How many people work from home in the US, and who are they?

About 13 in every 100 US workers usually work from home. The Census Bureau's American Community Survey commuting figures put the share at 13.3 per cent in 2024, down from 13.8 per cent in 2023 but still more than twice the 5.7 per cent of 2019. In 2023 that was more than 22 million people, against around 9 million in 2019.

That is the count of people for whom home is their usual place of work. Count everyone who works from home on some days, and the number is much larger.

How many people work from home at least some of the time?

About a third of employed people, on the days they work. In the Bureau of Labor Statistics' American Time Use Survey, 35 per cent of employed people did some or all of their work at home on the days they worked in 2025, and 70 per cent did some or all of it at their workplace. The two add up to more than 100 per cent because some people work in both places on the same day.

The two surveys ask different questions. The American Community Survey asks people how they usually got to work in the previous week, so someone who commutes on most days and works from home on the rest is counted as a commuter. The Census Bureau's definition of working at home is the narrower of the two. The time use survey records where people actually worked on the day they describe. Neither figure is wrong, but check which one you are using before it goes into a plan.

Who works from home?

The Census Bureau added home-based workers to its detailed commuting tables from 2023. Its first look at the new figures sums them up this way: remote workers "tend to be older, are more likely to be White and less likely to be in poverty". The figures behind that, all for 2023, each compare home-based workers with a particular group:

• Age. Their median age was about 43.5, against 41.7 for all workers.
• Race. Roughly two-thirds were White, compared with about half of workers who carpooled.
• Poverty. Around 3.6 per cent were below the poverty level, against 9.3 per cent of people who took public transport to work.
• Earnings. In five of the most populous metro areas, home-based workers had higher median earnings than all workers. In the New York and Chicago metros their median was over $80,000.

The time use survey adds education and sex. Among workers aged 25 and over, 51 per cent of employed people with a bachelor's degree or higher did some work at home on the days they worked in 2025, against 19 per cent of those with a high school diploma and no college. Employed women were more likely than men to work at home: 38 per cent against 31 per cent.

For another audience profile built the same way, from published government figures, see how many CEOs there are in the US and who they are.

Does it vary by place?

Yes, and the Census Bureau publishes the same breakdowns for states, metro areas and counties. Two examples from its own write-up, both for 2023: in North Carolina, the median earnings of home-based workers were $65,652, roughly twice the $30,664 of people who commuted by public transport. In Bronx County, New York, 39.3 per cent of home-based workers were in management, business, science and arts occupations, against 27.6 per cent of all workers.

What does this mean if you sell to people who work from home?

1. Say which group you mean. "Remote workers" can mean the 13 per cent or so who usually work from home, or the larger group who do it on at least some days. Part of the larger group also works at a workplace, so it can be reached both at home and on the way to work.
2. Expect a better-off buyer. Working from home is far more common among people with degrees, and home-based workers earned more than other workers in the places the Census Bureau compared. That bears on price, on message and on where you advertise.
3. Check the local picture. National figures may not describe your region. For a regional campaign, start from the state, metro or county tables rather than the national figure. A guide to using the American Community Survey for market research covers how marketing teams work with the survey.
4. Size it from the people, not the label. Count the workers who match your buyer rather than quoting the national share.

Where the numbers have limits

The figures come from two surveys and three years. The 2024 share and the 2023 characteristics are from the American Community Survey; the time use figures are for 2025. They also cover slightly different people: the American Community Survey counts workers aged 16 and over, the time use survey people aged 15 and over. Both are surveys of a sample, so every figure is an estimate with a margin of error, and small groups in small places carry wider ones. One-year American Community Survey figures are published only for areas of 65,000 people or more; for smaller counties, use the five-year estimates. The American Community Survey counts people by how they usually travelled to work, so it misses the home working of people who commute on most days. The time use survey counts anyone who did any work at home on a given day, so it includes people who spent only part of that day working at home.

Cambium AI builds synthetic populations from verified US public data, so a team can check the audience it is planning for before it commits a budget. For a campaign aimed at people who work from home, Cambium AI's marketing pilot is a way to test the audience and the creative before the spend is committed.

Data sources: U.S. Census Bureau, American Community Survey 1-year estimates, commuting (2024) and home-based worker characteristics (2023, published January 2025); U.S. Bureau of Labor Statistics, American Time Use Survey, 2025 results (June 2026).