Remote work has been claimed to improve mental health and to damage it. Both claims cite evidence. The resolution is that the effect depends on factors most studies do not measure.
Where it helps
Removing commutes, which are consistently among the least pleasant parts of the day in time-use research. Control over environment and schedule. Fewer interruptions for people whose work requires concentration.
For people with caring responsibilities, disabilities, or long commutes, the benefit is substantial and well documented.
Where it harms
Isolation, for people who live alone or whose social contact came primarily through work. Blurred boundaries, where the working day extends because there is no physical departure. And the loss of informal support – the colleague who notices you are struggling before you say so.
For younger workers and those early in a career, the missing mentoring and visibility has measurable effects that appear later.
Why the studies conflict
Because they average across people for whom the effect is opposite. A parent of small children and a recent graduate living alone are having different experiences, and a survey combining them finds a small effect either way.
What the evidence supports
Choice matters more than arrangement. People who chose their working pattern report better outcomes than people assigned one, regardless of which pattern.
Boundaries matter. Defined working hours, a separate space, and a ritual for ending the day all show effects.
The reasonable conclusion
Remote work is a tool that suits some people and situations and not others. Policies that impose one arrangement on everyone get the worst of both, and most organisations have learned that.