Job Tenure Statistics (2026)

Median employee tenure was 3.9 years in January 2024, meaning half of workers had been with their employer longer than that and half less.

That is down from 4.1 years in January 2022 and is the lowest reading since January 2002.

The BLS notes that tenure moves for two quite different reasons: how often people change jobs, and the age profile of who is working. A workforce that gets younger will show falling tenure without anyone changing behaviour.

Median tenure was 2.7 years for workers aged 25 to 34 and 9.6 years for those aged 55 to 64.

The short-tenure figures show the same thing from the other end: 70% of 16 to 19 year olds had been in their job a year or less, against 10% of workers aged 55 to 64.

Almost every other cut of this data is a restatement of age. That is worth knowing before reading anything into the differences between groups.

28% of White workers had ten or more years with their employer, against 25% of Asian, 22% of Black and 22% of Hispanic workers.

The BLS explains this directly: White workers tend to be older. 23% of White wage and salary workers were 55 and over, against 19% of Black, 18% of Asian and 16% of Hispanic workers.

It is a useful example of a gap that looks like a difference in behaviour and is mostly a difference in composition.

Among women aged 25 and over, median tenure was 3.8 years without a high school diploma, 4.9 years with an associate degree and 4.7 years for college graduates.

For men, the BLS reports that tenure for those without a high school diploma was little different from other education levels.

Education is one of the strongest predictors of pay in US labour data and one of the weakest predictors of how long someone stays.

Retirement plan vesting schedules were designed around a workforce that stayed. Cliff vesting can require up to three years before an employer's contributions become yours, and graded vesting up to six.

At a median tenure of 3.9 years, and 2.7 years for workers in their late twenties and early thirties, a large share of the workforce is leaving before a six-year graded schedule completes.

Your own contributions are always yours immediately. It is the employer's money that walks out of the door, and it does so silently.

The vesting schedule and your position on it, because a few weeks can be worth thousands of dollars if a cliff or a step is close.

Whether an annual employer contribution requires being employed on a specific date, which is common and is not the same as the vesting schedule.

Any unvested equity and its cliff, which for a one-year option cliff is the single most expensive date in an early departure.

Defined benefit pensions commonly require five years of service to vest at all, which is above the current median tenure.

Only 14% of private industry workers have access to a defined benefit plan, so this affects a minority, but for that minority the cliff is absolute rather than graded.

It is also why public sector tenure runs so much longer than private: the benefit structure rewards staying in a way most private plans no longer do.

Changing employer is generally the fastest way to a large pay increase, which is a well-established pattern in US wage data and is a large part of why tenure keeps falling.

The offsetting cost rarely appears in the comparison: forfeited unvested employer contributions, a new waiting period before plan eligibility, and a reset vesting clock.

A raise of a few thousand dollars can be genuinely smaller than what is left behind, and the only way to know is to price both sides.

The vested balance is portable. It can stay in the old plan, roll into the new employer's plan, or roll into an IRA.

Leaving a series of small balances scattered across former employers is the common failure mode, and it is expensive because forgotten accounts are rarely invested well and are sometimes cashed out automatically.

For someone with median tenure, a full career now means five to ten former employers, which is a consolidation problem that a single-employer career never had.

Falling tenure is often read as a generational change in attitude. The BLS data does not support that reading on its own.

Tenure falls when hiring is strong, because more of the workforce is newly hired, and it rises in recessions when nobody moves. It is partly a labour-market thermometer.

It also fell to its lowest reading since 2002 in a period of unusually strong hiring, which is exactly what that mechanism predicts.

Employee tenure is measured in a January supplement to the Current Population Survey, sponsored by the Department of Labor's Chief Evaluation Office and run biennially since 1996.

The CPS is a monthly survey of about 60,000 households covering the civilian noninstitutional population aged 16 and over.

January 2024 is the most recent published survey. The next is January 2026 and has not yet been released, so the figures on this page are the current official ones rather than current-year ones.

Every tenure figure is from the BLS news release Employee Tenure in 2024, USDL-24-1971, published 26 September 2024, tables 1 through 3.

The defined benefit access figure is from the BLS National Compensation Survey, Employee Benefits in the United States, March 2025.

The vesting schedule table is indicative plan design rather than a published statistic, and it says so. An individual plan document is the only authority for a specific employer.