P.S. on the Leisure of Keynes’s “Grandchildren”

A recent post of mine on “the intriguing twentieth century” focused on John Maynard Keynes’s prediction of an eightfold multiplication of GDP per capita in the century following 1930. Part of the comments and conversation dealt with a related prediction of Keynes—that leisure would increase up to the point where people would work only 15 hours a week. More should be said on why this last forecast failed.

Mark Brady of San Jose State University brought our attention to the book of Lorenzo Pecchi and Gustavo Piga (Eds.), Revisiting Keynes “Economic Possibilities for our Grandchildren” (MIT Press, 2008). It contains an interesting chapter by Harvard University’s Richard Freeman, titled “Why Do We Work More Than Keynes Expected?” (pp. 135-142). The crucial technical passage regarding leisure is the following:

[Keynes] missed the boat by failing to appreciate the power of economic incentives to induce people, even those with high standards of living, to work long and hard. He did not expect that the increased cost of leisure due to rising wages would dominate the income effect that induces people to take more leisure. … Textbooks often displayed backward-bending labor supply curves to illustrate the point. But the race between the substitution and income effects turned out to be more of a fair race than the sure-fire guaranteed winner that your local tout predicted.

An especially interesting chapter is that of Gary Becker (the famous Nobel economist) and Luis Rayo, “Why Keynes Underestimated Consumption and Overestimated Leisure for the Long Run” (pp. 179-184). Their analysis also makes much use of the distinction between the income and the substitution effects, combined with a very Beckerian human-capital analysis:

Keynes assumed that higher incomes would lead to increased demand for leisure through what is now called the “income” effect. But in the same year as Keynes published this article, Lionel Robbins published a classic article showing that higher hourly earnings have conflicting effects on hours worked. …

Keynes was misled in his predictions concerning the effect of higher income on hours worked by the behavior of gentlemen in Britain—who Keynes believed provided a window onto future behavior as everyone’s income rose. Their behavior gave a distorted picture of what to expect because these gentlemen had sizable wealth in the form of physical and financial assets, but not high human capital or earnings. So economic theory would predict that these gentlemen would take more leisure than would equally wealthy persons in the future who in fact would be holding the vast majority of their assets in human capital rather than land and other assets. English gentlemen indeed had mainly just an income effect, while those who would have to work would also have powerful substitution effects.

The reference to Robbins is interesting in itself because it betrays the kinship between the Austrian-influenced economist of the London School of Economics, and Becker, the quintessential neoclassical economist at Chicago. In my Regulation review of Robbins’s famous 1932 book An Essay on the Nature and Significance of Economic Science, I noted:

The Essay defined economics as the science that studies human behavior in allocating scarce means among competing ends, a definition that has become standard. We get a glimpse at this definition’s influence when we realize that it was adopted by Gary Becker, the standard bearer of mathematical and empirical economics, which is the polar opposite of the Austrian school.

“Polar opposite” is too strong (I should have written “appears to be”) as Robbins and Becker themselves demonstrated. Robbins’s book and my review survey these important ideas.