction was almost uncannily accurate, while the second has turned out to be completely wrong in regard to the United States, and largely wrong about Europe. (I’m not familiar enough with the relevant statistics in regard to the emerging economies of Asia and Latin America to comment on the salience of Keynes’ second prediction to them).

Regarding the first prediction, U.S. per capita GDP was about 6.2 times higher at the end of 2013 than it was at the end of 1929 (it increased from just over $8000 to just under $50,000 in chained 2009 dollars). Extrapolating out, Keynes high end prediction appears to have been if anything slightly conservative. Keynes thought that such a trajectory would result in something like a 15-hour work week for most people who worked for income:

For many ages to come the old Adam will be so strong in us that everybody will need to do some work if he is to be contented. We shall do more things for ourselves than is usual with the rich to-day, only too glad to have small duties and tasks and routines. But beyond this, we shall endeavour to spread the bread thin on the butter-to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!

At the time, this seemed like a reasonable projection, as work hours in the US and Europe had declined considerably over the previous half century, and Keynes assumed that the income effect — the declining marginal utility of income in relation to leisure — would cause this trend to continue. Since then, however, the decline in working hours has ceased almost completely in the US, and slowed down drastically in Europe (Europeans do work about 20% fewer hours than Americans however, which is not a trivial distinction).

Economics being a rather tautological discipline, there is of course a ready theoretical explanation for this as well: the substitution effect — i.e., to the extent that productivity increases are reflected in higher income per hour worked, each hour of forgone work in favor of leisure becomes more costly to the worker.

Of course this doesn’t explain why the substitution effect seems to have won out almost completely over the income effect in the US, and to a lesser extent in Europe. And here the biggest weakness of Keynes’ analysis in the essay — a surprising weakness given how relatively sensitive he was to sociological considerations — becomes obvious: there isn’t a word in the piece about distributional considerations.

Here are some numbers from the U.S. census bureau (all figures are in 2012 dollars):

Median household income in 1967: $42,323

Median household income in 2012: $51,017

Per capita GDP 1967: $21,893

Per capita GDP 2012: $49,231

In the late 1960s, median household income was nearly double per capita GDP, while today we have nearly a one to one relationship between the two metrics (Households are on average only slightly smaller today. I don’t have figures for 1967 handy, but in 1975 the average household included 2.89 people, while in 2012 it featured 2.54 persons). Or to put it another way, if over the past 45 years the nation’s increasing wealth as measured by output had ended up getting distributed equally across income groups as income, median household income in the US would be nearly $100,000 per year, rather than half that sum.

This helps explain, I think, why in the US in particular the substitution effect has been so much stronger than the income effect: it’s much harder to forgo additional income in favor of increased leisure when the relative wealth of those above you in the SES hierarchy is increasing faster than your own — especially in a culture obsessed with the conspicuous consumption of positional goods.