The argument for hauling people back to offices is usually framed as a productivity fix. The numbers tell a different story. Only a small minority of executives in the United States plan new mandates, workers would not uniformly comply, and the firms that force the issue risk a loss of experience at the top. Presence still signals power. That is the problem to solve.
What the RTO push really looks like
In February 2025, a Stanford and Atlanta Fed survey of more than 1,000 executives found that only 12 percent of leaders with hybrid or fully remote staffs intended to impose a return‑to‑office mandate in the next 12 months (Bloom, Davis and Barrero, 2025). Even if all those plans were executed, the share of paid days worked from home in the United States would fall by less than half a percentage point, from 21.2 to 20.8 percent. Workers are clear about their own red lines. In the December 2024 edition of the Survey of Working Arrangements and Attitudes, only 44 percent of American workers said they would comply with a fully in‑person policy, with the rest saying they would quit or look for another job (Bloom, Davis and Barrero, 2025). These are not the responses of a labor force that sees presence as the route to better work. They read as a bargaining map.
The gap between executive intent and worker response explains why the return wave has been patchy rather than tidal. It also reframes the stakes. Mandates are not just a scheduling decision. They are a visible test of authority.
Who leaves when presence is compulsory
A working paper by economists Zhendong Ding and Jiajia Ma analyzed 260 million resumes to study three high‑profile mandates at Microsoft, Apple, and SpaceX (Ding and Ma, 2024). Using distributional synthetic controls, the authors attribute a disproportionate exit of senior employees to the mandates. After Microsoft announced a 50 percent in‑office policy in April 2022, average tenure in the upper deciles fell by about two months, and the share of employees with 0 to 3 years at the firm rose by 6 percent. The effect scales with the rigidity of the policy. Apple’s one‑day‑a‑week requirement showed a seniority shift below 4 percent. Microsoft’s stricter policy exceeded 4 percent. SpaceX’s full‑time presence requirement was stronger still. The paper has not yet been peer reviewed, so its estimates deserve continued scrutiny, but the pattern is consistent and economically meaningful.
Losing veterans is not a cosmetic problem. Senior people carry context, informal networks, and judgment that take years to rebuild. If mandates push out the most experienced employees, the near‑term optics of a busier office mask a longer‑term risk to productivity and innovation. That cost is easy to underestimate and hard to reverse.
Presence does not fix a design problem
The deeper issue is not where people sit. It is how authority flows. Michael Lee and Amy Edmondson’s synthesis of self‑managing organizations defines them as firms that decentralize authority radically and systematically, eliminating the manager‑subordinate relationship as the default structure (Lee and Edmondson, 2017). Their review makes a simple point that cuts through much of the RTO noise. Classic managerial hierarchy works when the task is predictable and the environment stable. The same hierarchy inhibits complex, cross‑functional problem solving, tends toward rigidity, and reinforces status differences that stunt development.
That diagnosis fits the moment. Many firms face problems that do not respect org charts. Treating co‑location as a cure misses the target. If the work requires lateral coordination and rapid learning, doubling down on presence to reassert vertical control addresses the wrong bottleneck. Engagement data add a sharper edge. Gallup’s 2025 report shows global employee engagement at 20 percent, with disengagement costing the world economy about 10 trillion dollars, or 9 percent of global GDP (Gallup, 2025). Managers are struggling too. Their engagement has fallen by nine percentage points since 2022, including a drop from 27 to 22 percent between 2024 and 2025. Yet in best‑practice organizations, 79 percent of managers are engaged, nearly four times the global average. That spread is the point. Engagement looks like a design choice, not a byproduct of location (Gallup, 2025).
If engagement hinges on design, then office attendance cannot fix a weak operating model. It can, however, amplify it. A brittle hierarchy feels more brittle when enforced face to face.
Use presence to build safety, not surveillance
There is a better way to think about offices. Amy Edmondson’s classic field study introduced team psychological safety as a shared belief that a team is safe for interpersonal risk taking (Edmondson, 1999). In 51 work teams at a single manufacturing firm, psychological safety was associated with learning behaviors such as seeking feedback and discussing errors, and those learning behaviors mediated the link to performance. The study was cross‑sectional and confined to one company, so it does not claim universal causation. It does, however, establish the mechanism leaders should aim to strengthen when they bring people together: learning, not monitoring.
Practically, that means treating presence as a scarce resource to improve work quality. Use co‑located time for interdependent problem solving, design reviews, and decisions that benefit from rapid iteration. Make attendance purposeful and time‑boxed. Where teams can set their own cadence, let them, then audit outcomes rather than hours. Track seniority mix and regretted attrition before and after any policy shift, including by role and commute distance, to detect Ding and Ma‑style losses early (Ding and Ma, 2024). Invest in manager capability where it matters, running meetings that surface dissent, documenting decisions, and closing the loop on issues raised. Those moves aim at the learning channel Edmondson identified, not at performative presence (Edmondson, 1999).
The policy horizon is clearer than the headlines suggest. Most executives are not marching toward stricter mandates, and even a full follow‑through would trim remote work at the margins, not erase it (Bloom, Davis and Barrero, 2025). The risk that is easy to miss sits elsewhere. Mandates that reassert control can repel the very people whose presence matters most. Leaders who treat the office as infrastructure for coordination and learning, and who redesign authority to match the work, will end up with quieter badge data and stronger results.
Sources
- Bloom, Davis and Barrero (2025). Return to office? Not everybody is doing it, Stanford Report. Finds only 12% of U.S. executives planned new RTO mandates and shows most workers would not comply with full in‑person policies, framing the real scale and resistance to mandates. https://news.stanford.edu/stories/2025/03/return-to-office-not-everybody-is-doing-it
- Ding and Ma (2024). Return-to-Office Mandates and Brain Drain, arXiv. A working paper using distributional synthetic controls on 260M resumes finds RTO mandates at Microsoft, Apple, and SpaceX disproportionately reduced senior tenure, with effects scaling by mandate rigidity. https://arxiv.org/pdf/2405.04352
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing, Research in Organizational Behavior. Defines self-managing organizations and synthesizes evidence that traditional hierarchy fits predictable tasks but inhibits complex cross-boundary problem solving and reinforces status differences. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Edmondson (1999). Psychological Safety and Learning Behavior in Work Teams, Administrative Science Quarterly. In a field study of 51 teams, team psychological safety is linked to learning behaviors that mediate performance, clarifying the mechanism leaders should target when convening teams. https://journals.sagepub.com/doi/10.2307/2666999
- Gallup (2025). State of the Global Workplace 2025. Reports global engagement at 20% with an estimated $10T productivity cost and a 9‑point drop in manager engagement since 2022, while best‑practice organizations show 79% manager engagement, underscoring that engagement reflects organizational design. https://www.gallup.com/workplace/659610/state-global-workplace.aspx