And there’s a big mismatch between what employers are offering and what workers want.
By Wolf Richter for WOLF STREET.
Despite an endless series of increasing pressure tactics, reprimands, threats of termination, ultimatums, and actual terminations by companies to force the Return to the Office (RTO), and despite efforts by state and local governments to get workers back into the central business districts on a daily basis, and despite the RTO mandate from the federal government this January, progress of RTO has essentially stalled since early 2023. Working from Home (WFH) has become entrenched, hybrid models are now common and functional, and RTO languishes.
In 2019, the share of “full paid days worked” from home as a percent of total days worked was about 7%, after slowly rising from near-0% in the 1960s. During the pandemic in May 2020, the share of days worked from home spiked to 62%, in part because office workers switched to WFH, and in part because other employment, such as in accommodation and food services, collapsed, giving the working-from-home workers a larger share of total days worked. As service workers returned to their jobs, the share of WFH plunged and by early 2023 hit 28%, and got hung up.
But August 2025, full paid days worked from home was still 27.2%, same as in October 2024, and above May 2024, and nearly four times the share in 2019, according to data by Jose Maria Barrero, Nicholas Bloom, Shelby Buckman, and Steven J. Davis, published by WFH Research.
A full day is defined as working 6 or more hours per day. While the authors say that the dataset may over-sample people who are “more tech and internet savvy, especially among the least educated,” the trend still works out about the same.
But there is a big mismatch between what employers are offering and what workers want:
- 60% of workers worked zero days from home (first chart below, via WFH Research);
- But only 27.7% of workers wanted to work zero/near-zero days from home (second chart below, via WFH Research).
Of all full-time employees in the spring 2025 (essentially unchanged from last fall):
- Fully remote: 12.6%
- Hybrid: 26.4 %
- Fully on site: 61%.
These three industries lead in the percentage of full-time employees on full WFH:
- Finance & Insurance: 26% full WFH, 41% hybrid, 33% fully on site.
- Professional & business services: 22% full WFH, 36% hybrid, 42% fully on site.
- Information: 21% full WFH, 51% hybrid, 28% fully on site.
Office attendance also shows RTO has stalled.
Kastle’s weekly back-to-work barometer measures office occupancy by how many people enter an office building for which Kastle provides the electronic access system. It measures office occupancy by people entering these office buildings as a percentage of what it had been before Covid. If office occupancy returns to pre-Covid levels, the barometer would return to “100%.”
The barometer had plunged to 15% in March 2020 for the average of the 10 large metropolitan areas which it tracks. It then recovered fairly steadily through 2022 to roughly the 50% line. But since the start of 2023, it barely edged up further and has been zigzagging in the 50% to 55% range, except during non-holiday periods when it plunges.
In the chart, the top gridline (100%) denotes the pre-Covid level. Note how the 10-city weekly average (red line) has meandered in the 50% to 55% range since early 2023 (chart from Kastle):
The metros of Austin and Houston have been in the 60% range since early 2023, essentially unchanged.
The metros of Philadelphia, San Francisco, and San Jose have been in the 40-45% range since early 2023, also essentially unchanged.
At very high-end office buildings, representing only 2.3% of the buildings in Kastle’s Barometer, the occupancy rate has recently been roughly 20 percentage points higher than the 10-city average occupancy rate, except the weeks leading up to and including the Labor Day week – but even their occupancy rate of a little over 70% is still far below where it had been before covid.
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the mug to find out how:
Thanks wolf
Terrific analysis and well timed. We keep receiving investor offerings touting the “amazing returns” we could receive investing in WFH etc. It’s great to review unbiased data and opinion. Thanks!
We too continue to receive offers VG, but I just want to warn you NOT to DO any, repeat ANY vigilante efforts!!
Thanks for your comment, and please to continue to comment and support WR’s efforts,,,
IMHO THE best on now SO many themes.
During a town hall it was asked of leadership if we are changing from three days (T-Th) in the office and two at home (M&F). The panel all responded that they haven’t heard of any change or shift in that position. Good to hear and the audience appeared to like the continuity and balance.
As someone who works for a major corporate retailer, this has impacted me (4 days RTO, Fridays WFH), but there has been flexibility within reason ie kid is sick from school, dr appointment, something comes up, etc, etc, you can WFH more than just the one day a week, which actually was the way pre-pandemic anyway (that flexibility in a 5 day work week).
Not that I didn’t mind WFH so much, just not one of those bothered being asked to come back. There’s ups and downs as with anything else. IMO there is something to be said about being in a meeting with people in person/face to face, esp with those with the right # of stripes on their arms. Also, ‘watercooler’ convos, which I once thought silly, I will never under appreciate again after the pandemic days.
Having said that from a personal/anecdotal POV, I wonder what Wolf presents here is an indication on how employers might be willing continue to bite with this ‘mismatch’ as they better understand the use of AI in helping to alleviate this pressure.
An AI agent could well be indistinguishable from a certain segment of WFH employees in a few years.
At which point the ‘pressure’ on the employer can be one solved by simply letting the employee go.
I’m not sure it’s a good time to be playing brinkmanship games with your employer unless you’re near certain you’re an AI-proof employee.
Working from home is great as long as ther is productivity. It up to the middle managers to be leaders, not taskmasters
If you work from home, you are working with a computer and/or telephone. Data can be pulled from both to gauge individual production. I was the IT Director at my last place of employment. We measured staff’s productivity that way. We knew how many sales were closed, customer interactions, accounting entries, etc… Some folks excelled working from home, others languished. Those who languished were brought back in. If they complained we sent them back home but gave them daily quotas of work (always a higher number). Some made it, others recognized their shortcomings and came back in. We remodeled the offices and saved about 55% of our floor space. Our hybrid model was a Teams meeting where we pushed leadership topics and talked about bigger project issues. Today, I’m sure much of the staff doesn’t even live close to an office, so RTO is a non-starter.
Everyone worked from home for three years. If productivity had gone down, employers would have been screaming about it. Instead, there were crickets.
People are MORE productive working from home. Employers wanting people in the office has zero to do with productivity and everything to do with control.
You can’t paint with that broad a stroke. Some people are more productive WFH and some are less.
The problem is that executives and managers often paint it with a broad stroke and default to 100% in-office.
“You can’t paint with that broad a stroke.”
Nor did I.
“Some people are more productive WFH and some are less.”
Apples and oranges. You’re looking at trees and I’m looking at the forest.
Whether individual people are more productive at home or in workplaces is not relevant to my argument.
In the aggregate, companies experience higher productivity when their employees work from home. We just ran a multi-year experiment. Had overall productivity declined, employers would have been screaming about it and demanding people get back in the office, covid be damned.
You just want an environment where individual lazy people will be punished. That’s fine, but it also has nothing to do with overall productivity.
Eric Patton, I don’t “want” anything. It seems that your goal is WFH irrespective of productivity.
Working from the office is great as long as there is productivity.
Have you ever worked in an office before? People waste a ton of time.
I understand the slippery slope of losing jobs to overseas talent if you allow your company to go full WFH. I feel that this can remedied if companies are restricted to hiring only US residents.
I’m an exception with continuing 100% work from home but most agencies are 2 to 3 days a week. State mandated RTO 4 days a week but that was likely a move to negotiate 3% pay cuts across the board which we now have. Returning to the office was impractical as would have had to rent all the space they gave up plus of course all the expenses to get it setup plus equipment costs. After 5 years of telework going back into the office would be a shocker but given 50% of who I work with are all over the US so being on Teams in an office is much worse than in home, as just competing with your neighbor with volume.
Since Jan 2025 full time jobs are down 1.5 million. Full time gov
employees will cont to shrink. Mid management homes became
command centers with 3/4 monitors. They can be on conference calls
24 hr/day. Their co don’t pay them overtime and rent. Full paid workers are reserves workers, on dormant wages, before they get a call for WFH action.
“Since Jan 2025 full time jobs are down 1.5 million.”
LOL. Look at a 1-year chart of full-time employment. You will see that in January 2025, there was a 2.4 million UP-revision of full-time employment, part of which was then revised back down in the following months (if you had read my articles at the time, you’d know this). So you cherry-picked January because it suited your narrative. Now go to December 2024, and count from December, and you will see that since December, over those 8 months, full time jobs have soared by 970,000.
With how tight the labor market has been, remote and hybrid positions open the geographical range a job posting can reach just enough to actually fill a position. A lot of people are willing to have a 1.5 hour commute if it’s only for 3 or less days.
In the past few months, most of Canada’s major banks have announced 4 to 5 day per week return to office policies, as has the Province of Ontario public service, and major telecom provider Rogers. It may not be popular with workers but employers are forcing the return to office.
I recently learned that two of my coworkers I deal with almost on a daily basis, actually live on the other end of the state, almost 4 hours from where our home office is.
I honestly don’t know what to make out of this. We are a state agency and we’ve effectively been full WFH since covid, being called to the office or off-site events on a pure “as needed” basis (once every few months or so, enough to still remember what your boss looks like). I admit I sometimes miss water cooler chats and lunches with teammates, and our VPNs can be super slow at times, but I don’t realistically see this setup changing anytime soon. Besides, most of the counterparts we are working with are remote and often out-of-state, so there is nearly zero point in coming to the office to still spend the whole day in Teams calls.
It would be a crazy uphill battle if someone decides to return everyone to office, unless they come up with a valid business reason why it is needed, not just “because I said so”.
The company I work for only had WFH for a short period during covid, but they instituted one day a week WFH about a year ago. I think companies are using a limited version as a benefit to help keep wages down and improve retention. It’s an extra tool in the toolbox instead of an all or nothing type thing.
My wife’s company has a great work-from-home policy. Basically, full-time if you want it.
When she attends meetings, there are people literally sitting in cubes, attending these in-person meetings on Teams from their desks. It is so fucking stupid.
One reason why RFH has little affect on productivity:
“I am reminded of Bullshit Jobs: A Theory, by the anthropologist David Graeber, who argued that around a third of modern jobs are pointless, and simply make work for other people. These included “Taskmasters”: middle managers who create work that isn’t needed; and “goons” — lobbyists and marketers who try to sell things that no one needs or wants. Graeber’s thesis had a huge response — many wrote to admit that they themselves had a bullshit job, and were miserable.”
–Modern life is drowning in a sea of verbiage_Financial Times.html, December 27 2024
My insurance company USAA went to a WFH model. I had to file 2 claims over the past 2 years on my homeowners policy. The claims adjusters were inexperienced and working with laptops at home and talking to me on the telephone. Both claims were handled poorly with multiple mistakes and irregularities which had to be resolved with elevation to upper level dispute resolution personnel. If they were in an office with other claims people instead of WFH they could have consulted with people and resolved most of the issues at that level. The WFH model has become a sacred cow, and any opinion that it is not serving the customer well is met with resistance.
This is a weird attempt to explain what happened to you (sorry it didn’t go well, btw). It sounds like two things happened here. First, you were dealing with inexperienced employees who lacked training and support (a totally separate issue from wfh). Second, it sounds like the company has done a poor job setting up WFH in a way that enables employees to maximize their productivity.
Neither of those is an indictment against the WFH model as a whole, in my opinion. I work fully remote and while there are certainly different sets of challenges, they can all be remediated if you know what you’re doing, and if you’ve built an organizational culture that can actually support a higher level of worker autonomy. Workers also need the technical tools and information resources to help them do their job and keep them accountable.
What’s funny is that these same problems can exist in the office. I’m definitely of the opinion that most companies could be more generous with WFH if they decided to actually learn how to do it correctly. But most don’t want to be bothered.
” If they were in an office with other claims people instead of WFH they could have consulted with people and resolved most of the issues at that level.”
Maybe but also maybe not. You have no proof to back up this statement.