Remote, Hybrid or Back to the Office? What the Data Actually Says

On July 6, 2026, most of Canada’s core federal public servants were required to be at their desks four days a week, up from three. Executives had already gone to five days on May 4. The Treasury Board had announced the change in February, and by the time it took effect, several departments were quietly admitting they did not have enough desks. Immigration, Refugees and Citizenship Canada told most of its staff to stay at three days until it found space, according to The Canadian Press, and Global Affairs Canada phased its requirement in through mid-September.

That small farce captures where the remote work debate has landed in Canada. Employers in Ottawa, on Bay Street and at Queen’s Park have mostly decided the question. The evidence, though, has not moved in the same direction as the mandates. The best research we have says something more specific and, frankly, more useful than “the office is back” or “remote is the future.”

So let’s set the memos aside for a moment and look at what the studies, the statistics and the job postings actually show.

The study everyone cites, and what it really found

If you’ve read anything about hybrid work in the last two years, you’ve probably seen a reference to the Trip.com experiment. It’s worth knowing the details, because they matter more than the headline.

Stanford economist Nicholas Bloom and two co-authors, Ruobing Han and James Liang, ran a randomized controlled trial on 1,612 university-educated employees in engineering, marketing, accounting and finance at the travel company’s Shanghai offices. Starting in 2021, employees with odd-numbered birthdays could work from home on Wednesdays and Fridays; everyone else stayed in the office five days a week. The results were published in Nature in June 2024.

The headline numbers:

• Quits fell by a third. Attrition was 4.8% in the hybrid group versus 7.2% in the office group. The drop was largest for women, non-managers and people with long commutes.
• Performance didn’t budge. Over up to two years of reviews, the researchers found no difference in performance ratings or promotion rates. Software engineers wrote about as much code either way.
• Managers changed their minds. Before the trial, managers predicted hybrid work would cut productivity by 2.6% on average. Afterward, their estimate was a 1% gain.

Stanford’s own write-up called the result a “win-win-win” for the company, its workers and shareholders, since lower turnover saves real money on hiring and training. Two caveats are worth keeping in mind. The study covered university-educated office staff at one firm, and it tested two home days a week, not three or four. It says very little about fully remote work.

Fully remote is a different animal

This is the distinction that gets lost in most arguments. The Nature paper itself notes that earlier research on fully remote arrangements found mostly negative effects on productivity. In a 2023 review for the Journal of Economic Perspectives, Bloom, Jose Maria Barrero and Steven Davis put the productivity penalty of fully remote work at roughly 10%, as summarized by the American Enterprise Institute, while finding no measurable hit from hybrid setups.

The reasons are not mysterious. Fully remote work cuts down on the unplanned conversations where problems get solved quickly. It’s harder to learn by watching. One study cited in that review found junior engineers got about half as many code comments when their teams were fully remote rather than sitting together. A field experiment with data-entry workers in India found those working from home were 18% less productive, as Boston College’s Center for Retirement Research reported.

None of this means remote work is a failure. Fully remote firms save heavily on rent and can hire from anywhere, and for some jobs that trade is well worth a modest productivity dip. But it does mean that “remote” and “hybrid” should not be lumped together. The evidence for two or three home days a week is strong. The evidence for zero office days is mixed at best.

What Canadians are actually doing

Statistics Canada tracks where people work through the Labour Force Survey, and the trend line has been heading one way for four years. In May 2025, 17.4% of employed Canadians mostly worked from home, down from 18.7% a year earlier. That was the fourth straight annual decline since the pandemic low point for commuting in May 2021, when only 75.7% of workers commuted.

A few details from that release stand out:

• The share of workers who split their time between home and a workplace was about 10%. Among them, 44.7% spent at least half their hours outside the home, up from 43.0%.
• Ottawa–Gatineau saw the biggest jump among large cities. Its commuter share rose 9.5 percentage points to 76.0%, essentially level with Toronto at 76.1%. That’s the federal three-day rule from September 2024 showing up in the data.
• Toronto had the longest average commute among major cities, at 34.9 minutes, according to Benefits and Pensions Monitor’s coverage. Transit riders averaged 44 minutes.

The job market is moving the same way. Robert Half’s analysis of more than 48,000 Canadian postings in the second quarter of 2026 found 84% were fully on-site, 13% hybrid and just 3% fully remote. Technology and marketing roles were the most flexible, at about 20% hybrid or remote combined.

For comparison, the American numbers have flattened rather than fallen. Bloom’s WFH Research project, which surveys U.S. workers monthly, found that about 26% of paid workdays in August 2026 were worked from home, a level that has been fairly stable since 2023. Among full-time U.S. employees, 12% were fully remote, 26% hybrid and 62% fully on-site.

The mandates: who’s asking for what

Canada’s biggest employers have converged on a fairly aggressive standard. Here is roughly where things stand as of fall 2026:

The federal government’s common hybrid work model page confirms the four- and five-day rules. Its stated rationale, repeated in a Treasury Board question period note, is that working together on-site is essential “during this pivotal moment.” Federal unions have filed unfair labour practice complaints and policy grievances. The Professional Institute of the Public Service of Canada called the February announcement a matter of “optics,” CP24 reported.

Ontario went further. Premier Doug Ford announced in August 2025 that provincial public servants would move to four days in October and five days in January 2026, citing mentoring and downtown foot traffic.

The banks moved together in 2025. RBC, TD, Scotiabank and BMO all set four-day expectations starting in the fall, CTV News reported. RBC had previously asked staff to be in the office “the majority of the time,” according to an employment law summary, with the bank framing in-person connection as core to its culture.

Do mandates match the evidence?

Here’s where we’ll offer an opinion. The research gives decent support for asking people to come in two or three days a week, especially early-career staff who learn by proximity. It gives much less support for four or five days, at least for knowledge work that already happens on laptops. The clearest measurable cost of strict mandates is attrition, and the Trip.com study suggests it falls hardest on the people employers often say they want to keep: women, mid-level staff and those with long commutes.

Leaders may have other reasons, of course. Downtown cores, commercial real estate exposure and simple management preference are real factors. They just aren’t productivity findings, and it would be more honest to say so.

The tools that actually matter

Every hybrid debate eventually turns into a software shopping list. In our experience, the specific apps matter less than how teams use them. Microsoft’s own data makes the point uncomfortably well. Its June 2025 Work Trend Index report found that the average worker is interrupted roughly every two minutes during core hours by a meeting, email or chat, about 275 times a day. Meetings after 8 p.m. were up 16% year over year.

A few practices tend to separate hybrid teams that work from those that just feel busy:

1. Anchor days. If people come in, they should come in on the same days. An office full of people on video calls with colleagues at home is the worst of both arrangements.
2. Write things down. Decisions recorded in a shared document or ticket survive the meeting. This matters more the more your team is spread out.
3. Protect focus time. Calendar blocks, quiet hours for chat and fewer recurring meetings do more than any new collaboration tool.
4. Design for the newest hire. Mentoring is where fully remote work struggles most, so make sure juniors get deliberate time with senior people, in person where possible.
5. Book desks honestly. If attendance is mandatory, the space has to exist. The federal rollout showed what happens when it doesn’t.

Measurement tools deserve a caution. Health Canada, for instance, uses a Workplace Presence Management Tool where managers record employees’ daily work locations. Tracking attendance is easy. Tracking output is hard, and only the second one tells you whether your policy is working.

What this means if you run a team

If you’re a manager or founder in Canada, the honest summary of the evidence looks like this. Hybrid work with a couple of home days does not hurt performance and measurably helps retention. Fully remote work can work but carries real costs in mentoring and speed, which you need to deliberately offset. Four- and five-day mandates are mostly a cultural and political choice, and the people most likely to leave over them are often the ones you can least afford to lose.

For smaller companies, there’s an opportunity here. With 84% of Canadian postings now fully on-site, a well-run two- or three-day hybrid policy is a genuine hiring advantage. If you’re changing your policy, it’s worth talking to an employment lawyer first, since existing contracts and past practice can affect what you’re allowed to require.

The office isn’t dead and it isn’t back. The data says the best answer for most knowledge workers is somewhere in the middle, and that the middle is cheaper to run than the alternatives once you count the people who walk out the door.

Sources and further reading

• Bloom, Han and Liang, “Hybrid working from home improves retention without damaging performance,” Nature (2024)
• Stanford SIEPR: Hybrid work is a “win-win-win” for companies and workers
• AEI: What we’re learning about work-from-home and productivity growth
• Center for Retirement Research at Boston College: Hybrid work is having more success than fully remote
• Statistics Canada, The Daily: Number of Canadian commuters increases for fourth straight year in 2025
• Benefits and Pensions Monitor: The end of work from home may be closer than expected
• WFH Research: Monthly updates (September 2026)
• Robert Half: Canadian remote work statistics and trends
• Government of Canada: Common hybrid work model
• Treasury Board Secretariat: Question period note on hybrid work in the public service
• CP24: Federal government to order public servants back to office four days a week
• CP24: Office space scarce as federal public servants return four days a week
• North Shore News (Canadian Press): Ontario public servants back to office five days a week
• CTV News: BMO employees to start working four days in office
• NewsChannel 9: Amazon tells employees to return to office five days a week
• Microsoft Work Trend Index: Breaking down the infinite workday