PTO Statistics 2026: Twenty Years of US Service Buys the Leave a European Gets on Day One

Paid time off is the benefit everyone assumes is settled. It is not measured like one. The US has no statutory minimum, so "how much leave do people get" is an empirical question about employer behaviour rather than a legal fact — and the answer splits sharply by wage, hours, employer size and industry. Then a second question sits on top of it: how much of that leave is actually taken.

This page assembles the verified 2026 dataset across all four layers — who has PTO, how many days they get, how many they take, and how that compares to statutory entitlements elsewhere. Every figure is attributed to a named source with its reference period stated, because in this topic the vintage matters: the most-quoted "average vacation days" numbers circulating in 2026 are from 2021 fieldwork, and the most-quoted dollar figure for unused leave is a vendor estimate from a 1,500-person consumer panel. Sources are listed in Methodology and sources. The statistics that circulate widely but do not say what people think they say are in section 9, not repeated as fact.

1. Who actually has paid time off

The authoritative US measure is the Bureau of Labor Statistics' National Compensation Survey, whose March 2025 estimates were published on 25 September 2025. It reports access — whether a benefit is available to the worker, not whether they use it.

The government column surprises people. Public-sector workers have the best sick leave in the country (93%) and the worst vacation access (61%) — because a large share of state and local employment is school-year teaching staff, whose contracts are built around the academic calendar rather than an accruing vacation bank.

Access is not evenly distributed. The single largest determinant is not wage, industry or employer size — it is hours.

Industry compounds the same effect, because part-time and low-wage work is concentrated in the same sectors:

2. How many days — and how slowly they arrive

For workers who do have paid vacation, the BLS publishes the average number of days granted at each service milestone. This is the table to quote when someone asks what "normal" PTO looks like in the United States.

Two things are worth noticing. First, the ladder decelerates hard: the first five years deliver four extra days, the next five deliver three, and the following decade delivers two. Nearly half of everything tenure will ever buy arrives in the first quarter of a twenty-year career. Second, sick leave does not move at all in private industry — seven days at year one, seven days at year twenty. Tenure rewards vacation; it does not reward illness.

Averages hide a wide distribution. After one year of service, private-industry workers split almost evenly between 5–9 days (31%) and 10–14 days (31%), with 21% at 15–19 days. After ten years, 31% get 15–19 days, 24% get 20–24, and 20% get more than 24. On top of vacation, US employers providing paid holidays average eight per year — a figure from the BLS March 2018 holiday profile, which is the most recent detailed publication and should be quoted with that date attached.

3. The days that are never taken

Entitlement is not use. The cleanest measurement of the gap comes from the Pew Research Center, which surveyed 5,188 US workers between 6 and 12 February 2023: 46% of workers who receive paid time off take less than their employer offers.

The pattern inside that number is counter-intuitive. It is not the precarious who leave leave unused — it is the senior:

The most common reason Americans give for not taking their leave is not fear of the boss. It is that they do not feel they need it — a reason that gets less convincing the moment you notice 65% of the same workforce reports feeling vacation-deprived.

4. Why the days go unused

Pew asked workers who take less than offered why. Respondents could give more than one reason, so these do not sum to 100%:

Three findings inside that chart deserve to be quoted more than they are. Guilt is gendered: 48% of women cite feeling bad about colleagues absorbing their work, against 39% of men, rising to 53% among women aged 18–49. Fear is racialised: 21% of Black workers name the risk of losing their job as a reason, against 13% of White workers. And fear is front-loaded by tenure: 24% among workers with less than a year of service, 21% at one to five years, 11% beyond that. A PTO policy that looks identical on paper is experienced very differently depending on who is reading it.

5. The United States versus everyone else

US federal law requires zero paid vacation days and zero paid holidays. The Center for Economic and Policy Research has documented this across several editions of its No-Vacation Nation report: the United States is the only advanced economy that guarantees its workers no paid time off of any kind. Everything American workers receive is employer policy, and can be changed by employer policy.

In the European Union, four weeks is not a benchmark but a floor. Article 7 of Directive 2003/88/EC requires every member state to guarantee at least four weeks of paid annual leave, and further specifies that this minimum cannot be replaced by a payment in lieu except where the employment relationship ends — the opposite of the American PTO-payout convention. Most member states go further, and collective agreements go further still: Eurofound puts the average collectively agreed annual leave entitlement across the EU at 24.3 days in 2022, with a split between the pre-2004 member states (25.3 days) and those that joined later (20.9 days).

Comparisons of days actually taken are murkier, because no statistical agency collects them internationally. The widely-cited country rankings come from Expedia's Vacation Deprivation consumer survey — 11,580 respondents across 11 countries, fielded 26 March to 3 April 2024 by Harris Research Partners. Its most defensible finding is not the league table but the trend: 65% of US workers report feeling vacation-deprived, the highest reading in eleven years of the series, and 93% of French respondents describe vacation as a basic right against 83% of Americans.

6. Unlimited PTO: the two-day benefit

Unlimited or "open" leave is the most discussed and least adopted policy in the field. SHRM's Employee Benefits Survey puts adoption at 7% of US employers, against 98% offering a conventional vacation policy — and the figure has been flat to slightly declining, not rising.

The employer fear is that people take too much. The data says the opposite. Analysis by Empower, reported by SHRM in October 2024, found employees with unlimited PTO take 16 days a year against 14 days under a defined policy — a difference of two days. Meanwhile the accounting effect runs entirely one way: with no accrual, there is no balance to carry, no liability to book and nothing to pay out at termination.

7. What unused leave costs — and who owns the balance

In an accrual system, untaken leave does not evaporate. It sits on the balance sheet as an accrued liability, growing with every pay rise, and in several states it cannot legally be taken away. California treats accrued vacation as earned wages under Labor Code §227.3: use-it-or-lose-it forfeiture is unlawful and unused vacation must be paid out at separation. Colorado, Montana and Nebraska take comparable positions. The distinction that matters in policy drafting is between a cap (you stop accruing above a ceiling — lawful everywhere) and a forfeiture (you lose what you already earned — unlawful in those states).

The dollar figures attached to this are shakier than they look. The most-quoted estimate — that Americans held more than $312 billion in unused vacation, affecting some 68 million people — comes from Sorbet, a PTO-management vendor, based on a Sapio survey of 1,500 US employees aged 21+ fielded in December 2023. The same survey found 62% did not use all their vacation days, women took about 10% less PTO than men, and workers aged 21–34 took roughly a week less than those aged 55+. It is a real survey with a stated method; it is also a vendor with a product to sell, and the total has been reported elsewhere as $224 billion and $318 billion. Divide it out and you get about $4,590 per affected worker ($312bn ÷ 68m) — useful for scale, not for a headline.

8. Benchmark cheat sheet

The numbers most often needed in a policy paper, a board deck or a journalist's sidebar, with the source and reference period attached to each.

9. Statistics to handle with care

Four figures dominate PTO coverage and none of them means what it is usually made to mean.

"Americans take 11 days a year; the French take 29." This league table comes from Expedia's consumer survey, not from any labour statistics agency, and the edition year is almost always dropped when it is quoted — the figures have moved between editions. Days taken are self-reported and the samples are online panels of roughly 1,000 per country. Cite it as a survey of attitudes, which it is good at, rather than as a measurement of leave, which it is not designed for.

"$312 billion of unused PTO." A vendor estimate extrapolated from 1,500 respondents, reported elsewhere as $224 billion and $318 billion. The direction is well supported; the precision is not. If you need a defensible number for a specific employer, calculate it from your own accrued-liability ledger.

"Unlimited PTO means people take less time off." Widely repeated, and the best available data contradicts it: 16 days versus 14. What is true is subtler — unlimited plans remove the accrued balance, so the employee loses a payable asset and the employer loses a liability. That is a balance-sheet argument, not a wellbeing one, and it should be made as such.

"The average American gets two weeks of vacation." Only after a year of service, only in the 77% of jobs that offer paid vacation at all, and only as an average across a distribution where 31% of one-year workers get 5–9 days. The sentence is true of the average and false of the median experience in low-wage and part-time work, where the honest figure is often zero.

10. What to do with these numbers

Methodology and sources

Every figure on this page comes from a named source with its reference period stated. Where a number is derived, the inputs and the arithmetic are shown next to it and it is labelled as derived. Nothing is averaged across sources with different universes without saying so.

Citation. If you use these figures, please cite the underlying source rather than this page, and link here as the compilation. Derived metrics — the full-use rate, the part-time cliff and the twenty-year gap — are original to this page and should be attributed to Rewordin with their inputs.