Remote Workers Reported the Best Well-Being—But Productivity Wasn’t Tested

Fully remote employees reported the highest well-being in a study of 7,704 workers at one large Southwestern U.S. health-care organization. Fortune’s Aug. 21 coverage highlighted that result while noting a crucial boundary: the researchers did not track productivity.

The Aug. 21 report followed an Aug. 19 University of Colorado Boulder release about the peer-reviewed study. The evidence supports a narrow conclusion: remote work was associated with higher self-reported well-being in this workforce, while location was not a strong direct predictor of turnover and the research did not establish whether any arrangement produced better work.

What the study measured

The peer-reviewed Frontiers paper, published July 29, analyzed a 2023 engagement survey and employment records collected one year later. Its final sample included 1,869 fully remote employees, 2,099 hybrid employees and 3,736 fully onsite employees.

• Work arrangements: Managers classified employees from scheduled arrangements in the organization’s human-resources system. Remote meant zero onsite days, hybrid typically meant two or three onsite days per week, and onsite meant zero remote days.
• Well-being: Employees rated four statements about organizational care for their health, energy from work, managerial support and life balance on a five-point scale.
• Connection: Employees supplied three to five words describing organizational culture. Researchers used a purpose-built dictionary to identify words associated with inclusion, unity, care and collaboration.
• Retention: Employment status was checked one year after the survey, providing an observed turnover outcome rather than an intention to stay or leave.
• Productivity: The study collected no outcome for output, work quality, speed, revenue, customer results or individual performance.

The remote group’s average well-being score was 4.22, compared with 4.12 for hybrid employees and 3.89 for onsite employees. Those differences were statistically significant, making well-being the study’s clearest comparative finding.

Why productivity remains unanswered

Productivity appeared in the wider return-to-office debate discussed around the research, but it was not one of the study’s three outcomes. The statistical models examined well-being, culture-language indicators and turnover; none tested how much work employees completed or how well they completed it.

Higher well-being may coincide with effective performance, but it is not a performance measure. This dataset therefore cannot show that remote employees were more productive, that onsite employees were less productive or that changing an attendance policy would alter output.

The design also cannot establish that remote work caused the higher well-being scores. Employees were working under existing arrangements rather than being randomly assigned, and the three groups differed in the kinds of jobs they performed. Remote positions tended to be administrative or analytical, while onsite positions were more likely to involve clinical or operational work with potentially different physical and emotional demands.

Retention and connection were more qualified

Turnover after one year was 7.8% among remote employees, 8.3% among hybrid employees and 8.8% among onsite employees. The ordering favored the remote group, but the differences were small and not statistically significant; work arrangement was not a significant direct predictor once well-being was included in the model.

Higher well-being was significantly associated with lower odds of turnover across the sample. The researchers also estimated a small indirect relationship running from work arrangement through well-being to turnover, but an observational mediation model does not prove that changing work location will cause employees to stay.

The connection finding requires similar care. Remote employees used positive-connection words somewhat more often than the other groups, yet the weighted regression estimates were not statistically significant. The measure also captured the presence of selected culture words—not trust, relationship depth, interaction frequency or collaboration quality—so it challenges a simple isolation assumption without proving equivalent teamwork.

What the findings permit employers to decide

The study weakens the claim that office presence is inherently necessary for employee well-being or positive connection in this organization. It does not establish that every eligible role should become remote, and it offers no productivity evidence for or against an RTO mandate.

• Define the objective: Treat well-being, retention, connection, mentoring, innovation and productivity as separate outcomes rather than interchangeable reasons for office attendance.
• Use the study only for measured outcomes: It shows that fully remote work can coexist with comparatively high reported well-being and does not necessarily eliminate positive culture language.
• Measure performance directly: A productivity-based policy needs role-appropriate measures of output, quality, timeliness or customer impact that this research did not provide.
• Compare like with like: Job eligibility and job demands can differ systematically between remote and onsite groups, obscuring the effect of location.
• Evaluate policy changes: A limited or phased change can track the intended outcome alongside well-being, voluntary turnover and other possible consequences.

This framework does not assume that location is irrelevant. It requires an attendance rule to be judged against the particular problem it is supposed to solve.

The evidence stops at one employer

All participants worked for a single large health-care organization, which helped hold organizational culture and systems relatively constant but limits generalization. The final sample also included only respondents who completed the open-ended culture question, well-being was self-reported, and the connection measure relied on a researcher-created dictionary that was later reduced to a yes-or-no indicator.

The current evidence is therefore specific but useful: remote employees in this organization had the highest well-being scores, while connection and turnover analyses found no statistically significant direct disadvantage for remote work. Studies across additional employers, stronger controls for job type, policy experiments and objective performance measures are still needed before researchers can answer the productivity question or justify a universal RTO rule.