The Return-to-Office Mandate Experiment Has Two Years of Data. Here Is What It Actually Shows.

On a Tuesday morning in 2025, there’s something strange about passing any large corporate headquarters. Yes, the parking lots are more crowded than they were in 2022. The elevators are operating once more. However, a closer look reveals that the open-plan floors are filled with workers wearing headphones and sitting in back-to-back video calls with their laptops open, essentially replicating the remote setup they left at home. There are people in the office. The question of whether it is genuinely effective is quite different.

The return-to-office experiment that corporate America started with such conviction has now been the subject of two years of serious, peer-reviewed data. The proof is not nuanced. According to Stanford economist Nicholas Bloom’s Global Survey of Working Arrangements, which collected responses from more than 16,000 individuals in 40 countries, hybrid workers perform on par with full-office workers. comparable results. The distinction is that hybrid employees have a 33% lower resignation rate. It’s not a rounding error. This type of retention gap has the power to alter entire departments and budget projections.

Mark Ma of the University of Pittsburgh monitored S&P 500 companies and discovered something that merits greater consideration than it has gotten. Return-to-office requirements were announced by companies after, not before, their stock prices had declined. Additionally, the mandates did not result in a subsequent increase in firm value or financial performance—this is the section that executives typically omit from board presentations.

Millions of Glassdoor reviews show that they did result in measurably lower job satisfaction. It appears that businesses were looking to remote work as a handy way to solve issues that had nothing to do with where their workers were seated.

It is difficult to ignore how skillfully the corporate defense of RTO changed based on the target audience. The pitch to staff members focused on teamwork, impromptu conversations in the hallway, mentoring, and culture. Productivity and competitive performance were important to shareholders. When examined closely, neither narrative holds up very well. Innovation metrics reveal no discernible rise in new product development or patent filings as a result of required office presence.

Returning employees at JPMorgan Chase reportedly discovered crowded floors, slow Wi-Fi, and a lack of desks, but they still used their newly reclaimed office hours for back-to-back virtual calls. Executives’ imagined conversations in the hallway were not taking place. People were taking calls while seated on floors.

Researchers at Cornell University discovered that office rent expenses in a company’s headquarters city are actually among the best indicators of RTO policy, which helps to explain a lot of this paradox. not scores for teamwork. not standards for innovation. leases for real estate. When considering a long-term real estate commitment, a CFO needs documentation to support the decision.

RTO tells a tale about the balance sheet. Another unsettling layer was added by Mark Ma’s research: about 25% of executives said in a survey that they hoped return-to-office requirements would encourage some employees to leave on their own volition, avoiding the expense of official layoffs. Andy Jassy of Amazon was remarkably forthright when he revealed plans to reduce management layers by 15% and a complete five-day return mandate. Regardless of one’s opinion of the approach, at least the objective was made clear.

In contrast, between 2019 and 2024, remote businesses grew their revenue 1.7 times faster than those that needed an office. That is not a pandemic anomaly, but rather a five-year trend. According to tracked data, remote workers worked roughly 30 minutes more per day than their in-office counterparts and saved an average of 72 minutes of commuting time. According to a study of 60,000 Microsoft employees, remote workers put in 10% more hours per week. The productivity narrative presented in boardrooms differs from the productivity narrative found in the data.

This becomes especially expensive when it comes to the talent drain, which is the topic that usually comes up last in executive discussions. Senior women, highly skilled specialists, and long-tenured employees are the groups most likely to depart when RTO mandates take effect—exactly the people whose departure hurts the most.

They are familiar with the legacy systems. They are aware of the choices made and the reasons behind them three years ago. They have client relationships that were developed over a ten-year period. The hiring team is left attempting to recreate institutional memory using onboarding decks and Notion pages after they move to a competitor that maintained its flexibility policy.

More texture is added by the regional image. In China, workers typically work five days a week. Employees in the US and the UK are handling slightly more than two. Flexibility rights, such as the official right to disconnect after work, are being enacted by governments throughout Europe. These are not merely cultural distinctions. They represent conflicting presumptions about the source of productivity, the ownership of an employee’s time, and the true foundation of the employment relationship.

In one version of this tale, the data ultimately prevails. When enough senior leaders read the Stanford data, take note of the turnover rates, and subtly change their language from “five days” to “two or three structured days with intentional collaboration built in.” Regardless of what any particular mandate formally requires, hybrid arrangements have stabilized as the actual default for the majority of knowledge workers, according to some evidence. In most businesses, there is a significant discrepancy between office attendance tracking policy and practice.

When the next round of retention data comes in, it’s still unclear if the executives who are most dedicated to full-office returns will change their minds or if the control and real estate arguments will just outlast the evidence. It is quite evident from the research that the experiment has been conducted. Tens of thousands of workers’ worth of data collected over two years by impartial academics who have no stake in the result all point in the same direction. Performance did not improve as a result of the mandate. It did have an impact on who chose to stay.