Is Your Return-to-Office Mandate About Control or the Business?

When researchers went looking for the productivity payoff behind return-to-office mandates, they could not find it. A University of Pittsburgh study of S&P 500 firms found mandates did not improve financial performance or firm value, and did lower employee satisfaction. What fit the data instead was control: the authors concluded the pattern was consistent with managers “power grabbing” and blaming employees for poor performance, and that mandates were more common at firms with powerful CEOs. So if a return-to-office push feels less like a business decision and more like a need to see people at their desks, the research suggests your instinct is right. For a startup, that instinct is expensive.

This piece is about that expense. Not the morale cost you already suspect, but the hiring cost: the senior engineer who took the remote offer instead, the three weeks a role sat open, the applicants who never applied. If you are a founder deciding your own work-location policy, the useful question is not “is the office good?” It is “does this policy serve the company, or just my discomfort with an empty room?”

Are return-to-office mandates about control or productivity?

The research points to control. When economists at the University of Pittsburgh (Ma and Zhang) studied return-to-office mandates across S&P 500 firms, they found the mandates did not boost profits or firm value, and they hurt employee satisfaction. The pattern that best fit the data was managers reasserting control, not a measured gain in output. Mandates showed up more often at firms led by powerful CEOs, which is what you would expect if the driver were power rather than performance.

That is the empirical floor for the whole “it’s about control” argument, and it is worth being precise about what it does and does not say. It does not say every founder who likes an in-person team is insecure. It says that when you look at mandates in aggregate and search for the productivity justification, it is not in the numbers. What is left, once you subtract the business case that did not materialize, is the desire to watch the room.

What the research found, and what it didn’t

Here it helps to separate two things that popular coverage tends to blur.

The hard finding is the one above: mandates did not improve performance or value, they reduced satisfaction, and the behavior is consistent with control. That is measured, in real company data.

The softer layer is the personality language, the “narcissism,” “insecurity,” and “ego” framing you see in op-eds and LinkedIn threads. That part is interpretation, not a measured statistic. No credible peer-reviewed study has established a number linking a manager’s narcissism score to the odds they mandate RTO, and you should be suspicious of anyone who quotes one. What leadership psychologists actually do is layer a reading on top of the control data: the discomfort of not being able to see the team, the pull to reward whoever is visible, the unease of an empty office. Those are plausible readings of what remains once productivity is off the table. They are not a lab result.

Kept honest, the two halves reinforce each other rather than inflate a claim. The data says it is not productivity. Psychology offers a name for what remains. You do not need the word “narcissism” for the argument to land, and attaching a fake percentage to it would only weaken a case the real numbers already make.

The psychology of the empty room

Strip away the productivity story and a specific set of behaviors comes into focus. Most of them have names.

Proximity bias is the best-documented one. Managers tend to rate, remember, and promote the people they physically see over the people who quietly ship. It is not malice; it is a cognitive shortcut. But it means an office mandate is not neutral. It rewards visibility as a proxy for value, which is exactly the substitution a team serious about results should avoid.

The visibility reflex is the founder-level version. A room full of people at desks looks like momentum. An empty office at 2pm feels like the company is slipping, even when the work is shipping on time. That feeling is real. It is also not data. A founder who mandates attendance to soothe it is treating their own anxiety as a management metric.

None of this makes you a bad leader. It makes you a human running a company under uncertainty, reaching for the thing that feels like control. The problem is only that the thing that feels like control, at a startup, quietly taxes the one function you cannot afford to tax: hiring.

Why this is a startup problem, not a big-company one

The mandates in the headlines come from companies that can absorb the damage. A Fortune 500 can lose a slice of its workforce to an RTO policy and keep running; it has depth on every team and a brand that refills the funnel. A twelve-person startup has neither.

At seed or Series A, every engineer is load-bearing. One senior departure is not a headcount line; it is architectural judgment, code-review standards, mentorship for the juniors, and the undocumented context about why the system is built the way it is, all walking out at once. You cannot rehire that quickly, and the time it takes lands directly on your runway.

This is the trap in copying a big-company mandate: its costs scale down worse than they scale up. The headlines normalize a policy whose brain-drain math is survivable at fifty thousand people and potentially fatal at fifteen. Importing it means importing attrition onto a team with no absorption capacity, to buy a feeling of control you did not need in the first place.

The hiring math of an ego-driven mandate

Here is where the “expensive feeling” stops being a metaphor. The same Pittsburgh research group quantified what mandates cost on the talent side, and the numbers are blunt.

A separate Pittsburgh study (Ding, Jin, Ma, Xing and Yang) tracked more than 3 million workers’ employment histories at S&P 500 tech and finance firms. After RTO mandates, abnormal turnover rose roughly 13 to 14 percent, and the departures were not random. They concentrated among senior, high-skilled, and female employees, with women’s turnover increase nearly three times that of men’s. Mandates do not shed your weakest people. They shed the ones with the most options and the most institutional knowledge.

Then the funnel slows. In the same data, open vacancies took 23 percent longer to fill, from about 51 days to 63, and hire rates fell 17 percent after adjusting for market-wide trends. So the mandate empties senior seats faster and makes them slower to refill. For a startup counting weeks of runway per hire, that is a direct hit.

Gartner’s research points the same way from a different angle. In a 2024 survey of more than 3,000 managers, strict RTO cut intent to stay by up to 10 percent overall, but by 16 percent among high performers, and pushed “quiet quitting” up as much as 19 percent. A mandate repels your best people first.

And the funnel shrinks before anyone quits. LinkedIn’s own data shows remote and hybrid roles make up only about 20 percent of postings but attract roughly 60 percent of all applications. An onsite-only requisition fishes in a structurally smaller pond and signals low trust to exactly the senior candidates who have options elsewhere. You are advertising the single attribute the majority of applications are flowing away from.

Put it together and an ego-driven mandate is a triple tax: a smaller applicant pool at the top, your best and most senior people leaving out the back, and every remaining role slower and harder to fill.

“But I believe in the office” — the supported middle

None of this means the office is worthless or that fully remote is the only defensible answer. The honest reading of the evidence is narrower and more useful: a mandate driven by your need to see people does not buy the productivity you think, and it does buy attrition. Deliberate in-person time is a different thing.

The cleanest evidence here is a randomized controlled trial, not a survey. Stanford economist Nicholas Bloom and colleagues ran a hybrid-work experiment at Trip.com with more than 1,600 professionals, published in Nature in 2024. Employees assigned two work-from-home days a week resigned 33 percent less often, with no measurable hit to productivity or promotions. That is the middle the data supports: structured hybrid captures most of the collaboration upside while cutting the quits, without the brain-drain bill a rigid mandate runs up.

So the strawman to retire is “it’s control, so go fully remote to make a point.” The actual lesson is quieter. Decide in-person time on the merits of the work, not the ache of the empty room, and you can have the collaboration without the exodus.

The founder’s self-audit

You can settle the “control or business” question for your own company with one honest gut-check and one metric.

The gut-check is the mirror the research holds up to Fortune 500 CEOs, pointed at yourself. For this role, ask: what work genuinely needs a shared room, and what am I requiring only because an empty office makes me uneasy? If the honest answer is the second one, that is the ego tax, and you now know its price from the section above.

The metric turns the whole thing from a vibe into a hypothesis you can test. Decide each role’s arrangement on the merits, sized to whether you can absorb senior attrition (at seed, you cannot). State it plainly in the posting so candidates self-select before your pipeline fills with mismatches. Then watch the funnel. If an onsite-only role converts thin from application to offer, or offers keep dying late, your policy is the suspect, not your sourcing. “It’s about control” stops being an accusation you defend on vibes and becomes a claim you can falsify with your own hiring data.

How to run this in Kit

Kit is where that self-audit becomes something you run rather than debate, because the policy and its funnel effect live in the same place.

The most important piece is the apply-to-offer conversion rate Kit computes per posting on your dashboard. That is the falsification tool. If your onsite-only engineering role converts worse than your remote-friendly roles, or offers die at the last stage, you are looking at the policy tax in your own numbers rather than arguing about it. You replace “I feel people should be here” with a measured signal.

Upstream of that, Kit makes the arrangement visible before anyone applies. A posting carries a remote flag and location, and Kit renders a remote badge on the public job page, so candidates self-select on the real deal instead of discovering it at the offer stage. The same trust logic extends to pay: employment type and salary-range fields sit right alongside location, so you disclose the whole arrangement, not just where the desk is. Signaling openly beats signaling control, and it does it before a single application lands. (If pay transparency is your next question, see our piece on stating salary ranges up front.)

For the senior engineers you are specifically trying not to hand to flexible competitors, Kit’s candidate portal is magic-link and passwordless, so a self-selected candidate moves through your stages without the login and scheduling friction that quietly costs you the people with the most options. And posting and process templates let your team encode “state the arrangement, state the pay, on every req” once, so no founder’s bad-room-day ships a control-signaling mandate that dies at offer.

For the broader retention and disclosure math behind all of this, see our companion piece on return-to-office mandates and engineering retention.

The research keeps landing on the same conclusion: return-to-office mandates are usually about a leader’s need for control and visibility, not the business. That is an expensive feeling to indulge anywhere, and at a startup it is unaffordable, because it shrinks your applicant pool, repels your best senior candidates, and slows every hire at the exact moment one bad quarter can sink you. The fix is not to prove a point by going fully remote. It is to decide each role’s arrangement on the merits, state it plainly, and let your own apply-to-offer rate tell you whether your policy is winning candidates or quietly repelling them.