- A 2025 FlexJobs report found 23% of US workers took no days off at all in the preceding year, and most leave at least some annual leave unused.
- Jessica de Bloom's research found the wellbeing benefits of a vacation typically fade within two to four weeks of returning, and sometimes faster.
- The fade-out is not evidence that vacations are pointless. It is evidence that they are being asked to fix a chronic condition with an acute intervention.
- The dominant stated reason for not taking leave is workload, followed by concern that taking it signals reduced commitment. Neither is irrational in 2026.
- PTO-maxxing, the practice of stacking leave around public holidays, optimises the calendar without addressing the reason the leave was going unused.
One research-backed insight per week on stress and nervous system regulation — free.
No spam. Unsubscribe anytime.
In This Article
Nearly a quarter of American workers took no days off at all last year. Not a short break, not a long weekend built out of a public holiday. Nothing. The 2025 FlexJobs Work and PTO Pressure Report put the figure at 23%, and among those who did take leave, most left some of it unused.
It is late July. The stretch of the year in which taking time off is most socially permitted is more than half gone, and a large number of people are going to arrive at Labor Day having quietly not used the thing they are entitled to.
The interesting question is not why people are bad at resting. It is why the leave keeps going unused when the entitlement is sitting there, unspent, costing nothing to claim. The answers turn out to be fairly rational, which is the part that makes them difficult.
Why the Days Go Unused
Survey data on this is remarkably consistent, and two reasons dominate.
The first is workload. This is the one that gets dismissed as an excuse and should not be. Taking a week off does not remove a week of work from your life. In most roles it relocates it, into the fortnight before and the fortnight after. What is nominally a week of rest is in practice a week of absence purchased with two weeks of intensified effort.
Run the arithmetic on that and the calculation people are making stops looking like a failure of self-care. A week that costs two weeks of elevated load, and which the research says will stop helping within a month, is a genuinely marginal trade. Plenty of people have run that calculation, consciously or not, and concluded it is not worth it. They are not wrong about the arithmetic. They are trapped in a system where the arithmetic comes out that way.
The second reason is signalling. Workers report concern that visible use of leave reads as reduced commitment.
In 2026 that concern is not paranoia. AI was cited in over 100,000 US job cuts in the first half of this year, and in a June survey of 1,000 US employees, more than a third said they believed going above and beyond would not save their job. When a workforce believes its position is precarious and only loosely connected to performance, the calculation around visible absence changes. Being away is a risk with no offsetting protective effect. We wrote about the broader psychology of that state in Quiet Cracking.
So the unused leave is not a mystery of human psychology. It is what you get when the cost of taking it is real and the benefit is short-lived.
The Uncomfortable Finding About Vacations
Which brings us to the second half of the problem, and it is the part that rarely gets said out loud in wellbeing communications.
Vacations stop working quickly.
Jessica de Bloom, whose research programme has focused specifically on vacation effects, has found across multiple studies that the health and wellbeing improvements produced by a holiday typically return to baseline within two to four weeks of getting back. In some studies the fade-out is faster than that, closer to a week.
This finding is often reported as though it were evidence that vacations are useless. It is not. The improvements are real and measurable while they last. De Bloom's work consistently finds genuine gains in health, mood, energy and sleep quality during and immediately after time off.
The correct reading is narrower and more useful: a vacation is an acute intervention, and it is being asked to resolve a chronic condition.
If you return to the same workload, the same lack of control, the same volume of communication, the same conditions that produced the depletion, then the depletion re-establishes itself. Not because you rested wrong, but because the input that caused it is still running. The vacation was a withdrawal from a debt that continues to accrue. We made the fuller version of this argument in Why Vacations Don't Rest You.
Why Longer Is Not Proportionally Better
There is a related finding that changes how the calendar should be used.
The benefits of a vacation do not scale linearly with its length. The research generally finds improvements peak within roughly the first week and then plateau. A two-week break is not twice as restorative as a one-week break.
Combine that with the fade-out timeline and a fairly clear implication falls out. If the effect plateaus after about a week, and decays within two to four weeks regardless of how long the break was, then a single annual two-week absence is close to the worst available way to distribute a fixed quantity of leave. It produces one good fortnight and eleven months at baseline.
Several breaks of four to seven days, spread across the year, keep you nearer to recovered more of the time.
The caveat is real, though. Short breaks make psychological detachment harder to achieve, because detachment is not instant. Sabine Sonnentag and Charlotte Fritz's 2007 work in the Journal of Occupational Health Psychology identified detachment as the strongest determinant of whether time off actually reduces exhaustion, and it typically takes a few days to establish. The first two days of a break are often spent still mentally at work.
So the practical shape is not a scatter of single days, which rarely achieve detachment at all, and not one long annual reset, which wastes most of its own benefit. It is several breaks long enough to detach and frequent enough to matter.
What PTO-Maxxing Does and Doesn't Solve
The tactic circulating this year is PTO-maxxing: placing leave days adjacent to public holidays and weekends so that a small number of days produces a much longer continuous break. One day either side of a three-day weekend turns three into five. Applied deliberately across a year, it can substantially extend total time away from a fixed allocation.
As a calendar tactic it is genuinely good. It is precisely the distribution pattern the fade-out research supports, and it does so at low cost.
It is worth being clear-eyed about what it does not touch. PTO-maxxing optimises the arrangement of leave. It does nothing about the workload that makes leave expensive to take, and nothing about the signalling risk that makes it feel unsafe to take. A worker who was not taking leave because of either of those does not start taking it because they have learned to place it more cleverly.
There is also a mild irony in the framing, which is that a strategy for maximising rest efficiency is itself an optimisation behaviour applied to the one domain that is supposed to be exempt from optimisation. It is a much more benign example than most, and it does not merit a lecture. But it belongs to the same family we describe in The Optimization Paradox, and it is worth noticing when the calendar becomes a project.
The Version That Actually Helps
If the fade-out is real and the workload cost is real, what is the honest recommendation?
Stop treating annual leave as the primary recovery mechanism. This is the central reframe. If the effect of a holiday decays within a month, then a system that relies on holidays for recovery is going to spend most of the year unrecovered by design. The recovery that determines your baseline is daily and weekly, not annual. It is the evening that ends, the weekend day with nothing scheduled, the hour that belongs to you. Those are the mechanisms carrying the load, and they are the ones worth defending.
Take shorter breaks more often. Four to seven days, several times a year, beats one long absence, given both the plateau and the fade-out. PTO-maxxing is a good tool for making this cheap.
Protect the detachment specifically, not just the time. A week away spent checking email delivers materially less than a long weekend spent genuinely disconnected, because detachment rather than duration is doing the work. If you have to choose between a longer break and a properly disconnected one, choose the second. Which Type of Rest Do You Actually Need? covers how to tell whether you are achieving it.
Do not expect the break to solve the job. This is the most useful thing on the list. A great deal of post-vacation disappointment comes from having expected an intervention to fix something it structurally cannot. If the conditions produced the exhaustion, the exhaustion returns when you return to the conditions. Knowing that in advance removes a layer of self-blame that people otherwise carry, in which the failure of the holiday to fix things becomes evidence that they rested wrong.
Notice the guilt separately. For a significant number of people, the reason the days go unused is not workload or signalling. It is that time off without a justification feels indefensible. That is a distinct problem with a distinct fix, and we wrote it up in Why Do You Feel Guilty When You Rest?.
What the Unused Days Actually Represent
There is a version of this essay that ends by telling you to book the leave. It would be reasonable advice and it would miss most of what is going on.
Unused annual leave is not primarily a failure of individual resolve. It is a fairly precise measurement of the gap between what an organisation formally offers and what it actually makes possible. A benefit that a quarter of workers cannot use is not really a benefit. It is an entitlement with conditions that are not written down.
And the fade-out finding says something similar from the other direction. If a week of genuine rest is erased within a month by a return to normal conditions, then normal conditions are the thing generating the deficit, and the annual holiday was never going to be more than a temporary offset against it.
None of which is an argument for not taking the days. Take them. The gains are real while they last, and several shorter breaks will do more for you than the single heroic absence you have been postponing since spring.
Just do not take them expecting a reset. Take them knowing that what you are buying is a few good weeks, and that the thing determining the other forty-eight is happening on ordinary Tuesdays.
Get one essay a week on the science of rest and the cost of optimization culture. Subscribe to the newsletter. Free, one essay a week, no upsell.
Related reading: Why Vacations Don't Rest You · Why Do You Feel Guilty When You Rest? · Slow Summer Incubation · Everyone Else's Summer Looks Better
The research is actionable. So is the newsletter. One insight per week, no fluff.
No spam. Unsubscribe anytime.
Frequently Asked Questions
Why do people not use their vacation days?
Two reasons dominate the survey data. The first is workload: taking leave does not reduce the work, it compresses it into the surrounding weeks, so a week off costs two weeks of intensified effort either side. The second is signalling: workers report a concern that visible use of leave reads as reduced commitment. In 2026 that concern has sharpened considerably, with AI cited in over 100,000 US job cuts in the first half of the year alone. Neither reason is irrational. Both are responses to conditions rather than personal failures of self-care.
How long do the benefits of a vacation last?
Not long. Jessica de Bloom and colleagues, reviewing and running studies on vacation effects, found that improvements in health and wellbeing typically return to baseline within two to four weeks of getting back, and in some studies within a week. Her research also found that the benefits are real while they last, so the finding is not that vacations do nothing. It is that a vacation functions as an acute intervention, and that acute interventions do not resolve chronic conditions. If the work you return to is unchanged, the state you were in before it will re-establish itself on roughly a fortnightly timescale.
What is PTO-maxxing?
PTO-maxxing is the practice of scheduling annual leave adjacent to public holidays and weekends so that a small number of leave days yields a much longer continuous break. Placing one or two days around Memorial Day, July 4 or Labor Day can turn a three-day weekend into a five-day break, and applied across a year it can substantially extend total time off. It is a genuinely sensible calendar tactic. It is worth being clear about what it does not do: it optimises the arrangement of leave without addressing the workload and signalling pressures that cause leave to go unused in the first place.
Does a longer vacation work better than a short one?
Somewhat, but less than people expect, and the relationship is not linear. The research on vacation length generally finds that health and wellbeing improvements peak within the first week or so and then plateau, meaning a two-week break is not twice as restorative as a one-week break. What matters more than duration is the quality of psychological detachment achieved, which Sabine Sonnentag and Charlotte Fritz identified in 2007 as the strongest determinant of whether time off actually reduces exhaustion. A fortnight spent monitoring email delivers less than a well-detached long weekend.
Is it better to take one long break or several short ones?
The evidence leans toward more frequent shorter breaks, given that benefits fade within two to four weeks regardless of how long the break was. If the restorative effect has a roughly fortnightly half-life, then distributing time off across the year keeps you closer to baseline more often than a single annual reset does. The practical caveat is that shorter breaks make full psychological detachment harder to achieve, since detachment takes a few days to establish. The reasonable compromise for most people is several breaks of four to seven days rather than either one long absence or a scatter of single days.