How Much Do We Sleep? Americans Don't Know

But we spend billions of dollars on insomnia drugs, anyway.

Wikimedia Commons

Six hours and 48 minutes. That's how much the typical American sleeps, according to the most recent Gallup Survey.

Eight hours and 42 minutes. That's how much the typical American sleeps, according to the most recent U.S Department of Labor's American Time Use Survey.

How are American estimates of our own sleep times off by more than 13 hours per week?

Catherine Rampell wonders if Americans are providing two accurate answers to two separate questions: How much time do you spend in bed vs. how much time do you spend asleep? Perhaps. But even if it's old-fashioned self-reporting error, this isn't the only life category where Americans dramatically miscalculate their own lives. In work-hour surveys, she writes, people who claim to work 40 hours per week actually work an average of 37. People who claim a usual week of 75 hours really work 50.

And sleep time is a little trickier to accurately estimate than work hours on account of your being entirely unconscious. In Elizabeth Kolbert's New Yorker essay on the science of sleep, she visits a clinic to monitor her own rest and restlessness and comes away with a big surprise...

The first table showed that I’d spent six hours and forty-two minutes in bed, of which I had slept for four hours and two minutes. But I hadn’t lain awake for nearly three hours and then dropped off, as I’d imagined. A graph, known as a hypnogram, traced how I’d cycled from wakefulness through each of the sleep stages—1, 2, 3, and REM. A good sleeper’s hypnogram looks like sets of facing staircases. It traces a steady descent from wakefulness to REM sleep, then an ascent to a lighter sleep stage, then another steady descent, a pattern that gets repeated three or four times in a night.

My hypnogram looked like the Manhattan skyline. It turned out that I’d fallen asleep within about ten minutes of getting into bed. But, after only a minute or so, I’d woken up for about fifteen minutes. I fell asleep again and woke up, fell asleep again and woke up, fell asleep a fourth time and then remained awake for nearly an hour. Even when I’d felt that I’d finally conked out, I had kept waking up, for a mind-boggling total of a hundred and forty-one times. Most of these awakenings—a hundred and eleven—were brief, under fifteen seconds. The tables also showed that I’d stopped breathing eight times, which, Palat assured me, was not unusual, and had experienced seventeen “periodic limb movements,” also not uncommon. In his comments, Palat had written that my “sleep architecture” suggested “difficulty with sleep maintenance.” He advised me not to go to bed until I was sure I was tired; not to stay in bed when I couldn’t sleep but to go read in another room; and to eliminate alcohol.

To sum up, she was in bed for seven hours, fell asleep 141 times, woke up 141 times, and actually slept for four hours. If the Labor Department had asked her how many hours she slept, Kolbert would have said three. If Gallup had asked her hour many hours she spent in bed, she might had said seven. The actual sleeping time, however, was neither.

Americans aren't sleeping any more than we were in 1990, according to Gallup (even if the self-reported sleep survey is wrong, I'm going to assume it's equally wrong each year and therefore comparable over time). In that time, the number of U.S. insomnia diagnoses has gone from less than one million to more than five million. The medical dose has quintupled, but the sleep times haven't budged.

Research shows that insomnia drugs—for all their quirky risks and drowsy side-effects—add no more than 20 minutes in total sleep time, on average. That's 80 percent less than the gap in sleep-time between Gallup and the Labor Department. Maybe Americans need better data to go along with better drugs.