Most US employees can shift their hours, but control has limits

A job can be described as flexible even when a worker has little say over next week's hours. In survey results released September 24, the US Bureau of Labor Statistics (BLS) found that 57% of wage and salary workers could vary when they began and ended work in 2024–25. Among those workers, 36% could change those times frequently, 45% occasionally, and 19% only rarely. The headline measure counts all three groups.

A separate question captures the opposite kind of uncertainty: how much notice a worker gets. Across wage and salary workers, 59% knew their schedule at least four weeks in advance, while 9% learned it no more than one day ahead. These figures do not measure the same thing as the ability to adjust a start time. A worker may have a fixed but predictable schedule, or flexible hours within a schedule set by someone else.

The distinction matters most when “flexibility” is used as a shorthand for worker control. BLS found that 36% of workers lacked flexible hours and had schedules set by employers without their input. Yet 70% of that group knew their schedule four or more weeks ahead. Employer control and short notice overlap, but neither implies the other.

Working from home is another question

The BLS survey estimates that 37% of wage and salary workers could work at home and 34% sometimes did. A narrower 28% had days on which they worked exclusively at home. These are different measures of permission, practice, and full workdays; none means that a third of employees held fully remote jobs.

The opportunities are uneven. Among workers age 25 or older, 57% with at least a bachelor's degree worked at home at least occasionally, compared with 15% of those with a high school diploma and no college. Among workers who could work at home, 78% could vary their start and end times; among those who could not work at home, 44% could. This association does not show that remote-work permission itself causes schedule control. Occupation and other job differences may affect both.

A separate Federal Reserve survey of 2025 job quality offers context from a different question: 39% of workers said they worked from home at least some of the time in the week before that survey. It also found 16% had a schedule that varied based on employer needs, while 10% had one that varied at their own request. Those numbers cannot be subtracted from the BLS figures to calculate a change. The Fed includes self-employed workers and asks about a recent week; BLS covers wage and salary workers' main jobs and asks whether they worked at home at least occasionally during 2024–25.

The BLS figures come from about 6,600 respondents to a supplement of the American Time Use Survey. Federal shutdowns interrupted collection in October and early November 2025, so the report combines a 12-month average for 2024 with an 11-month average for 2025. It is a portrait of that period, not a live count of jobs available today.

For a worker weighing an offer, the practical questions are separate: Can I choose when I start and stop? How often may I change it? How far ahead will I know my hours? Can I work from home, and on which days? A single “flexible” label answers none of those by itself.