Two recent studies reached opposite verdicts on remote work and mental health. Both leave out much of what shapes your actual day. Here are the numbers, side by side.

The distress data. Researchers from the New York Fed, the University of Virginia, and Harvard analyzed more than 588,000 workers for a study published in Science. Nearly 9 in 10 people in remote-capable jobs spend the workday completely alone. They also socialize less after hours, and entire days can pass with no face-to-face contact. Distress, mental health visits, and antidepressant prescriptions all run higher. The authors estimate remote work accounts for about a third of the rise in mental distress from 2011 to 2024, mostly among people who live alone. USA Today reports that distress for people living alone nearly doubled when they worked remotely. Co-author Emma Harrington, a University of Virginia economist, says everyone logs more hours alone, largely because of the workday. For people who live alone, that can mean a full day with no social contact and no trip outside the house.

The belonging data. A separate team tracked 7,704 employees at a large healthcare system for a year. Remote workers used words like belonging, collaboration, and inclusion to describe the culture 61% of the time. Onsite workers did so 55% of the time. Remote workers were also the most likely to report a positive sense of connection. Turnover was statistically indistinguishable between the groups. What predicted who left was wellbeing, which ran higher among remote workers.

The pushback. Stanford economist Nick Bloom points to two randomized trials in which remote workers reported better mental health. He says working from home cuts commuting stress and gives people more control over their time, and that it affects people differently. Flexible work researcher Gemma Dale notes that the worst symptoms in the Science data sit with people who live alone. She calls the remote experience contextual. Psychotherapist Ruth White, who worked remotely for nearly a decade, reports seeing clients who counted on the office to meet all their social needs.

One named data point. Newton Cheng spent 17 years overseeing health and performance programs at Google. After leaving to work on his own, he says the evidence favors remote and hybrid work for wellbeing, yet he personally finds remote work isolating. His fix: a Tuesday night pizza truck at a neighborhood park with his two kids, and lifting at a friend's gym instead of in his garage.

The shared gap. Both studies compare desk locations and skip workload, manager support, caregiving, and demographics, according to Brian Elliott of Work Forward. The Science study tracks people in remote-capable jobs, not confirmed remote workers, and people who live alone are only part of the workforce. In the healthcare study, remote roles skew administrative while clinical roles stay onsite. One variable location data can hide: University of Toronto researchers found women face more bias in the office than working remotely.

Two datasets, two verdicts, one blind spot.

Sources: Brian Elliott, "Why the Office Isn't a Cure for Loneliness" (Charter); USA Today, "Too many Zooms, too little face time. Is remote work making us lonely?"