Remote Work Statistics (2026)

Remote work has settled well above its pre-pandemic level. In September 2026, about 21.9% of US workers teleworked or worked at home for pay, roughly 34.7 million people, according to the BLS Current Population Survey (see the table below) (BLS). That is slightly below the 22.9% recorded in early 2024 but far above where things started.

A broader lens from the BLS American Time Use Survey shows 35% of employed people did some or all of their work at home on days they worked in 2025. The gap between 22% and 35% is not a contradiction: it reflects two different questions, which the next section unpacks.

There is no single 'remote work rate,' which is why headline figures range from 22% to 35% to over 50%. The BLS CPS telework question asks whether someone worked from home for pay in a reference week. The ATUS asks whether any work was done at home on days worked. Gallup restricts its split to remote-capable jobs only.

So the measure you cite depends on the question. Roughly 22% telework in a given week (CPS), about 35% do at least some work at home (ATUS), and among jobs that can be done remotely, only about 22% are fully on-site (Gallup). All three are true at once; they just count different things.

The arc is dramatic. Before 2020, only about 6.5% of private-sector workers regularly worked from home. The 2020 lockdowns pushed that above 60% of paid days for a time, then it retreated as offices reopened (see the chart below). By early 2023 the CPS telework rate was 19.6%, climbing to 22.9% in early 2024.

Since then it has plateaued. The rate has bounced in a narrow 17.9% to 23.8% band since late 2022 and read 21.9% in September 2026, and WFH Research finds the share of full paid days worked from home has held near a quarter since 2023. The post-pandemic retreat from remote work has, in the researchers' words, largely bottomed out.

Among jobs that can be done remotely, hybrid has won. Gallup's tracking of remote-capable US employees finds 52% hybrid, 26% fully remote, and 22% fully on-site as of May 2026 (see the chart below), a hybrid share that has held relatively steady since 2022 (Gallup). Expectations for the long term closely mirror today: 52% hybrid, 25% fully remote and 23% on-site.

Worker preferences run even more strongly toward flexibility: asked in November 2025, 61% of remote-capable employees said they prefer a hybrid schedule, 33% fully remote, and only 6% on-site. Hybrid lines up closely with preference (52% work it, 61% want it), while on-site is overrepresented (22% work on-site versus 6% who prefer it).

Remote work is heavily concentrated at the top of the education ladder. In September 2026, 40.6% of workers with an advanced degree teleworked, versus 8.3% of those with only a high-school diploma and 4.4% of workers without one (see the table below) (BLS). The ATUS shows the same split: roughly half of bachelor's-degree holders did some work at home, against 19% of high-school-only workers.

The reason is task type, not privilege alone. Jobs requiring a degree are more likely to be knowledge work that travels over a laptop, while roles in construction, food service, and transportation must be done on-site. Education is largely a proxy for whether a job is remote-capable in the first place.

Remote work skews toward women, prime-age workers, and certain groups. In September 2026, women teleworked at 23.9% versus 20.1% for men, a gap that has persisted since 2022 (see the table below). Asian workers had the highest rate at 29.4%, followed by White (22.5%), Black (15.0%), and Hispanic workers (12.8%).

Age matters sharply at the young end: only 5.7% of workers aged 16 to 24 teleworked in September 2026, versus 26.2% of workers aged 35 to 44. Younger workers are overrepresented in customer-facing service jobs, and early-career employees are also more likely to be asked on-site for training and mentorship. Married workers (27.2%) and parents of children under 18 (25.9%) also telework above the average.

Remote work is an occupation story before it is anything else. In September 2026, computer and mathematical roles led with a 63.7% telework rate, business and financial operations ran at 53.8%, and the finance and insurance (57.6%) and information (53.5%) industries were both above half, while management sat at 37.7% (see the chart and table below) (BLS). These are the classic laptop-and-login jobs.

At the other extreme, physical and customer-facing work barely moved: construction and extraction at 2.2%, transportation and material moving at 2.0%, and food preparation and serving at just 1.4%. The dispersion is enormous, which is why national averages hide the reality that remote work is nearly universal in some fields and essentially nonexistent in others.

For those who do work remotely, the intensity has eased. In early 2024, teleworkers worked an average of 27.2 hours a week from home, down 2.1 hours from a year earlier, and the share who worked entirely from home fell 6.1 points to 47.9%. The center of gravity is shifting from fully remote toward hybrid.

That matches the hybrid-days data: employers that require office attendance asked for an average of 3.8 days a week in 2025, up from 2.6 days in 2023, while WFH Research puts the economy-wide average near a quarter of all paid days done from home. In practice, a two-or-three-day-a-week split is now the norm.

The evidence on hybrid work is more settled than the debate suggests. Stanford economist Nicholas Bloom's experiment on more than 1,600 workers, published in Nature, found hybrid schedules had zero measurable effect on productivity or promotion while cutting resignations by 33% (see the takeaways above). Home days save workers about 70 minutes each in commute and prep time.

Fully remote is more contested, but self-reports are positive: Pew found 56% of remote workers say their productivity improved and 71% report better work-life balance. Bloom's read is that hybrid is close to a free lunch for employers, boosting retention and morale without denting output.

Workers treat remote flexibility as real compensation. Field experiments find the average worker will accept about an 8% pay cut for the option to work from home, and Bloom estimates the typical employee values remote days at roughly the equivalent of an 8% raise. For some tech workers, one NBER study put the figure as high as a 25% pay cut.

That valuation is why remote flexibility has become a recruiting and retention lever, not just a perk. It also explains the standoff behind return-to-office fights: when employers revoke remote days, workers experience it as a pay cut, and many respond by looking for a new job.

Large employers are tightening the rules even as workers resist. By mid-2025, 54% of Fortune 100 companies required five days a week in the office, up sharply from just 11% a year earlier, and 83% of CEOs told KPMG they expect a full return by 2027 (see the table below).

Enforcement is catching up to policy but unevenly: the share of firms actively enforcing attendance rose to 37% in 2025 from 17% in 2024, yet actual office attendance ticked up only 1-3%. Some executives have candidly described mandates as a way to trim headcount without formal layoffs.

The public sector saw one of the fastest reversals. Gallup data indicate hybrid work among federal employees collapsed from about 61% in late 2024 to roughly 28% by mid-2025, with fully on-site federal workers rising to around 46%, following a government-wide return-to-office directive.

That is a much sharper swing than the private sector, where hybrid barely moved over the same span. By September 2026 the BLS measured federal workers teleworking at 18.4%, below the 21.0% rate for private-industry workers (BLS). It is a reminder that top-down mandates can shift arrangements quickly when a single employer controls a large workforce.

Remote and hybrid work reshaped commercial real estate. The US office vacancy rate eased to 18.3% in the second quarter of 2026 from 18.6% in the first, still historically high, while the prime segment was tighter at 12.3% as tenants concentrated demand in the best buildings (see the table below) (CBRE). About 72% of office tenants have adopted hybrid or remote strategies, shrinking their footprints.

The market is adjusting supply rather than filling space: sublease inventory has fallen about 28% from its peak, only 2.2 million square feet of office space was completed in Q2 2026 (the lowest first-half level since CBRE began tracking in 1990), net absorption nearly doubled to 12.6 million square feet, and 2025 was the first year since 1988 that demolitions and conversions outpaced new office completions. The overhang is being worked off slowly, not reversed.

The remote-work shift is not a passing anomaly; it is a structural change with clear market winners and losers. Software, cloud, cybersecurity, and collaboration-tool companies benefit from a permanently larger distributed workforce, while traditional office landlords and REITs face a slow, grinding repricing of their core asset.

For a thematic investor, the durable read is that hybrid work has stabilized at roughly a quarter of all paid days and is not going back to the 2019 baseline. Building a thesis around that, on the technology that enables it and caution on the real estate it disrupts, is more useful than betting on a full return to the office that the data does not support.