I want to thank Kevin Drum from Mother Jones for surfacing a 2011 Bureau of Labor Statistics study that confirms something I’ve long suspected: Virtually anyone you know who claims to be working more than 60 hours a week is not telling the truth.
Bureau of Labor Statistics researchers reached this conclusion by comparing regular survey data to diary data from the American Time Use Survey, a Census project that asks Americans to track, diary style, how their weekly time is divided among 163 different activity categories, from sleeping to shopping to pet care.
The ATUS data has a couple of advantages over simpler survey data where you just ask people how many hours a week they typically spend working for pay, or taking care of the house, or attending church. One is that people are bad at estimating: When you give people a long list of activity categories and ask them how many hours a week they spend on each, they tend to give an answer that sums to more than 168 hours. The ATUS diary structure has the advantage of limiting respondents to 24 hours of time when reporting a day’s activity. Another issue is that questions about one specific activity are more subject to social desirability bias: If you ask people to tell you everything they did on Sunday, rather than asking specifically how much time they spent at church, they may be less likely to fib in a way that projects virtue.
The BLS study found respondents in the ATUS tend to give an estimate of typical working time that is 5 to 10 percent higher than what shows up in their diaries. But the divergence was not uniform across the population. The largest overestimates came from the people providing the highest estimates: People who said they typically worked 75 or more hours per week tended to provide diaries reflecting 25 hours’ less work per week than they estimated. People claiming to typically work between 65 and 74 hours weekly tended to be overestimating by 18 hours.
Again, this sort of misreporting is not limited to work hours. People overestimate how often they do all sorts of things they “ought” to be doing, often by even larger margins than with work for pay. One study from the 1980s found swim and tennis club members provided estimates of their frequency of use of recreation facilities that was double what was reflected in their clubs’ logbooks.
As George Costanza says, it’s not a lie if you believe it. Do the people who give inaccurate survey answers about their time use know they’re not telling the truth? Like Drum, I lean toward the view that the deception here is mostly self-deception, especially in the case of gyms and other recreational facilities. Their business models often rely on members overestimating how often they will attend, so they will be willing to pay monthly fees that in some cases exceed what it would cost to pay per-visit.
So you can think of your friend who claims to work 75 hours a week at the bank similarly to your friend who claims to go to the gym six days a week: He’s probably lying to you, and, possibly, to himself.