Most people who end up reading about this have already tried the obvious. The bedroom is cooler. Coffee stops after lunch. There was a wind-down routine, and it lasted about nine days. The hour at which the lamp actually goes off has not moved, and that is the interesting part, because advice which fails this reliably is usually not failing. It is aimed at something else.
Going to bed later than intended, with nothing outside yourself stopping you, has a name in the research literature. Kroese and colleagues described it in 2014 as a self-regulation problem rather than a sleep disorder, and that framing has held. What the framing does not supply is a next move, because the behaviour is a destination rather than a cause. At least four quite different situations arrive at the same half past midnight, and the only thing they have in common is the timestamp.
33.2%
of US adults slept under seven hours a night in 2020
80.6%
of young adults in one survey called themselves binge viewers
20.2%
of those binge viewed at least a few times a week
Sleep hygiene is advice for somebody trying to sleep
Cool the room, cut the caffeine, keep the hours regular, no screens for an hour before bed. Irish and colleagues reviewed the individual components in 2015 and found the evidence behind them uneven and far more individual than the uniformity of the advice suggests. Chung and colleagues showed in 2018 that education of this kind, delivered alone as a treatment for insomnia, performs poorly enough that guidelines no longer recommend it as a standalone intervention. It survives as public health messaging, which is a reasonable thing to be and not the same thing as a fix.
But the deeper mismatch is not about effect sizes. Every item on that list presupposes a particular person: in bed, lights off, genuinely attempting sleep, not succeeding. Somebody sitting upright at one in the morning with a lamp on and a clear view of the clock is not that person and never was.
The months of failed advice were not a failure of character. They were a routing error.
Four roads to the same half past midnight
| What is driving it | What the late hour is solving | What actually moves it |
|---|---|---|
| Hours you feel owed | No part of the day belonged to you | Time of your own in daylight, scheduled rather than found |
| Friction at the end | A procedure you have not started | Fewer steps, most of them done earlier |
| Dread of tomorrow | Lights off is when tomorrow begins | Dealing with tomorrow while still upright |
| Pull of the device | Nothing in it ever says stop | An ending supplied from outside |
The right-hand column is the whole argument. Four problems, four answers, and only the second one is even adjacent to sleep advice. Twelve situational questions sort between the four without scoring severity, which matters because the behaviour looks identical from outside whichever road it came down.
The hours that belonged to other people
Hill and colleagues interviewed twenty-eight people in the first years of their careers about what the late hour was for. Seven themes came out of the transcripts, and the one participants kept returning to was the need for time that was theirs. Alongside it sat two things that explain why advice bounces off this road: active self-negotiation to keep going, and complete knowledge of what sleep is worth. Nobody there needed informing. The title the authors chose came from a participant calling it satisfying and destructive at once.
An hour nobody assigned to you is worth more than an hour of rest, which is why rest does not substitute for it.
That line has an evidence base behind it, from a literature not usually read next to sleep. Sonnentag and Fritz, building a measure of how people recover from work, found four separable experiences rather than one. Psychological detachment, relaxation, mastery, and control. Control is its own factor: the experience of deciding what happens in a stretch of time, independent of whether the stretch is restful. This is why a pleasant evening organised by somebody else leaves the urge for the midnight hour completely intact. Relaxation was delivered. Control was not.
So the prescription for this road is not a sleep prescription at all. It is that the day has to contain time that belongs to the person living it. Twenty minutes counts, and two conditions decide whether twenty minutes does anything. The content has to be chosen rather than assigned, and the slot has to be on the calendar rather than scavenged from a gap, because scavenged time delivers no sense of control and therefore does not satisfy the thing the midnight hour is satisfying. An evening of chores somebody happens to enjoy is not a substitute and never works as one.
There is a version of this where the honest answer is uncomfortable. For a parent of a small child, a carer, somebody on a rotating shift, somebody whose calendar is written entirely by other people, the late hour is solving a real problem with tomorrow's money, and being told to go to bed earlier amounts to being asked to surrender the only autonomy on offer. They will not do it. They should not be expected to. If a week genuinely contains no twenty minutes anywhere, the thing to work on is the shape of the week, which is a conversation about what gets refused or handed over, not about what time the light goes off.
Planning in advance does something measurable, and it is worth knowing exactly what. Valshtein and colleagues ran two trials of a short written exercise pairing the wanted outcome with the obstacle in the way, then fixing a concrete if-then plan. The gap between intended and actual bedtime narrowed in both. Self-reported sleep duration did not change, so the technique moves the discrepancy rather than buying hours.
The phrase for all of this, where it came from and what has been measured about it sit in the entry on revenge bedtime procrastination. This piece is about the routing.
The last ten steps
The second road looks like the first from across the room and is almost its opposite. The person wants to be asleep. They said so, out loud, forty minutes ago. What stands between the decision and the bed is a procedure, and somewhere in the procedure is a step they cannot make themselves start.
Nauts and colleagues named this directly in 2016. The aversion is not to sleep. It is to the preparatory tasks sleep is attached to: brushing teeth, washing a face, taking out contact lenses, changing clothes. None of it is difficult and all of it is tedious, arriving at the hour when tedium is least tolerable. Their preliminary data found that aversion to the bedtime routine contributed to how late people went.
Kamphorst and colleagues looked at the resource side of the same hour and found that people who had resisted more desires across the day procrastinated more at bedtime. It rests on self-report, and the depletion framework it sits inside has had a difficult time in replication, so the result describes how the evening feels rather than stating a law about a tank running dry.
Either way the fix is environmental and faintly mechanical. Anything that can be done at nine should be done at nine, while the near side of tired still has capacity. Strip the midnight list down to standing up and turning off a lamp. Put the charger in a different room, not as an act of virtue but because it converts a decision into a walk that has to happen anyway. And deal with the sofa, which is the actual trap on this road: falling asleep there does not skip the routine, it adds a waking to it, and the version of a person who has to get up at one in the morning has less left than the version at eleven.
What this road needs is a shorter list, not a stronger character. It is the cheapest of the four to fix and the one people are hardest on themselves about, because from outside it is indistinguishable from not caring.
Staying up against tomorrow
The third road is not about tonight. Sleep is the door tomorrow comes through, so the person stands in front of it with the lamp on.
Harvey's cognitive model of insomnia set out why the dark is the difficult part. Worry and the monitoring that comes with it maintain sleep problems on their own, independently of whatever started them, and the moment the room goes quiet is the first moment all day with no competing input. With something playing there is nowhere for the thinking to go. Without it, the thinking starts, and it starts with whatever has been outrun since four in the afternoon. So the next episode goes on, and none of it is about the episode.
Exelmans and Van den Bulck found something that fits this precisely. Binge viewing was associated with poorer sleep quality, more fatigue and more insomnia symptoms, while ordinary television viewing was not, and cognitive pre-sleep arousal fully mediated the relationship. The content is not the agent. The state of being switched on is.
Sunday evening is where this road becomes legible, and most people on it recognise the description. The same evening as any other, the same things being done, and the whole thing sitting differently because of what is on the other side of it.
This is also where sleep hygiene is most clearly pointed the wrong way. A darker, quieter, cooler, screen-free room is an exact description of the conditions the person is staying up to avoid, and handing it to them as a solution is handing them more of the thing they are postponing.
What helps is dealing with tomorrow while still vertical and still capable, rather than negotiating with it horizontally in the dark. One written line about the first thing that happens tomorrow, early in the evening, because a good deal of what the dark gets to work on is an unspecified tomorrow rather than a bad one. A fixed slot for worrying, in daylight, on paper, so it has somewhere else to be. For Sunday in particular, something actually wanted placed early in the evening rather than late, since the standard advice to keep Sunday restful tends to leave it empty and empty is the one thing this road cannot handle.
Underneath this road is usually anxiety, and sleep is simply where it becomes visible. If the days are spent braced as well as the nights, the bedtime is not the thing to treat. And the loop is worth naming because it does the real damage: the later the night, the harder tomorrow is, which makes the next night worse.
The device has no ending built into it
The fourth road deserves the honest version of the design argument, which is that none of it is accidental. Autoplay exists because it measurably increases watch time. Infinite scroll exists because a page with a bottom is a page people leave. The next round loads before the last one has resolved, because the gap is where somebody would have stood up.
Baughan and colleagues built a custom Twitter client and gave it to forty-three people to find out what the state actually is, and the answer was not addiction. It was normative dissociation: absorption deep enough that self-awareness and the sense of agency both drop, which is why the honest report afterwards is not remembering what was read. The useful half of that study is that design changes reduced it. Reading-history labels, custom lists, time-limit dialogues, usage statistics. The state belongs to the interface at least as much as to the person in front of it.
The responsibility does not land entirely on the product, though, and the version where it does is a dead end, because the design is not going to change on anyone's behalf. Putting an ending back in is the work. Autoplay off is one switch and the highest-yield minute on this road, since it converts a single decision at eleven into a decision at every episode. The endpoint decided before starting rather than during. One more made to cost an action: logged out so it needs a password, charger in another room.
And if the scrolling is not pleasurable at all, if it is flat and faintly grim, then the device is the instrument rather than the reason, and one of the other three roads is playing it.
When staying up is not the problem
One split matters more than anything else here, and it is easy to miss from inside.
If getting into bed at a sensible hour reliably produces forty minutes of lying there awake, then staying up is not procrastination. It is a rational response to a bed that has stopped working, and the thing to look at is sleep onset. The European insomnia guideline sets out what the diagnosis requires, and it is a different pattern with different treatment: difficulty initiating or maintaining sleep, plus a daytime cost, persisting over months. Hill and colleagues found difficulty falling asleep sitting among the themes in their interview data too, which is a reminder that the two coexist happily. Sorting out the going-to-bed part first is still worth doing, because it is impossible to judge how long sleep takes to arrive when every attempt starts at one in the morning.
Four roads, one timestamp. The reason it matters which one is yours is arithmetic: a month spent on the wrong one changes nothing, and the evidence that it was the wrong one looks exactly like evidence that you are undisciplined.
Sources
- Kroese, F. M., De Ridder, D. T. D., Evers, C., Adriaanse, M. A. (2014). Bedtime procrastination: introducing a new area of procrastination. Frontiers in Psychology, 5, 611.
- Irish, L. A., Kline, C. E., Gunn, H. E., Buysse, D. J., Hall, M. H. (2015). The role of sleep hygiene in promoting public health: a review of empirical evidence. Sleep Medicine Reviews, 22, 23-36.
- Chung, K. F., Lee, C. T., Yeung, W. F., Chan, M. S., Chung, E. W. Y., Lin, W. L. (2018). Sleep hygiene education as a treatment of insomnia: a systematic review and meta-analysis. Family Practice, 35(4), 365-375.
- Hill, V. M., Ferguson, S. A., Vincent, G. E., Rebar, A. L. (2024). It's satisfying but destructive: a qualitative study on the experience of bedtime procrastination in new career starters. British Journal of Health Psychology, 29(1), 185-203.
- Sonnentag, S., Fritz, C. (2007). The Recovery Experience Questionnaire: development and validation of a measure for assessing recuperation and unwinding from work. Journal of Occupational Health Psychology, 12(3), 204-221.
- Valshtein, T. J., Oettingen, G., Gollwitzer, P. M. (2020). Using mental contrasting with implementation intentions to reduce bedtime procrastination: two randomised trials. Psychology & Health, 35(3), 275-301.
- Nauts, S., Kamphorst, B. A., Sutu, A. E., Poortvliet, R., Anderson, J. H. (2016). Aversive bedtime routines as a precursor to bedtime procrastination. The European Health Psychologist, 18(2), 80-85.
- Kamphorst, B. A., Nauts, S., De Ridder, D. T. D., Anderson, J. H. (2018). Too depleted to turn in: the relevance of end-of-the-day resource depletion for reducing bedtime procrastination. Frontiers in Psychology, 9, 252.
- Harvey, A. G. (2002). A cognitive model of insomnia. Behaviour Research and Therapy, 40(8), 869-893.
- Exelmans, L., Van den Bulck, J. (2017). Binge viewing, sleep, and the role of pre-sleep arousal. Journal of Clinical Sleep Medicine, 13(8), 1001-1008.
- Baughan, A., Zhang, M. R., Rao, R., Lukoff, K., Schaadhardt, A., Butler, L. D., Hiniker, A. (2022). I don't even remember what I read: how design influences dissociation on social media. Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 18.
- Riemann, D., Espie, C. A., Altena, E., et al. (2023). The European Insomnia Guideline: an update on the diagnosis and treatment of insomnia 2023. Journal of Sleep Research, 32(6), e14035.
- Hill, V. M., Ferguson, S. A., Vincent, G. E., Rebar, A. L. (2024). Beyond circadian timing: a new model of bedtime procrastination. Sleep Health, 10(1), 5-6.
- Pankowska, M. M., Lu, H., Wheaton, A. G., Liu, Y., Lee, B., Greenlund, K. J., Carlson, S. A. (2023). Prevalence and geographic patterns of self-reported short sleep duration among US adults, 2020. Preventing Chronic Disease, 20, 220400.