It Was Never About Sleepiness
The clock says 12:47. You have an early morning, you know exactly what tomorrow looks like on four hours of sleep, and you are watching a stranger restore a barn in Norway. Your eyes sting. You keep watching. And here is the detail that makes the scene strange: nothing is stopping you from going to bed. The door is open, so to speak. You just do not walk through it.
Most sleep advice is built for a different problem โ the person lying in the dark, willing themselves to sleep. This is the opposite. Researchers call it bedtime procrastination: going to bed later than you intended, with no external reason for the delay. Not a medical term for a disorder. A description of a behavior most of us recognize from the inside.
The definition came with three parts when Floor Kroese and her colleagues at Utrecht University introduced it in 2014, and the parts matter. The delay has to be later than intended. It has to happen without external circumstances preventing the original bedtime. And it has to be something you would still call voluntary if a lawyer asked.
Fail any one of those tests and you are dealing with ordinary sleeplessness or a demanding job, not this. Pass all three, though, and you are in genuinely odd territory: a person sacrificing sleep to an activity they are not even enjoying much.
That last part deserves the emphasis. Plenty of bedtime procrastinators report the scrolling is mediocre โ mildly entertaining, occasionally interesting, rarely worth the morning. The behavior survives its own poor reviews. Any useful account of why has to explain that, and the discipline-and-willpower story does not.
- Tomorrow night, try naming the moment out loud: 'this is a delay, not insomnia.' The label changes which tools apply.
- If someone in your house stays up late for shift work or childcare, their situation is external โ this article's framework is not aimed at them.
One Survey of 177 People Started It
The paper that named the behavior was modest, and it is worth knowing exactly how modest. Kroese's group ran an online survey of 177 adults recruited through Mechanical Turk, measured a new bedtime procrastination scale against general procrastination and self-regulation measures, and asked people about their sleep.
What came back: people who scored higher on bedtime procrastination reported more insufficient sleep and more daytime fatigue, and the behavior tracked with broader self-regulation problems โ the same machinery behind unpaid parking tickets and unfinished tax forms, pointed at the act of ending the day.
Cross-sectional is the honest label for that kind of study. It captures one moment, so it cannot say whether the procrastination causes the bad sleep, the bad sleep erodes the self-control that guards bedtime, or some third factor drives both. Kroese's team said as much in the paper.
The contribution was not proof; it was the naming, plus the observation that this behavior predicted insufficient sleep above and beyond demographics and general self-regulation. A variable with its own identity.
The research since has kept the same shape. A 2023 survey of 560 Indian university students found just under ten percent with regular sleep habits and about a fifth scoring in the bedtime-procrastination range, with higher scores correlating modestly with daytime tiredness.
Another 2023 study of 495 undergraduates โ published by a Saudi-Polish team โ linked it to insomnia in the other direction too: students with medium or high bedtime procrastination scores were roughly twice as likely to report insomnia as their low-scoring classmates.
Notice what all of this is. Surveys, self-reports, correlations. Useful for sizing the phenomenon, weak for proving mechanisms. Anyone selling you a cure for bedtime procrastination is selling ahead of the evidence, and this article will keep flagging that line as it goes.
- Prevalence numbers vary wildly by population and cutoff, so treat any single statistic โ including the ones above โ as a snapshot, not a census.
The Word That Made It Spread: Revenge
Around 2020, a Chinese internet phrase โ bao fu xing ao ye,ๆฅๅคๆง็ฌๅค โ crossed into English as revenge bedtime procrastination. The revenge is the whole payload. The story it tells: your daytime belongs to other people. The job takes the mornings and afternoons, the commute and the chores take the evening, and midnight is the first stretch of hours that belongs to nobody but you. Staying up is not carelessness.
It is repossession.
That framing explains the behavior's signature oddness โ that people keep scrolling through content they describe as mediocre. The reward was never the barn restoration video. The reward is autonomy itself, taken in the only denomination available: minutes, at night, spent on nothing anyone can bill you for.
Sleep researchers had already noticed that people with less control over their daytime hours delay bedtime more, and the revenge framing gave that pattern a name vivid enough to travel.
An evidence note, because this term invites overclaiming. Revenge bedtime procrastination as a distinct construct is young; most of what exists is survey work, some of it in college populations, and the formal research on the revenge variant specifically remains thinner than the coverage implies. The underlying bedtime-procrastination literature is more solid, but even there the causal arrows are mostly untested.
What can be said honestly: the behavior is common, it correlates with insufficient sleep and insomnia reports, and the motives people give โ reclaiming personal time after a day that offered none โ are plausible and internally consistent, which is not the same as experimentally demonstrated.
There is also an older version of this insight, and it did not need a hashtag.
What the Old Texts Call It: A Shen That Will Not Come Home
The Huangdi Neijing, compiled roughly two thousand years ago, treats sleep as a matter of descent. Defensive qi โ wei qi, the waking, guarding, outward-facing portion of vital activity โ travels the surface of the body by day. At night it is supposed to sink inward, into the yin layers, and the sinking is the sleep.
The Ling Shu has a passage on insomnia built exactly on that model: when the wei qi cannot enter the yin and lingers in the yang, the yang grows full and restless, and the eye does not close.
Read the midnight scroller against that frame and the fit is uncomfortable. Brightness at the eye, attention pulled outward, a mind deliberately held at the surface โ the old text describes a physiology of not-descending, and the modern evening manufactures it on demand.
Where the classical account blames a failure of the qi to settle, the research blames a self-regulation system outmatched by content engineered to prevent settling. Different vocabularies, same shape.
The tradition had a second observation that maps even more precisely. The Heart houses the shen, the spirit-mind, and it is the sovereign of the whole ensemble โ the Suwen's famous cabinet metaphor gives it the throne. Sleep, in that system, is the sovereign returning to residence.
And the Suwen's chapter on emotional causes of disease includes a line that reads like it was written for the 1 AM feed: prolonged thinking knots the qi. Rumination, looping attention, the mind gnawing the same bone โ that is the mechanism, named, with the consequence spelled out. Everything tightens and stops moving.
A boundary, stated plainly: none of this is offered as treatment. These passages are diagnostic metaphors โ a pre-scientific system describing patterns with the tools it had. They earn their place here because they notice something the modern framing keeps rediscovering, not because they prescribe a fix. The fixes in this article come from the research side.
- If you borrow one classical idea, borrow this one: sleep is a descent. Ask what in your last hour is actively preventing it โ brightness, outrage, unfinished loops โ and start there.
The Phone in Bed Does Two Different Jobs
It is tempting to file all of this under blue light, but the phone's effects split into two different problems, and only one of them is optical.
The optical problem is real. Anne-Marie Chang's team at Brigham and Women's Hospital ran a small inpatient crossover study โ twelve adults, tightly controlled โ comparing four hours of reading on a light-emitting tablet against four hours in a printed book before bed.
The tablet delayed the circadian clock, suppressed melatonin, pushed sleep latency out by roughly ten minutes, cut REM sleep, and left people groggier the next morning. A ten-minute latency increase sounds small until you notice the clock shift: the device was not just delaying sleep, it was arguing with the body about what time it was.
The content problem is harder to measure and probably bigger. A 2016 Belgian survey of 844 adults aged 18 to 94 โ Exelmans and Van den Bulck's work โ found that six in ten took a phone to bed, and that messaging or calling after lights out predicted worse sleep quality, more insomnia symptoms, and more fatigue.
Age complicated the picture in an interesting way: for younger adults, more bedtime phone use tracked with later rise times and more fatigue; for the oldest group, it tracked with shorter sleep. The mechanism the authors emphasized was not light alone โ it was arousal. A conversation at 1 AM is stimulating regardless of spectrum.
So is the algorithmic carousel, which is built by enormous teams of engineers for exactly that property.
And there is a third layer, easy to miss: the phone is the prosthetic excuse. As long as it is the alarm clock, the late-night notepad, and the morning's first contact, it has a standing invitation to the nightstand, and every other plan for the evening has to negotiate with it. That is the practical thread the rest of this article pulls on.
Why the Discipline Advice Keeps Failing
The standard advice โ just decide to go to bed and go โ assumes the behavior is a decision problem. The research says it is closer to a friction problem. Kroese's original finding pointed at self-regulation: bedtime procrastination travels with the same trait patterns as general procrastination, and nobody solves general procrastination by being told to stop procrastinating.
There is a modest intellectual honesty debt here. The idea that willpower is a depletable fuel took real damage in the last decade โ replication attempts on the classic ego-depletion experiments came back mixed, and some came back empty.
So the strongest claim available is a modest one: self-control performance varies with state and circumstance in ways people systematically overestimate, and bedtime is exactly the wrong hour to run an experiment on your own reserves. You are tired, decision quality is at its daily low, and the opposition is a recommendation algorithm with a few billion hours of engagement data.
The practical upshot, which behavior-change research supports well enough: make the environment cast the vote instead. People who never see the phone after 11 PM do not need the willpower to resist it at 11:40. People whose evening has a shape โ a closing ritual, a sequence, a physical cue โ do not spend the last hour deciding every five minutes whether to start winding down.
None of this requires believing any particular theory of willpower. It only requires noticing that the current arrangement asks you to win a negotiation nightly, against a professional, at your worst hour.
One more thing worth saying: the goal is not an eleven o'clock bedtime imposed from nowhere. It is a bedtime you chose in advance, respected by an environment that makes the choice easy. If the honest bedtime is 12:30, build for 12:30. The mechanism being repaired is the gap between intention and action, not the clock time itself.
The Mechanics That Actually Move Bedtime
The concrete version, assembled from what the studies above actually tested. No step requires more than a week to try.
The anchor move is moving the charger out of the bedroom. This single change deletes the standing invitation described earlier: the phone stops being the alarm, so it stops having a job at the bedside, so it can charge in the kitchen without a policy debate each night.
A ten-dollar battery alarm clock fills the vacancy โ the companion piece to this article on this site reviews the options โ and the phone's overnight absence turns out to matter more than any screen-time setting, because the settings live inside the phone and the phone is the problem.
Second, the restriction trial. The most direct experimental evidence available is a four-week randomized pilot from a Chinese team: 38 college students instructed to stop using their phones in the thirty minutes before bed fell asleep faster, slept longer, and reported better sleep quality and calmer pre-sleep arousal than the control group.
Thirty-eight people is a pilot, not a verdict โ the honest scale of the claim. But the intervention costs nothing and the pilot is the kind a real-world attempt replicates: not 'no screens tonight,' a rule with the lifespan of a mayfly, but a repeating thirty-minute window with a replacement activity already in it.
Third, the replacement. The thirty minutes needs content or it collapses back into the phone. Paper works โ a book, or, for the ruminators, a notebook: the day's loose ends written down, on the theory (Suwen's, mostly) that thinking which lands on paper stops circling the ceiling.
Fourth, the implementation intention, which sounds managerial and is not: decide the sequence once, in if-then form โ after the kettle goes on, phone goes to the kitchen; after the kitchen, book. Deciding nightly is the failure mode. Decide once, then obey yourself.
Small, physical, dull. Which is roughly how the last hour of the day should feel.
- Charger moves to the kitchen tonight, before any other step. Everything else in this list is downstream of that one.
- Write the if-then sequence on paper and put it where the phone used to sit. The cue should live where the habit died.
- Give the trial two weeks before judging it โ the first three nights feel wrong in a way that reverses.
What We Can and Cannot Claim
The ledger, in the open.
Solid enough to act on: bedtime procrastination is a recognized, measured behavior; it correlates with insufficient sleep and insomnia reports across several surveys; restricting pre-bed phone use improved sleep measures in at least one randomized pilot; light from screens delays the circadian clock in controlled settings.
Weakest links: nearly everything is self-reported; the revenge variant specifically runs far ahead of its literature; causal direction between poor sleep and the procrastination is unresolved; the pilot trials are small.
Also worth being plain about: chronic difficulty falling asleep is not always self-inflicted timing. Insomnia disorder, sleep apnea, restless legs, medication effects, and mood disorders all masquerade as bad habits, and the fix for a masquerade is a clinician, not a kitchen charger.
If falling asleep is hard even on nights the phone stayed out, or snoring is interrupting your nights, this article's framework is the wrong tool and a proper evaluation is the right one.
The older texts would probably not be surprised that the modern problem is not falling asleep but stopping. A sovereign that never leaves the throne room ends up governing nothing well โ the Neijing's cabinet metaphor makes that argument in six characters.
Whether the price shows up as fatigue, mood, or just the sense that days have no edges, the mechanism is the same: nothing ends anymore, so nothing can begin again.
Open question, then, for the next time 12:47 finds you watching a barn get restored: what would the last hour of your day need to contain, for going to bed to feel like getting something instead of giving something up? That question has no algorithm behind it. Which may be the point.