Psychology says people who find it hard to disconnect from work when they get home aren't workaholics or ambitious overachievers — they're often people who learned that being useful was the safest way to feel secure

It’s 7pm. Dinner is on the table, the laptop is shut, the phone is face down, and somewhere behind the eyes a message thread from four hours ago is still quietly running. Nobody is working. The work is simply still on.

The usual label for this is workaholism, delivered by a partner with a raised eyebrow. It is a tidy explanation and often a wrong one, because workaholism describes something much narrower than most people assume.

Workaholism means something quite specific

Occupational psychologists treat workaholism as two things happening at once: working excessively, and working compulsively. Wilmar Schaufeli and colleagues have argued for years that the compulsive part is the discriminating ingredient. Engaged workers can put in similar hours while enjoying themselves, and what they lack is the inner drive they cannot resist.

That distinction matters at 11pm. Plenty of people who cannot stop thinking about work are not chasing more of it. They would take less in a heartbeat. What they cannot do is set it down.

Two ways of thinking about work at 10pm

Mark Cropley and Fred Zijlstra split after-hours work thinking into three types: affective rumination, problem-solving pondering, and detachment. In plain English, that is stewing, planning, and letting go. The three feel roughly identical from the inside, which is part of the trouble.

Stewing is the expensive one. In a survey of 719 workers across a range of occupations, affective rumination was the strongest predictor of increased fatigue, while problem-solving pondering predicted less of it. Both keep the lights on upstairs, and only one of them leaves a mark by morning.

So the diagnostic question is what kind of thought is running. Working out tomorrow’s staffing problem in the shower is a different animal from replaying a meeting and feeling slightly ill.

Where the habit of being useful gets built

Jennifer Crocker and Connie Wolfe proposed that self-esteem is not one flat number. It rises and falls in whichever domains a person has staked their worth on. Stake it on being competent, and a mediocre performance review lands as a verdict on the entire self rather than a note about one quarter.

Where does the staking happen in the first place? One well-studied route is parental conditional regard, the pattern where affection and approval seem to arrive with conditions attached. Children who experience it tend to absorb the standard in a shallow, pressured way that self-determination theorists call introjected regulation. The everyday translation is a sense of internal compulsion, the feeling that something inside is doing the pushing, plus shame when the standard slips. A 2023 meta-analysis found consistent links to contingent self-esteem and depressive symptoms.

Worth a caveat. That literature is about parenting, measured mostly in adolescents and young adults. Applying it to a 41-year-old who checks email in the bath is a reasonable inference rather than a demonstrated finding.

Switching off is a skill, and stress makes it harder

Recovery researchers use the term psychological detachment for the state of being mentally away from work during off hours. Sonnentag’s own research team has tracked this day by day and found that detaching in the evening is linked to less exhaustion at bedtime, which in turn predicts waking up more recovered.

There is a cruel wrinkle here, one Sonnentag calls the recovery paradox: recovery processes tend to break down exactly when job stressors are highest, which is exactly when they are needed. The evenings that most demand a clean break are the evenings when getting one is hardest.

What tends to actually help

Unfinished tasks are a reliable trigger. One study followed employees into their weekends and found that jobs left hanging on Friday predicted worse weekend sleep, with affective rumination doing the damage in between.

Happily, finishing everything is not the fix, since nobody ever does. In a small laboratory experiment, 57 healthy young adults spent five minutes writing before bed. Those who wrote a specific to-do list for the days ahead fell asleep around nine minutes faster than those who journalled about what they had already finished. One study, one narrow sample, so hold it loosely. The underlying move, getting the loop out of the head and onto paper, costs nothing to test.

Three things are worth trying beyond that. Write the shutdown list at the desk rather than in bed. Notice which mode is running and, when it is stewing, name the fear underneath it, which is usually some version of being found out or found useless. Then run small experiments in being unhelpful. Leave the message until morning, decline the extra task, slip away for an afternoon without telling anyone, and watch closely for the consequences that never arrive.

None of this argues for caring less about the job. The experiment is simply to stop being useful for three hours and notice whether it feels like rest or like exposure. For a lot of people, the answer is informative.