The people who scroll longest after the lights should already be off usually aren’t chasing another laugh, another video, another dopamine hit — at least according to two separate strands of research, run years apart on entirely different questions, that point toward the same underlying behavior when read together: they’re avoiding the exact moment they’d otherwise have to sit alone with their own thoughts, not chasing whatever happens to be on the screen.
Neither study measured late-night scrolling directly. This is an inference from putting the two side by side, not a finding either research team set out to prove.
What happens when the entertainment is taken away entirely
The foundational finding here has nothing to do with phones. In a 2014 Science study, the psychologist Timothy Wilson and his colleagues asked people to simply sit alone in a room with their thoughts for between six and fifteen minutes, no phone, no book, nothing to do but think. Across a long series of experiments, most participants rated the experience as unpleasant, difficult to sustain, and something they had to actively resist cutting short. In the most widely cited version of the study, a subset of participants who had already said beforehand that they’d rather pay a small amount than be shocked again chose to shock themselves anyway rather than sit with their thoughts for even a few empty minutes: 67 percent of men and 25 percent of women, among those who had that option, shocked themselves at least once. Being alone with an unoccupied mind wasn’t neutral. For a lot of people, it was measurably, consistently aversive, not just mildly unpleasant.
That finding predates smartphones being the default distraction of choice, which is part of why it’s useful here. It establishes that the discomfort of unstructured, thought-only time is real and strong on its own, with no screen anywhere near the experiment.
What made the result hold up under scrutiny was how many ways the researchers tried to make the unoccupied time more tolerable before concluding it genuinely wasn’t. They tested the effect at home rather than in an unfamiliar lab, in case an unfamiliar room was the actual source of the discomfort. They tested it with people given advance suggestions for pleasant things to think about, in case the problem was simply not knowing what to think. Across each variation, the core discomfort persisted. The unoccupied mind wasn’t struggling because of the setting or a lack of good material to think about. It was struggling with the basic condition of having nothing external to attend to.
What happens right at the edge of sleep specifically
Separately, researchers studying what they termed bedtime procrastination — delaying sleep later than intended with no external reason for doing so — found the behavior tracked with weaker self-regulation, not simply with enjoying the delaying activity more than sleep. The term was coined in a 2014 study of 177 US adults. A larger 2016 follow-up, surveying more than 2,400 Dutch adults, found a similar pattern: the later bedtime wasn’t well explained by people simply having something better to do. It tracked more closely with a breakdown in the ability to follow through on the intention to stop, at precisely the point in the day when willpower and cognitive resources are at their lowest ebb and the mind is least defended against whatever it’s been successfully avoiding since morning.
Put the two findings next to each other and a shape emerges. Bedtime is the one part of the day explicitly designed to remove all remaining distraction. The lights go off, the tasks are done, and for a few minutes before sleep arrives, a person is left with something close to the exact unoccupied, thought-only state Wilson’s participants found so difficult to sit inside. A phone within reach at that exact moment isn’t competing with sleep for entertainment value. It’s competing with the one part of the day that most resembles the empty room in the shock experiment, and for anyone carrying unresolved thoughts, worries, or low mood, the phone is simply the nearest available exit.
This also helps explain why the pull feels so much stronger at night than during an equivalent idle moment earlier in the day. A quiet ten minutes at midday still has competing claims on it: something else to do next, people around, daylight signaling the day isn’t finished. A quiet ten minutes at midnight has none of that scaffolding left. The mind is more depleted, the usual defenses against unwanted thoughts are weaker, and the room is, both literally and psychologically, closer to Wilson’s empty chamber than anything encountered during waking hours.
Why this reframes what “just entertainment” means at night
This matters because it changes what the actual craving might be. A person reaching for a video at 3 in the afternoon during a boring meeting is plausibly bored and looking for stimulation. A person reaching for the same video at midnight, having already scrolled for forty minutes past their intended bedtime, is more plausibly avoiding a specific internal state rather than seeking a specific external one. The specific content may barely matter in the second case. Almost anything sufficiently absorbing could serve the same function, which is one reason the specific app or video someone lands on late at night is often forgettable within days, while the pull to keep scrolling something, anything, remains fairly consistent.
What actually follows from this
Framing this as a willpower problem alone doesn’t fit what the bedtime-procrastination research actually found; it specifically frames the behavior as a self-regulation failure, not a bad habit that discipline can override on a tired brain at the end of a depleting day. It does suggest that treating the urge as being about the content, another episode, another feed refresh, misses what’s actually driving it for at least some people.
The more useful question, for anyone who notices the pattern in themselves, may be less “what do I want to watch” and more what’s waiting in the fifteen minutes of unoccupied thought the phone is being used to skip, and whether that’s something worth sitting with rather than outrunning every night.
Neither study offers a clean fix for that. Together, they offer a more accurate name for what might be happening when the screen is still lit and sleep is still on the other side of it.