Boards still talk about office attendance as if it were a proxy for management. It is not. The most credible data suggest the era of big return-to-office pushes has peaked, employees are willing to vote with their feet, and the firms that force presence risk losing the experience they most need.
What the new numbers really say about presence
A broad survey of more than 1,000 U.S. executives finds that only 12 percent of leaders with hybrid or fully remote staff plan to tighten return‑to‑office rules in the next year (Bloom, Davis and Barrero, 2025). Even if those plans materialize, the total share of paid workdays from home in the U.S. would fall by less than half a percentage point, from 21.2 to 20.8 percent. This is a modest adjustment, not a wholesale reversal. On the other side of the table, the December 2024 wave of the Survey of Working Arrangements and Attitudes reported that only 44 percent of workers would comply with a fully in‑person RTO policy. The rest said they would quit or look for another job (Bloom, Davis and Barrero, 2025).
The message is not subtle. Most executives are not planning a march back to 2019, and a majority of workers will not accept a blanket mandate without considering exit. The tension is not about office desks. It is about who decides when presence matters and how that decision is justified.
When mandates backfire where it hurts
A recent working paper tracks what happened after high‑profile firms imposed stricter attendance policies. Using distributional synthetic controls on a dataset of 260 million resumes, the authors find that return‑to‑office mandates at Microsoft, Apple, and SpaceX were followed by a disproportionate exit of senior employees, with effects that scale with mandate rigidity (Ding and Ma, 2024). After Microsoft’s April 2022 mandate, average tenure in upper deciles fell by roughly two months, with a 6 percent shift in the distribution toward employees with 0 to 3 years at the firm. Apple’s one‑day‑a‑week policy showed a smaller seniority shift, Microsoft’s 50 percent presence policy showed a larger one, and SpaceX’s full‑time presence requirement showed the largest movement. The paper frames this senior talent loss as a threat to productivity and innovation. It is a working paper, not yet peer reviewed, but the design is careful and the dose‑response pattern is hard to dismiss.
This is the power problem in concrete terms. A blanket policy may assert authority, but it can hollow out the mentorship, institutional memory, and judgment that complex work depends on. The price is paid quietly, in product decisions that lose nuance and in slower problem solving when the people who have seen the movie before are no longer in the room.
Power travels through learning, not hallways
The office should serve performance, not the other way around. The best classic evidence we have on how teams actually improve points to a different lever. In a field study of 51 work teams, psychological safety, the shared belief that it is safe to take interpersonal risks, was linked to learning behaviors that in turn supported better performance, while team efficacy alone did not predict learning when safety was accounted for (Edmondson, 1999). The mechanism is mundane and powerful. People speak up about errors, ask for help, and try new approaches when they believe they will not be punished for doing so. Attendance, by itself, does not produce that climate.
The broader organizational literature comes to a similar conclusion about power. Self‑managing organizations, defined as firms that decentralize authority in a formal and systematic way and that strip out the manager‑subordinate relationship, show how control can be designed into the work rather than embedded in a layer of hierarchy (Lee and Edmondson, 2017). Hierarchy works well for reliable execution in stable contexts, but it tends to inhibit complex, cross‑functional problem solving and to reinforce status differences that stifle development. That is the crux of presence and power. If the aim is better judgment across boundaries, then mandating time at the same address is a coarse substitute for clarity on who decides what and how information flows.
None of this says the office is irrelevant. It says presence is an input that must be matched to the outcome. Use the building for activities that amplify learning and trust, such as onboarding, thorny cross‑team design, and post‑mortems where candor matters. Treat everything else as a question of measurable output.
Managers are exhausted. Attendance will not fix that
There is another signal worth heeding. Global employee engagement fell to 20 percent in 2025, the lowest level since 2020, with Gallup estimating a cost of about 10 trillion dollars in lost productivity, or 9 percent of global GDP (Gallup, 2025). Manager engagement has dropped sharply too, down nine percentage points since 2022, with the steepest fall between 2024 and 2025, from 27 to 22 percent. In top‑quartile organizations, by contrast, 79 percent of managers are engaged, a gap that Gallup frames as a design issue rather than a temporary mood (Gallup, 2025). This is survey evidence, not causal proof, but the contrast is striking. Firms that treat engagement as a system problem find ways to create it. Those that reach for attendance as a cure signal that the underlying measures of good work are fuzzy.
Taken together, the evidence points to a simple reading. Presence is not a reliable proxy for control, capability, or culture. Where authority is clear and learning is safe, hybrid attendance becomes a coordination choice. Where authority is muddled, mandating presence becomes a surrogate for management and a tax on senior talent.
The practical read
Start with roles, not rules. Specify which activities truly benefit from co‑location and let teams schedule those moments, then measure whether those sessions improve cycle time, quality, or rework. Protect experienced contributors’ autonomy, because the best current evidence links rigid mandates to senior attrition with the sharpest effects where rules are strictest (Ding and Ma, 2024). Run lightweight checks on psychological safety in the teams that convene most often, and make leaders explicitly accountable for the conditions that drive learning rather than for headcount in chairs (Edmondson, 1999). Where you do require presence, tie it to a clear objective and a post‑hoc review of impact.
Finally, treat RTO as a trust audit. The Stanford and Atlanta Fed survey work suggests few leaders plan to tighten further and many workers will not accept it if they do (Bloom, Davis and Barrero, 2025). That is not a stalemate. It is a prompt to redesign how power is exercised so that the office becomes a tool for judgment and connection, not a stage for supervision.
Sources
- Bloom, Davis and Barrero (2025). Stanford Report — Return to office? Not everybody is doing it. Surveys show only 12% of executives plan tighter RTO and most workers would not comply with fully in-person mandates, framing the presence-power tension. 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. Working paper using distributional synthetic controls finds stricter RTO preceded disproportionate senior attrition at Microsoft, Apple, and SpaceX, with effects scaling by mandate rigidity. https://arxiv.org/pdf/2405.04352
- Edmondson (1999). Psychological Safety and Learning Behavior in Work Teams. Field study linking psychological safety to learning behaviors and performance, showing that presence alone is not the mechanism for improvement. https://journals.sagepub.com/doi/10.2307/2666999
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Theory synthesis arguing hierarchy suits stable execution but hinders complex problem solving, informing how power should be designed beyond attendance. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Gallup (2025). State of the Global Workplace 2025. Global survey shows engagement at 20% with large productivity costs and a steep drop in manager engagement, emphasizing that attendance does not substitute for sound organizational design. https://www.gallup.com/workplace/659610/state-global-workplace.aspx