Most people will tell you they love working from home. No commute, no small talk with a coworker they cannot stand, no one looking over their shoulder. Surveys consistently find that the large majority of remote-capable workers say they would be happiest working from home, and many say they would take a real pay cut to keep it that way.
And yet a major new study, published this month in Science, found something that complicates that preference considerably: remote work has significantly deepened isolation and psychological distress in this country, and it accounts for roughly a third of the overall decline in American mental health between 2011 and 2024.
This is worth sitting with, because it runs counter to what most people believe about their own choices. The thing that feels most personally liberating may also be one of the most significant drivers of a worsening mental health landscape. That contradiction is not really about remote work specifically. It is about something more fundamental: how much human beings depend on incidental, unplanned, in-person contact with other people, and how easily that dependency goes unnoticed until it is gone.
What the Research Found
The study compared workers in jobs that could plausibly be done remotely, such as finance and software engineering, with workers in jobs that require in-person presence. People in remote-capable roles worked from home roughly three times as often in 2024 as they did in 2019, and as they did, something measurable happened to the texture of their days.
The majority of remote workers spend their entire workday completely alone. Over half report feeling less connected to their colleagues. Even in digital communication, remote workers receive less feedback from coworkers and have meaningfully less contact with people outside their immediate team.
What is particularly striking is that people did not compensate for this lost workplace contact by socializing more elsewhere. More days passed with no social contact of any kind — no greeting from an office mate, no small exchange with a barista, no nod to a fellow commuter. These are the kinds of interactions that are easy to dismiss as trivial. The research suggests they are not.
Workers in remote-capable jobs saw steeper increases in psychological distress, mental health visits, and antidepressant prescriptions than workers whose jobs required them to be physically present. And the pattern began in 2020 and has not let up since — pointing toward remote work itself, rather than more recent anxieties like AI displacement, as the driving factor.
The effect was not distributed evenly. People who lived with a spouse and children saw their mental health hold relatively steady. People who lived alone experienced a roughly 20 percent decline in psychological wellbeing. Isolation, in other words, compounds. The less embedded a person already is in other forms of daily human contact, the more remote work appears to cost them.
Why the Cost Is So Easy to Miss
If remote work is doing this much damage, why does it not feel that way to most people experiencing it?
Part of the answer is the pace at which the cost accumulates. Loneliness that builds gradually does not announce itself as loneliness. It gets attributed to other things — a hard year, a breakup, a friendship that drifted, the ordinary tiredness of getting older. The texture of an isolating life is rarely dramatic. It is just quieter than it used to be, in ways that are easy to rationalize as unrelated to where you happen to be sitting during the workday.
This is consistent with something I see often in clinical work. People rarely arrive in therapy saying "I think my life lacks adequate community." They arrive describing low mood, flatness, a vague sense that something is missing, irritability, or difficulty finding motivation. The absence of community does not present as a clearly labeled deficiency. It presents as depression, anxiety, or a diffuse dissatisfaction that does not trace easily to any single cause — which is exactly what makes it so important to ask about directly.
There is also a structural reason the cost is hard to see: a half-empty office is not an appealing alternative to working from home. When most of your colleagues are also remote, going into an office does not restore the social environment that used to exist there. The choice many people are actually weighing is not "office community versus home isolation." It is "isolation at home versus isolation in a quiet office." Neither option, as currently structured, delivers what used to happen by default.
The Office Was Doing More Than We Realized
One of the more striking findings referenced in this research is that the workplace has historically been the single most common place where American adults form friendships — ahead of religious communities, neighborhoods, children's schools, and sports teams.
This is worth pausing on, because it reframes what was lost when offices emptied out. It was not just a commute and a desk. For a very large number of adults, it was the primary infrastructure through which adult friendship actually happened. Adult friendship, unlike childhood and adolescent friendship, rarely has a built-in structure that produces it automatically. Work was that structure for millions of people, largely without anyone noticing it was serving that function until it stopped.
This matters clinically because friendship and social connection are not peripheral to mental health. They are among the most well-established protective factors we have. Strong social ties are associated with lower rates of depression and anxiety, better physical health outcomes, longer life expectancy, and greater resilience in the face of life stressors. Loneliness, conversely, has been associated with health risks comparable to those of smoking and obesity. This is not a soft or sentimental claim. It is one of the more robust findings in health psychology.
When the primary structure that produced incidental adult connection disappears, and nothing replaces it, the consequences are not abstract. They show up, as this research demonstrates, in mental health visit rates and prescription data.
This Is Not an Argument for Returning to 2019
It would be easy to read this research as an argument for mandatory full-time office return, and that is not the right conclusion to draw from it.
The prepandemic norm, in which work occupied every hour of the workday and frequently crowded out time with friends and family, was its own kind of problem. Many people who value remote work do so for genuinely good reasons: more time with their children, more flexibility during illness, freedom from long commutes, escape from difficult office dynamics. None of that should be dismissed.
The point is not that offices are good and remote work is bad. The point is that human connection requires structure, and that structure does not happen automatically. Whatever the working arrangement, something has to actually produce the in-person contact that our nervous systems and our psychological wellbeing depend on. For decades, the office provided that structure as an incidental byproduct of simply showing up. Remote work removed the structure without anyone deciding to remove the connection it produced, and most people have not yet found something to replace it.
What This Means in Practice
The research points toward a conclusion that is genuinely useful, even if it requires intention that used to be unnecessary: connection has to be built deliberately now, because it is no longer happening by accident.
This looks different depending on your circumstances, but a few things are worth naming directly.
If you work remotely, audit your week for incidental human contact. Not scheduled, purposeful socializing — actual incidental contact. A coffee shop where someone knows your order. A walk where you might run into a neighbor. A coworking space, even occasionally. These small, low-stakes interactions are not filler. The research suggests they matter more than most people assume.
If you live alone, take the finding about compounding isolation seriously. The mental health cost of remote work fell hardest on people without a spouse or family at home to provide a baseline of daily contact. If that describes you, building deliberate structures for connection is not optional self-care. It is a meaningful protective factor for your mental health.
Consider what structures actually produce connection, rather than just opportunities for it. A standing weekly lunch with a friend produces connection more reliably than a general intention to "see people more." Structure that requires no ongoing willpower to maintain tends to outperform good intentions.
If you manage other people, recognize that this is a workplace mental health issue, not just an individual one. Organizations that have taken this seriously have restructured physical spaces to centralize rather than isolate, rethought how they recognize the often-invisible work of connecting teams, and built deliberate touchpoints like regular one-on-ones into how teams function. These are not perks. They are interventions with measurable mental health value.
A Closing Thought
Robert Putnam wrote, more than two decades ago, that Americans were increasingly "bowling alone" — disengaging from the associational life that had once structured American communities. What this new research suggests is a continuation of that trajectory into the texture of an ordinary workday: many of us are now, in a real sense, typing alone.
The remedy is not nostalgia for a five-day office week that had its own real costs. It is a recognition that community does not assemble itself. It has to be built, and in the absence of the structures that used to build it without anyone trying, building it now requires intention that can feel unfamiliar, even effortful.
That effort is worth making. Community is not a nice-to-have layered on top of mental health. For many people, it is one of the load-bearing structures of it.
APA CITATIONS
Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10(2), 227–237. https://doi.org/10.1177/1745691614568352
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. Simon & Schuster.
Yang, E., & Pallais, A. (2026). The mental health costs of remote work. Science. [As referenced in The New York Times, June 2026]
The New York Times. (2026). Why remote work has made America lonelier — and what to do about it. The New York Times.