Strip the collaboration language away from most return-to-office mandates and a plainer motive shows through. Remote work did not just move desks. It dissolved a form of control that managers had relied on without ever naming it: the ability to supervise by looking. A mandate is the attempt to get that control back. Once you see RTO as a fight over supervision rather than productivity, the pattern of who pushes it and how hard makes far more sense.
What remote work took from the manager
For most of the industrial era, management meant oversight you could perform with your eyes. You watched who arrived, who stayed, who looked busy, and you inferred effort from all of it. Distributed work broke that instrument. The manager could no longer see the room, so the old proxy for control stopped working. What remained was the harder discipline of judging people by their output, which requires actually knowing what good output is.
Lee and Edmondson (2017) describe the deeper structure at play. The defining feature of conventional hierarchy, in their synthesis, is the manager-subordinate relationship of oversight, and hierarchy is suited to the reliable execution of known tasks while tending toward rigidity. Remote work did not abolish the hierarchy, but it removed the visual supervision that made the manager-subordinate relationship feel real day to day. A return-to-office mandate restores the visibility. It puts the subordinate back within sight of the manager, which is the relationship the hierarchy is built around. That is the control being reclaimed, whether or not anyone says so.
The tell is in the evidence gap
If mandates were really about productivity, you would expect them to rest on productivity evidence. They rarely do. The strongest causal work on the consequences of mandates points the other way. Ding and Ma (2024), using distributional synthetic controls across 260 million résumés, found that mandates at Microsoft, SpaceX and Apple caused disproportionate senior departures and, in the authors' framing, threatened the very productivity and innovation the mandates were meant to protect. It is a working paper, not peer-reviewed, but its design is the best available, and it finds the policy working against its stated goal. When a policy persists despite evidence it undermines its own justification, the justification is probably not the real driver. Control is.
The survey record reinforces the reading. The Stanford and Atlanta Fed research program found that only 12 percent of executives with hybrid or remote staff planned a mandate, and that only 44 percent of workers would comply with a fully in-person policy (Bloom, Davis and Barrero, 2025). A minority of leaders reaching for a lever that a majority of workers would resist is not the profile of a productivity intervention. It is the profile of a contest over authority, pursued by the leaders for whom restoring visible control matters most.
Control that costs more than it delivers
The irony is that supervision by presence buys a weaker form of control than the one it replaced remote work's discipline with. Gallup (2025) reports global engagement at 20 percent, the lowest since 2020, and finds manager engagement itself falling sharply, down 9 points since 2022 to 22 percent by 2025. The managers doing the supervising are themselves disengaging. Reasserting the visual hierarchy has not produced a committed workforce or even committed overseers. It has produced a filled room presided over by people who are also checked out.
That is the trap of control as a goal. It is satisfying to the controller and hollow in effect. A manager can see the team again and still have no idea whether the work is good, because seeing was never the same as knowing. The mandate restores the feeling of oversight while leaving the substance of accountability exactly where it was, undefined and unmeasured.
There is a further cost that the control frame usually ignores. The oversight relationship the mandate reasserts is the very thing that suppresses candor. When people feel watched rather than trusted, they manage the impression rather than surface the problem. Edmondson's broader tradition of research treats that safety to speak as the engine of team learning, and a supervision-first posture runs directly against it. Control bought by visibility is therefore self-undermining: it purchases the appearance of command at the expense of the honesty a manager needs to command anything well.
The practical read
The way past the fight is to name the thing being fought over. If a leader wants control, the durable version of it is not visual. It comes from defining good work, measuring it, and holding people to the result regardless of where they sit. That kind of control survives distance, because it never depended on the eye in the first place.
The presence-based version is a regression to a supervision that knowledge work outgrew. It reclaims the appearance of control and forfeits the experienced people, as Ding and Ma document, who are the least willing to be watched and the most able to leave. Leaders who insist the mandate is about collaboration should test the claim honestly: would they accept the same output delivered out of sight? If the answer is no, the issue was never the work. It was the watching. Control was always the point, and it is the most expensive point a company can insist on.
Sources
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. The defining feature of hierarchy is the manager-subordinate oversight relationship, the visual supervision a mandate reclaims. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Ding and Ma (2024). Return-to-Office Mandates and Brain Drain. Mandates caused senior departures that threaten the productivity they claimed to protect, evidence the stated goal is not the real driver. https://arxiv.org/pdf/2405.04352
- Bloom, Davis and Barrero (2025). Return to office? Not everybody is doing it (SWAA/NBER). Only 12% of executives planned a mandate and only 44% of workers would comply, the profile of a fight over authority rather than productivity. https://news.stanford.edu/stories/2025/03/return-to-office-not-everybody-is-doing-it
- Gallup (2025). State of the Global Workplace 2025. Manager engagement fell 9 points to 22% by 2025, showing reclaimed visual control has not produced committed oversight. https://www.gallup.com/workplace/659610/state-global-workplace.aspx