by John Q on October 18, 2007

Following up my post on consumption and living standards in the US, there was a fair bit of discussion of what’s been happening to leisure. Juliet Schor and others have argued that the long-term trend towards reduced hours of work and more leisure reversed some time in the 1970s, and people have been working harder since then. A study by Aguair and Hurst (the final QJE article is subscription-only, but I found a preliminary version here) has been widely quoted as proving the opposite (here, for example, by Tyler Cowen) and the abstract seems to support this interpretation, saying “We find that a dramatic increase in leisure time lies behind the relatively stable number of market hours worked between 1965 and 2003.”

However the data periods don’t exactly match up. It turns out that, using any of the definitions of leisure considered by Aguair and Hurst, the majority of the increase in leisure time took place between 1965 and 1975, and most measures show little change since 1985.

There’s an important gender/family dimension too. On Aguair and Hurst measures 1 and 2 (which exclude child care), leisure time for women peaked in 1985 or 1993 and has declined since then, while leisure time for men has been increased marginally since 1985.

So that readers can make their own comparisons, I’ve extracted the relevant table, which is over the fold. I’d say it matches Schor’s story (increasing leisure until the late 70s followed by a decline) at least as well as that suggested by the authors (“dramatic” long-term increases in leisure)

There’s lots more data in the Aguiar and Hirst paper and one point worth noting is that the trend in the distribution of leisure time is the opposite of that in income. High income, high education people have experienced a significant decline in leisure relative to those with low income and low education. That somewhat offsets the growth in income inequality over the same period. Also, combined with the gender pattern I already mentioned, it almost certainly means that educated women have, on average, less leisure than in the late 1970s.




Michael B Sullivan 10.18.07 at 11:26 pm

It looks to me like those data suggest that leisure time for women in measures 1-3 peaked in 1993, not 1985.
Youre right. I was just fixing this as you posted – JQ


derrida derider 10.18.07 at 11:32 pm

The trouble with this framework is that it doesn’t distinguish between leisure and non-market production.

Americans – especially women – spend less time cooking, cleaning and looking after their kids than they did in 1970. These functions have been largely outsourced to the market. So you’d expect the time saved to be divided in some proportion between leisure and increased market work hours.


SG 10.18.07 at 11:57 pm

I don’t know if this is the same elsewhere, but you can read a lot about the conservative/free market view of the peons’ leisure time from their response to the “problem” of the ageing population. In Australia, rather than find a government-backed method to encourage women to work (thus pushing up the participation rate), the government asked working men to retire later.

i.e. we don’t have enough workers anymore, and we don’t want to even consider some kind of government-subsidized childcare to solve the problem, so would you lowly peasants mind putting off your leisure time until you die? After all, you can sleep as much as you want then.


Robert 10.19.07 at 12:31 am

Not exactly comparable, but here are some cross-national comparisons of average hours worked.


vivian 10.19.07 at 12:55 am

#2: Americans – especially women – spend less time cooking, cleaning and looking after their kids than they did in 1970. These functions have been largely outsourced to the market.

Um, didn’t you ever hear of the double day? (that’s just the top correct google hit, not the full bibliography, because I’m busy with my kids now that I’m home from work.) Yes, women work outside the home more frequently now, so they get others to provide childcare, but quite often it isn’t ‘the market’ so much as family and friends. But despite what you saw on the Jetsons, housework and arranging the patchwork of activities, doctor’s visits, etc. takes time.

C’mon Belle, you can do this much more entertainingly than I can. Where are you?


harry b 10.19.07 at 1:19 am

Don’t denigrate yourself, vivian.
And, there is a huge difference between spending the whole day looking after the house and kids, and spending the whole day working outside the home and then fitting all the household work and childcaring/relating into the time spent at home before and after work (Hochschild’s The Time Bind is very good on this).


Tracy W 10.19.07 at 9:34 am

Out of curiousity, where do gap years and OEs (overseas experiences) fit into leisure time?

As a NZer, I know it’s fairly common to take a leave of absence from your job, or quit it, and go travelling around the world for a year or two. This seems to be a result of increased incomes allowing this to be financed. Taking a year’s holiday all at once, rather than taking an extra 1.3 weeks each year over a 40 year working-life has some advantages. And as a third-generation member of Servas I have encountered a number of Americans doing an OE. But of course, any American off spending a year mooching around Outer Mongolia is not going to be answering the door to any statisticians. Is taking your leisure time in large lumps accounted for in these surveys of time-leisure trade-offs?


Dylan Thurston 10.19.07 at 10:38 am

#2, the study does talk extensively about non-market work, in several categories, and mentions, e.g., that time spent on total non-market work has fallen by 3.8 hours/wk. It’s up 3.75 hours/wk for men, down 10.3 hours/wk for women, but still with a substantial gender difference of 9.1 hours/wk. Most non-market work was excluded from all measures of leisure time, except for pet care and gardening. You might want to question some of the choices, but they certainly consider the issue.


Neil B. 10.19.07 at 2:03 pm

Is it possible that those who think “we” have more leisure time (versus the clear evidence of the opposite) have made a crude statistical fallacy? For example, what if we just took those who worked and found “average working hours.” That would be summing the hours of all working persons (in the past, mostly men) and dividing by n_e, the number of working persons – not n the total adult population n. Suppose for simplicity and illustration, all the men worked 40 h/week, no women work, equal numbers of each exist. Then, the “average” work hours comes to 40 h/week. (But the real average hours worked for adults is 20h/week.)

But suppose some woman start working, say around 20 h/week. Clearly, more hours are worked by adults and their combined average free time is reduced. If half of them now work, the true average is now 25 instead of 20. But if we sum hours and divide by the new value of n_e (now increased because of the working women), the supposed “average hours worked” goes down to 33.33, due to dilution by the PT workers. See the difference? Whether or not someone who should know better made that specific error in arguing a point, I bet it has gummed up some statistics and been abused somewhere.



Kenny Easwaran 10.19.07 at 5:22 pm

It looks like on measures 3 and 4, the amount of leisure time has actually gone monotonically upwards. However, 3 includes childcare, and 4 includes everything that isn’t market work. So perhaps the change since the early ’90s has been in terms of increased time spent on childcare?


leederick 10.19.07 at 6:52 pm

A lot of interpretations of this study do not seem sustainable when you bear in mind the major proviso that the study excludes people under 21 and over 65, as well as those who are students or retirees.

I’m also not sure the male-female comparisons people seem intent on making are sound, given that the study demographically weights the sample across time, but not across gender.


sara 10.20.07 at 2:37 am

It looks as if the decrease in leisure parallels the baby boomers taking up family life; this may merely be my impression instead of statistically valid, but will we see a massive increase in leisure time around 2015 when the boomers retire, or not? A trend is that many will not retire, unless this is manufactured to make the rest of us keep working.


robert inkol 10.20.07 at 1:44 pm

Technology and lifestyle changes make this subject a difficult one to analyze. For example, many people spend considerable time at their home computers, but a proportion of this time is work related – indeed computers make it possible, or even necessary, to bring your work home.

One recollection that I find striking is that when we were looking for a house in the early 1980s, we found that a substantial proportion had finished basements with wet bars. From what I’ve seen, this isn’t very common any more (I know of friends who ripped theirs out).


Kaleberg 10.21.07 at 7:46 pm

I wrote up a little piece on this paper for Daily Kos a while back: http://www.dailykos.com/storyonly/2007/9/6/12411/05257

Basically, work time has been declining since 1900 or so. Less educated workers, who are typically paid by the hour, are generally working fewer hours, and this trend has been continuing. More educated workers, who are usually salaried, are working longer hours. Women are working longer hours than they used to. People stopped spending time with their friends in the late 1960s, and all the new leisure time has gone to watching television watching. I really wish I were making this up, but that’s what the study shows, and it looks like a pretty reasonable estimate.

It’s a rather grim picture. Of course, most people reading this blog are probably exempt, that is, exempt from wage-hour laws, so they are probably working longer hours and screaming, “Where’s my #^$%%^$ leisure time?”


Elizabeth 10.23.07 at 7:22 pm

I blogged about this paper a while back at:

I’m pretty sure the key issue is whether you count time spent doing child-focused activities like birthday parties or soccer games as “leisure.” They can be fun, sure, but they also feel obligatory (at least to middle-class parents).

Comments on this entry are closed.