6 Comments
User's avatar
Liam Baldwin's avatar

Twice as much leisure is ridiculous, but 20% more is quite substantial!

Expand full comment
Jordan Peeples, PhD's avatar

Yeah, but not crazy, especially if it might be driven by age as well. Younger people make less money and don't have kids. Leisure time doesn't (shouldn't) include household chores devoted to children.

Expand full comment
Jordan Peeples, PhD's avatar

Aside from the horrible graph from Twitter, this is a classic example of a substitution effect at play -- the more money you make, the costlier it is to have leisure time. So, of course, we should see an effect where poorer people may work less due to that effect. Granted, it's not a crazy difference.

Expand full comment
Liam Baldwin's avatar

Exactly. Plus there are reasons to think the substitution effect dominates in aggregate for Americans as high incomes are driven largely by human capital. Gary Becker and Luis Rayo argued this in an essay trying to explain why working hours have not fallen as Keynes predicted. Consider a counterexample where high incomes are driven by physical capital:

“This difference is illustrated by the working habits of wealthy individuals in the various Gulf States, who typically get the vast majority of their income from oil revenues. It is said that in many of these countries, such as the Emirates, Qatar, or Kuwait, the typical working day for natives—as opposed to the imported laborers who do not share in oil revenues—is about three to four hours a day. This is actually very close to Keynes’s estimate of how many hours would be worked in advanced countries after another century of economic growth.”

In such a case the income effect dominates, and high earners will work less!

Expand full comment
Jordan Peeples, PhD's avatar

Good insight! I'm much more familiar with Becker's work in family economics, but yes, I'm sure income effect doesn't begin to dominate in the US until a certain threshold of income or a certain number of hours has been met. But that's probably a 95th percentile threshold. Probably a paper on it somewhere out there.

Expand full comment