Why Do We Work So Much?

by ed dolan
Klaus Ohlenschlaeger/Alamy Stock Photo

ed dolan is a senior fellow at the Niskanen Center, a nonpartisan think tank in Washington.

Published December 14, 2018


The Wall Street Journal recently carried a story on the FIRE movement. The acronym stands for Financial Independence, Retire Early — a trend (or at least an aspiration) among young professionals who want to take control of their lives by living frugally, saving up to three-quarters of what they earn, and retiring at 40. The story spurred me to revisit an old question: why, despite dramatic increases in productivity, do people continue to work so much and enjoy so little leisure?

The question is posed in a 2013 book by Robert and Edward Skidelsky, How Much is Enough?. The authors begin with a quotation from a 1928 lecture by the great John Maynard Keynes, “The Economic Possibilities for our Grandchildren.” Guesstimating long-term economic growth, Keynes argued that those living a century in the future would enjoy a standard of living four to eight times higher than his contemporaries. 

In our own lifetimes … we may be able to perform all the operations of agriculture, mining and manufacture with a quarter of the human effort to which we have been accustomed. … Thus for the first time since his creation man will be faced with his real, his permanent problem — how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest will have won for him, to live wisely and agreeably and well. … Three-hour shifts or a fifteen-hour week … is quite enough to satisfy the old Adam in most of us!

As it turns out, we are actually ahead of Keynes’s schedule. Yet most people in developed economies have only cut their work week by about 20 percent since the 1930s. The Skidelskys offer some explanations for the persistence of long working hours, but I don’t find them especially persuasive. Here are my own.

Keynes, the Material Guy

Keynes’s prediction focused largely on people’s supply and demand for material goods— “the operations of agriculture, mining and manufacture.” The surprising fact is that we’ve already gotten where Keynes anticipated, at least by one interpretation of what he meant. 

In the United States, mining, logging, manufacturing and construction account for only about 14 percent of all non-farm jobs; adding farming brings the total to 16 percent. Workers in these goods-producing industries work an average of 40 hours per week, a little more than the 33-hour average for the labor force as a whole. So the production of goods accounts for about 19 percent of all hours worked. Thus, working Americans spend an average of just 6.3 hours a week producing material goods of all kinds.

But, you may say, isn’t that just because so many of the goods Americans consume are produced in China and elsewhere? Fair enough. Let’s try again.

Instead of starting with employment, look at GDP. Consumer goods, including durables (e.g., SUVs) and non-durables (e.g., toilet paper), farm and non-farm — but excluding services — account for 25 percent of U.S. GDP. That includes consumption of imported goods, which exceeds the value of exported goods by about 6 percent of GDP. If we suppose that Americans were to produce all goods consumed in the United States at an average level of productivity (and to stop exporting them) it would still only take 10 hours of our average 33-hour week — well below Keynes’s prediction of 15 hours.

These numbers cast the puzzle of leisure in a different light. The question is not why we spend so many hours a week producing lawn furniture and golf clubs that aren’t critical to survival, but instead, what is important enough to occupy the other three-quarters of our working hours, when leisure is an option? 

Working for the IRS

A big part of the answer is that many of those other hours pay for the services of government. We expect far more from government now than in Keynes’s day. Government employees account for about 15 percent of all workers (including 1.3 million full-time military and about 800,000 local police). So that adds up to about 5 hours out of the average workweek for the labor force as a whole.

To be sure, just counting the hours put in by government employees understates the amount of time we work to pay for government, since many government services are performed by private contractors. A better measure might be government consumption expenditures and gross investment, a line in the national accounts that equals 17 percent of GDP. That item includes the salaries of civil servants, those of government contractors, and all of the goods purchased by government at federal, state and local levels. If we convert that to a share of the average workweek, it brings the total to 5.6 hours.;

An alternative measure of the amount of labor absorbed by government would be the number of hours we have to work to pay taxes. Tax revenues at all levels of government equal about 26 percent of GDP in the United States, which translates to about 8.6 hours of work per week. Note that this figure is considerably higher than the aforementioned 5.6 hours because it pays for trillions of dollars in cash transfers to the dependent population — everyone from the unemployed to the retired to the disabled.

We could, of course, free up more time for leisure if we accepted a smaller government. Instead, following a tendency sometimes known as Wagner's Law, governments in most countries have tended to grow faster than the rest of the economy. There is abundant evidence that as long as the quality of government remains high, larger government is not in itself detrimental either to freedom or to prosperity.

The question is not why we spend so many hours a week producing lawn furniture and golf clubs that aren’t critical to survival, but instead, what is important enough to occupy the other three-quarters of our working hours, when leisure is an option?
What Do We Do With Extra Days?

We have now accounted for about 19 hours of our average 33-hour workweek — 10 to produce goods and 9 to pay for government. Why do we work the other 14 hours rather than enjoying more leisure?

Clearly, we work some of those hours to pay for services that we could reasonably consider as necessities, but are not provided by government. In the United States, the biggest item in that category is the portion of health care services not covered by government. That comes to about 9 percent of GDP, or a little under 3 hours per week. Many families also view higher education and preschool education as expensive necessities. Suppose, for the sake of discussion, we assume that paying for privately supplied service necessities requires an average of six hours per week. Adding 6 hours to the 19 still leaves 8 hours — one full working day that could be devoted to leisure. Why don’t Americans use the time in this way?

A good starting point is to think more closely about the meaning of leisure. The Skidelskys have a rather narrow conception, dismissing mere rest and relaxation. They are positively disdainful of, say, watching television or getting tipsy at the neighborhood pub. They put creative activities at the top of their list — sculpting, playing in an orchestra, painting, writing (whether poetry or popular books on the economics of leisure). Apparently not wanting to sound too highbrow, they put their imprimatur on playing team sports, making one’s own furniture and strumming a guitar with friends.

Note that few of these activities are purely solitary. Painters might work alone in their studios, but some of their pleasure in painting comes from other peoples’ enjoyment of what they put on canvas. The singer wants someone to listen, the writer wants a reader, and the cook wants someone to share the coq au vin. Although you can walk on the beach by yourself, playing bridge with your friends is more fun than playing against the computer.

Looking at these leisure activities from an economist’s perspective, what we see is an exchange of leisure services in which John views Maria’s paintings, Maria enjoys Howard’s cooking, Howard cheers for John’s basketball team, and so on. I agree that those seem hours well spent, and that people who live to work are missing some of the best things in life. But, as enjoyable as they can be, there are ways to enjoy them even more that require more labor hours in addition to the time directly spent in leisure. For example, John and Maria might appreciate painting even more if they viewed the works of professionals in museums, for which they would need to work enough to buy tickets, or perhaps travel to a distant city. And maybe Maria and Howard would like to share their love of food by dining out.

These activities are still leisure, and still involve an exchange of services, but now the exchange is mediated by, or perhaps we could say enhanced by, the market. Is that in any way bad? If we had to depend entirely on art and food produced by amateurs, would we really be better off? I can’t see that we would.

Keynes Revisited

How, then, to resolve the puzzle posed by Keynes’s “mistaken” prediction of a 15-hour work week? Keynes was right in thinking that his grandchildren would be able to satisfy their demands for the material goods with a 15-hour work week. But we still need government services that require more work to produce, and there are still plenty of services, like higher education, that qualify as work-worthy purchases.

We still appear to work about a day longer a week than needed to provide all of the above. We spend much of that time earning extra money to spend on market-mediated exchange of leisure services, like watching Nascar and attending rock-and-roll concerts. 

I have framed the calculations in this post in terms of the length of the work week. If we want more leisure, we work shorter days, or take off Fridays to play basketball or fool around in the kitchen. But, of course, we have the option of taking leisure in more concentrated doses. 

The United States is, famously, the “No-Vacation Nation.” (Though it sure doesn’t look like it on a summer afternoon in Yosemite or San Francisco.) The average paid vacation is just 10 days. Contrast that with Danes, who get 25 paid days off a year, New Zealanders, who get 30, or the famously industrious Germans, who get 35. We could also demand more paid sick leave, more parental leave and the like. Still another option would be to beaver-as-usual, live frugally, and retire early — as opposed to never retiring, which an increasing number of Americans seem to be doing to the applause of economists worried about the rising rate of dependency.

Mr. Keynes, Meet Mr. Veblen

Any of these patterns — shorter workweeks, longer vacations, earlier retirement — could mean a step toward “living wisely and agreeably and well,” to use Keynes’s phrase. We should celebrate the fact that rising productivity has made this possible for many. Still, many others — here I am talking only about people who have left the threat of absolute poverty far behind — have chosen to keep right on working in order to sleep in larger houses at night and drive fancier cars to work in the morning.

Perhaps the determination with which they make the choice would surprise Keynes most of all. But that is another story…

main topic: Workforce