Local AI: Governance Before Hardware Purchase
Benedikt Langer
10 min readFour developments over two weeks show that locally operated AI goes far beyond the tech stack. ...
Read Article
8 min read
Five years after the forced home office experiment, we now have enough data to settle the debate. A Stanford study published in Nature, involving 1,600 participants, shows: hybrid work has zero impact on productivity or career progression – but cuts turnover by 33 percent. At the same time, German data warns that overall productivity declines once remote work exceeds 60 percent of working time. Both camps are right – and both are wrong. The truth lies at 2 to 3 days per week.
TL;DR
- Zero productivity loss: The Stanford/Trip.com study (n = 1,612, published in Nature 2024) shows hybrid workers (2 days remote) are just as productive and receive promotions at the same rate as their office-based colleagues.
- 33 percent fewer resignations: Turnover dropped significantly under hybrid arrangements – especially among women, non-managers, and commuters with commutes exceeding one hour.
- 24.5 percent home office share in Germany: Stable at just under one-quarter of all workdays (ifo Institute, March 2025). This figure has barely shifted over the past 18 months.
- The “sweet spot” is at 40 percent remote work: Focused individual work is 20 percent more efficient at home. Yet overall productivity declines once remote work exceeds 60 percent of total working time (bidt).
- 20 percent of companies are rolling back: One in five German companies has abolished home office – often without a data-driven rationale (Bitkom, 2025).
What the Nature Study Shows – and Why It’s the Strongest Evidence Yet
In June 2024, the research team led by Stanford economist Nick Bloom published the most methodologically robust study on this topic to date in Nature – the world’s most prestigious scientific journal. The setup: 1,612 employees of Chinese travel conglomerate Trip.com were randomly assigned to two groups. One group worked hybrid (from home on Wednesdays and Fridays); the other worked five days per week in the office. Duration: two years.
What does hybrid work mean? Hybrid work describes a model where employees split their working time between the office and another location – typically home. The most common models prescribe 2 to 3 remote days per week. Unlike full remote work, hybrid preserves face-to-face contact; unlike full office work, it grants employees flexibility and reclaims commuting time.
The results: no difference in performance reviews. No difference in promotion rates. No difference in code quality among developers. What did change: turnover fell by 33 percent – most sharply among three groups: women, non-managers, and employees with commutes longer than one hour.
What makes this study especially valuable: it ran for two years – not six months. Common objections (“honeymoon effect,” temporary motivation boosts) lose weight at this timescale. And because it was randomized, it avoids self-selection bias – no apples-to-oranges comparisons.
German Data: Stable at 25 Percent – and That’s Fine
The ifo Institute reports a home office share of 24.5 percent of all workdays in Germany for February 2025. This figure has held steady between 24 and 25 percent for 18 months. The pandemic peak reached 50 percent; the pre-COVID baseline was under 5 percent. A new normal has settled – in clear excess of the old normal, yet well below pandemic highs.
Sources: Nature/Stanford (2024), ifo Institute (2025), bidt Themenmonitor
According to Bitkom, 30 percent of German companies offer home office – 10 percentage points more than in 2014. Yet 20 percent of firms that previously offered home office have since revoked it. RTO mandates from major corporations – including Amazon, Dell, and JPMorgan – dominate headlines. The question remains: Are these decisions grounded in data?
The Sweet Spot: Why “2 to 3 Days” Isn’t Just a Compromise Formula
Data from multiple sources converge on the same point. The Bavarian Research Institute for Digital Transformation (bidt) adds the nuance missing from black-and-white debates: focused individual work is roughly 20 percent more efficient at home than in an open-plan office. But once remote work exceeds 60 percent of total working time, overall productivity drops. Information flow and team cohesion suffer without regular in-person interaction.
Nick Bloom’s own synthesis, drawn from hundreds of studies, is unequivocal: hybrid work with 2 to 3 office days per week delivers a “triple win” – for productivity, employee satisfaction, and retention. Fully remote work, by contrast, carries measurable downsides: weaker innovation, slower knowledge transfer, and heightened risk of social isolation.
The implication: “Remote vs. office” is the wrong question. The right question is: Which 2 to 3 days – and for which tasks? The most successful models differentiate by work type – not by day of the week: focus work (analysis, programming, writing) goes home; collaboration (workshops, brainstorming, onboarding) happens in the office.
Why RTO Mandates Are Rising – Despite the Evidence
If the data is so clear, why are giants like Amazon, Dell, and JPMorgan calling staff back? Three explanations go beyond “bad management”:
Real estate logic: Companies that invested billions in office space can’t publicly admit 40 percent of it is redundant. Depreciation schedules run; leases remain active. In some cases, RTO is a balance-sheet decision – not a productivity one. Naturally, no one says this aloud.
The illusion of control: Visible employees feel controllable. Presence gets conflated with performance – a cognitive error managers openly acknowledge in surveys, yet still practice daily. Microsoft’s Work Trend Index reveals managers trust their remote teams less than their office-based ones – even though output metrics show no difference.
Hidden layoffs: Strict RTO mandates reliably trigger 10-30 percent voluntary departures. For companies seeking headcount reduction without severance payouts, this is a calculated outcome. Amazon has never confirmed this – but the numbers speak for themselves.
The consequence: RTO mandates lacking data backing are expensive experiments. Bloom himself estimates the average value of hybrid flexibility to employees at 8 percent of salary. That means abolishing hybrid work forces employers to either raise pay by 8 percent – or absorb 33 percent higher turnover. Most do neither – and then wonder why roles stay unfilled.
What’s Settled After 5 Years – and What Isn’t
Settled: Hybrid work (2-3 office days) has no negative impact on productivity or careers. It slashes turnover significantly (−33%). Focus work is more efficient at home (+20%). Germany’s home office share has stabilized at 25 percent – and will likely remain there.
Not settled: How fully remote work affects innovation and corporate culture over 5-10 years remains unknown. Most long-term studies have only been running for 2-3 years. Likewise, whether early-career professionals who’ve never set foot in an office achieve the same career trajectory as predecessors remains an open question.
The counterpoint: Some organizations genuinely need more presence than others. Startups in growth mode, teams requiring high coordination, or creative-sector firms may benefit from more office time. The data reflects averages – individual organizations may deviate. Crucially, such deviations must be justified – not simply follow the RTO trend.
Conclusion
Five years of data – published in Nature, gathered by Stanford, corroborated by the ifo Institute – confirm hybrid work works. Not as a stopgap, but as a proven, effective model. Its sweet spot sits at 2 to 3 office days per week. Demand more, and you lose talent. Offer less, and you risk productivity loss. The ideological “remote vs. office” debate is over. The data has spoken. Now calendars – and the leaders who fill them – must catch up.
Frequently Asked Questions
Is remote work more productive than office work?
Focused individual work is about 20 percent more efficient at home than in an open-plan office (bidt). Hybrid models (2-3 days remote) show zero productivity loss compared to full office work (Stanford/Nature, 2024). Once remote work exceeds 60 percent of total working time, overall productivity declines due to weakened team communication and knowledge transfer.
How many Germans work from home?
According to the ifo Institute, Germany’s home office share stands stably at 24.5 percent of all workdays (as of February 2025). Thirty percent of German companies offer home office. Pre-pandemic, the figure was under 5 percent; during the pandemic, it peaked at 50 percent. Today’s level represents the new normal – and changes little month to month. In the U.S., the figure is 25 percent – nearly identical. Eighty percent of U.S. companies now offer some form of remote work. Globally, around 100 million people work in hybrid models – many times the pre-COVID baseline. The infrastructure (tools, processes, culture) exists. Whether remote work can function is no longer in doubt. The open question is: how exactly?
What does the Stanford study actually say?
The study by Nick Bloom and colleagues (Nature, June 2024) tracked 1,612 Trip.com employees over two years in a randomized trial. Result: hybrid work (2 days remote) had zero effect on productivity, performance reviews, or promotions. Turnover dropped by 33 percent – especially among women, non-managers, and commuters.
Why are companies like Amazon recalling everyone anyway?
Three drivers: First, sunk real estate costs (billions invested in office space; leases still active). Second, preference for control (presence equated with performance – even when data contradicts it). Third, calculated workforce reduction (strict RTO mandates trigger 10-30 percent voluntary exits, avoiding severance costs).
What’s the optimal hybrid model?
Evidence consistently points to 2-3 office days per week, with the remainder remote. Top-performing models differentiate by task type – not weekday: focus work (analysis, coding, writing) at home; collaboration (workshops, brainstorming, onboarding) in the office. Companies implementing this rigorously report the highest employee satisfaction alongside stable productivity.
Editor’s Reading Recommendations
More from the MBF Media Network
Header Image Source: Pexels / (px:4050315)