Atlas › The world in numbers

Case study 33 of 198

A random American's day, as Jev pictures it

Pick an American at random on a random day: how long did they sleep, work, watch TV, exercise? Does Jev's picture of that day match 181,000 time diaries?

result39 questions

Asked how a random American spent a random day, Jev orders the activities moderately like 181,335 real time diaries (rank correlation 0.66), but its day is a brochure, not a diary. It overstates relaxing (198 vs 18 minutes), education (88 vs 26) and leisure computer use (66 vs 10), understates work (120 vs 197) and sleep (476 vs 521), and pictures every day containing everything: it says almost no one spends zero time on the computer (7%), where 89% of real days have none.

0200400600

sleeping

0% none

working

57% none

20% none

travel

14% none

eating

4% none

grooming

20% none

socializing

63% none

housework

64% none

cooking

46% none

education

92% none

childcare

79% none

shopping

60% none

reading

77% none

exercise

81% none

relaxing

80% none

games

90% none

computer

89% none

religion

91% none

volunteering

94% none

calls

86% none

Jevpeopleminutes per day (magenta Jev, ink diaries)

How to read this: One row per activity: minutes per day in the diaries (diamond) and in Jev's picture (square); the figure on the right is the share of diary days with none of that activity.

20 activities, 181,335 diary days; mean similarity of the distributions 0.53.

In short

Jev gets the broad shape of a day right. Sleep comes first and TV time is close to the real figure.
Real days are lopsided and Jev's are balanced. It pictures a bit of everything every day, where most real days have none of most activities.
In its picture, doing nothing in particular takes more of the day than work.

What the data shows

Long day factory meme: Jev's American after a long day of relaxing

Sleep comes first (476 minutes for Jev, 521 in the diaries), and the order of activities matches moderately (0.66).
Too much leisure and learning: relaxing 198 minutes vs 18, education 88 vs 26, leisure computer use 66 vs 10, calls and email 37 vs 6.5, eating 110 vs 66.
Too little work: 120 minutes vs 197. Jev does say most days have no work (65%), close to the diaries (57%), but gives the workers shorter days.
Most days have a bit of most things: Jev says almost every day includes some computer use (7% none vs 89%), some calls (4% vs 86%) and some relaxing (0% vs 80%).

What it means, and what it doesn't

Jev pictures an average day as a balanced one: a bit of everything, plenty of downtime, some learning. Real days are lopsided: a few long blocks and most activities at zero. That's the gap between how people describe a day and how they live one, and Jev sits on the description side.

It doesn't mean Jev has everything upside down: sleep comes first and TV (181 minutes, against 163 in the diaries) is close. But in its picture, doing nothing in particular takes more of the day than work, and that says a lot about where the picture comes from.

Caveats

Narrow diary categories. The diaries code each stretch of time as one main activity. "Relaxing and thinking" and "phone calls, mail and email" count only time coded as exactly that, so they're small in the diaries. Jev likely read them more broadly, which inflates the gap for those two.
The diaries are old. They cover 2003 to 2016. Screen time has grown since, so Jev's higher computer figure is partly the world changing, not only Jev being wrong.
Coarse answers. Jev answered in nine time bins (none, 1-29 minutes, ..., 10 hours or more); minutes come from the middle of each bin, so they're rough. The share of days at zero is exact.
The question is about one day. "A random American on a random day" includes weekends, holidays and people who don't work. That's the diaries' whole point, and where a picture of a typical weekday goes wrong.

Jev on this experiment

Would a person find it interesting to read?: Yes80%
Does it describe you?: No70%
Would you have predicted it?: No60%
How fair is the comparison?: The comparison is reasonable
How much should a reader rely on it?: Moderately
Which caveat matters most?: Narrow diary categories65%

Why ask this

Time-use diaries are the least flattering mirror of daily life. On a random day, most Americans don't work, most don't exercise, and TV takes more time than anything but sleep and work. Ask people what their day looks like and you get something closer to the brochure.

A model asked to picture "a day" has read a lot of brochures. Does it know the diary?

How this was done

The people and the data

The American Time Use Survey, run by the Bureau of Labor Statistics: Americans aged 15 and older record everything they did on one day, minute by minute. The comparison uses 181,335 diary days from 2003 to 2016 (public domain), weighted to represent the population, and for each of 20 activities the share of days falling in each time bin.

What Jev was asked

One question per activity, framed exactly as the diaries measure it:

Pick an American aged 15 or older at random, on a random day of the year. How much time did they spend relaxing and thinking, doing nothing in particular that day?
None at all · 1 to 29 minutes · 30 to 59 minutes · 1 to 2 hours · 2 to 3 hours · 3 to 5 hours · 5 to 8 hours · 8 to 10 hours · 10 hours or more

(In the diaries, 80% of days have none.) There was one such question for each of the 20 activities, each asked with the bins in shuffled orders and averaged.

How it was measured

For each activity, the share of days Jev says have none of it against the diaries', and the average minutes (from the middle of each bin) against the diaries' weighted average. The analysis also checks how well Jev orders the activities by time spent (rank correlation: 1 same order, 0 no relation).

Where these questions live

39 questions across 2 topics of the map. Each opens on the map with every question in it.

Society & Culture

How Americans Spend Their Day20

Lifestyle & Taste

Habits & Routines19

Every question

All 39 questions behind this result, the telling ones first: the examples the analysis points to, then the ones where Jev misses, biggest gap first.

Jev’s own answer

Showing 0 of 39