Skip to main content
It’s Hard to Get Off the Couch When You’re Unemployed

Mark, 22 and unemployed, sleeps late in the morning.

His roommate has to get up for work, but Mark has nowhere to be. He rolls out of bed at 11 a.m. He checks his email — still no response to his last round of resumes — and heads out for a run. When he gets home, he spends 45 minutes filling out job applications, then plops down in front of the television for a couple hours before cleaning up the house — he’s taken on more chores since his roommate is cutting him a break on the rent. In the evening, his buddies are catching a game at the local bar, but Mark has class at the local community college, where he’s working toward a certificate in HVAC repair.

More Economics

Mark isn’t a real person. Or rather, he is an amalgamation of thousands of people struggling to get by in an economy still stuck in second gear.

That deep divide between those with jobs and those without them reveals itself not just in well-known statistics on hiring and income but in the day-to-day details of how people live their lives. The unemployed have higher rates of depression, obesity and suicide. In interviews, they frequently report that the social and emotional impacts of joblessness — isolation from friends, the loss of a daily routine, feelings of uselessness — can be as hard as the financial toll. Many say it’s hard just to get out of bed in the morning.

Government data released Wednesday helps put numbers to those anecdotes. Every year, researchers from the Census Bureau1 ask thousands of Americans for a minute-by-minute accounting of how they spend their days.2 The result, the American Time Use Survey, provides a remarkably detailed look at how much time Americans spend doing everything from grocery shopping (6.4 minutes a day, on average) to helping their children with their homework (7.5 minutes for the average parent) to lying awake in bed (more than an hour and 20 minutes for the average insomniac).3 It isn’t much good as an economic indicator — it’s released just once a year, and the year-to-year changes are small and fairly volatile — but it paints a detailed picture of how the economy affects people’s lives.

Overall, the unemployed spent more time sleeping, watching television and taking classes than the employed, and less time eating out and going to parties. They spent about an hour and a half a day, on average, on activities related to finding a job, including career-related education. But the most interesting details come when you zoom in on specific demographic groups.



Take Mark. To construct this character, I combed through the survey to identify unemployed, unmarried men ages 18 to 29 without a college degree, a group that had an unemployment rate of 14.3 percent at the end of last year.4 There were 50 of them in the 2013 survey, a small number, but enough to get a sense of how people like Mark live their lives. There were another 210 people in the survey who were like Mark in age, sex, education and marital status, but exhibited one key difference: They had jobs.

Some of the differences in how the two groups spent their time are obvious. The employed worked 5 hours a day on average (the averages include weekends, so that’s the equivalent of a 35-hour workweek) and spent another 20 minutes a day commuting. The unemployed spent essentially no time working or commuting, but spent about 45 minutes a day searching for jobs.5

Perhaps more revealing is how unemployed people in Mark’s demographic filled the rest of that extra time. They spent about five hours a day watching television and playing games, two and a half hours more than their employed counterparts. They spent about 45 minutes a day working out, versus about half an hour for the employed, and they got 40 more minutes of sleep, for a total of nearly 10 hours a night. They also spent half an hour more per day in class working toward a degree or certificate.6 But despite the extra free time, they reported spending less time relaxing, watching and playing sports, and eating and drinking. (An important conceptual note: These figures are based on the average amount of time spent on each activity by all members of the group, regardless of whether they participated in the activity at all. So if one person spends an hour in class and another doesn’t go to class at all, their average time spent in class will be half an hour. Where relevant, I’ve included more detailed breakdowns in the footnotes.)

Unemployment doesn’t affect all groups the same way. Young, single men like Mark are one big category of the unemployed. Another big group is single mothers, who had an unemployment rate of 10.4 percent in late 2013.7 They, too, got significantly more sleep than their employed counterparts, and watched more television. They also spent more time with their children and, unlike unemployed young men, spent more time relaxing.

Other activities show the isolation of unemployment: Unemployed single moms talked to friends and family on the phone less often than those with jobs, and they spent 15 fewer minutes per day socializing and going to parties. They also spent less time working out, more time eating and 12 minutes less per day on personal grooming.

And unsurprisingly, joblessness takes a financial toll: Unemployed single moms spent less time doing virtually everything that costs money, such as shopping and going out to eat,8 and more time trying to scrape together a living either from government services or on their own via babysitting or selling things on eBay.9

Education and marriage offer at least some protection. People are less likely to become unemployed if they’re married or have a bachelor’s degree, and particularly if they have both10; the unemployment rate for married college graduates was 2.7 percent in late 2013. And when they do lose their jobs, they are at least partially buffered from the effects of joblessness. The difference in time use between the employed and unemployed was much smaller for married college graduates than for other groups. They slept essentially the same amount, perhaps because many had an employed spouse to help keep them on a regular schedule, and used some of their extra free time volunteering and socializing, suggesting they were less isolated.

Perhaps surprisingly, the time-use data doesn’t show much evidence that the social and emotional effects of unemployment worsen over time. In fact, there was very little difference in time use between the long- and short-term unemployed. That’s consistent with other evidence that the two groups are very similar demographically, and with research that shows relatively little evidence that physical or mental health deteriorates as unemployment drags on.11 The time-use data suggests that the long-term unemployed haven’t given up looking for a job, despite concerns among many economists that they have drifted permanently from the workforce. People out of work for more than six months actually spent somewhat more time on average looking for a job (about 35 minutes a day compared to 20 for the short-term unemployed). But they were less than half as likely to get an interview.12

There are other, mostly small differences between different demographic groups. But across all groups, the more basic truth is the same: The 5-year-old economic recovery has done little to close the gap between those with jobs and those without them. In fact, the gap may be widening. By various measures, the economy is returning to normal for the employed, even as the unemployed — particularly the long-term unemployed — have seen little improvement in their job prospects. The employed, on average, are working more hours than in the recession, have more job security and might at last be in line for a raise. Mark, meanwhile, is still home on the couch.


  1. The survey is conducted by the Census Bureau on behalf of the Bureau of Labor Statistics, which releases the data.
  2. In 2013, the American Time Use Survey included about 11,400 individuals. Respondents were asked to recall how they spent their time on a single day (they were warned ahead of time, and called the following day), and those responses were weighted to create a representative sample of the entire country.
  3. On average, Americans reported being “sleepless” for just 3.5 minutes a night. But that’s because the vast majority of respondents didn’t report being sleepless for any time at all. Among those who did report sleeplessness, the average was 83.3 minutes. Of course, you’re more likely to remember your period of insomnia if it lasted for a while.
  4. This unemployment rate, and other detailed labor force statistics in this article, are calculated from Current Population Survey microdata from the fourth quarter of 2013. Because of differences in the surveys, the precise definitions of each group may differ slightly between them.
  5. The relatively small sample sizes, particularly for the unemployed, mean that only large differences in time use are statistically significant by conventional measures. In this story, I focus on the biggest disparities, but I haven’t run a formal significance test for every number. So interpret all figures, especially small differences, with caution.

    Perhaps more importantly, the size of the survey makes it impossible to control for more than a handful of variables at once. That means that while Mark and his theoretical employed roommate are similar in age, sex, education and marital status, they may well be different in other ways, such as race and employment history. And of course, no broad survey can do a good job of capturing factors like motivation or intelligence. As a result, there may be important differences between the two groups that are correlated with — but not caused by — unemployment.

  6. The unemployed were more than twice as likely to report any time spent in class (12 percent versus 5 percent) and spent more time there (close to six hours versus three and a half for the employed).
  7. I’ll define this group as unmarried women ages 20 and up with a child under 18 at home.
  8. The time-use survey doesn’t distinguish between eating in and eating out, but it does collect information on “travel related to eating and drinking” (a separate category from travel for grocery shopping), which should be a reasonable proxy for frequency of dining out.
  9. This last category highlights a challenge in economic data. The average single mother spent seven minutes a day on “income-generating services” such as babysitting, lawn mowing and the like. That kind of work qualifies as “employment” under the official definition, meaning these women likely shouldn’t, strictly speaking, have been counted as unemployed.
  10. This is true even controlling for age.
  11. To be clear, there’s lots of evidence that unemployment itself is bad for both physical and mental health. It just isn’t clear that longer spells of unemployment make these problems worse.
  12. Distinguishing between the short-term and long-term unemployed is somewhat problematic in the time-use survey. The survey is an offshoot of the Current Population Survey; a subset of CPS respondents are re-interviewed for the time-use survey two to five months after their final CPS interview. Basic information, such as age, sex and employment status, are collected in both surveys. But duration of unemployment is collected only in the CPS. As a result, the unemployment duration data is several months out of date when the time-use data is collected. For this story, I defined the long-term unemployed as anyone who was out of work at least six months at the time of the CPS survey and was unemployed at the time of the time-use survey (meaning it’s possible some of these people found a temporary job in between). The short-term unemployed are those who were unemployed in the time-use survey and either were employed or were out of work less than 15 weeks at the time of the CPS.

Ben Casselman is a senior editor and the chief economics writer for FiveThirtyEight.

Filed under Unemployment 67 posts, Our Changing Economy 30, Employment 14, Long-Term Unemployed 8, American Time Use Survey 7, Time Use 3

Comments Add Comment