Remote Work and Rising Mental Distress, 2011-2024

The Convenience of Remote Work and Social Contact

Working from home cuts commute time and makes scheduling more flexible. A lot of people prefer those upsides. At the same time, it’s worth examining what happens when opportunities to meet coworkers in person shrink.

In June 2026, Science published a paper on remote work and mental health. The researchers analyzed data from five US surveys covering 588,322 people between 2011 and 2024. That doesn’t mean all of them were tracked continuously for 13 years — rather, the team compared how things changed before and after the pandemic across jobs that can be done remotely versus jobs that require in-person presence.

The team estimated that the rise in remote work accounts for roughly a third of the increase in psychological distress observed over the study period. It’s a finding that suggests even a work arrangement people actively prefer may carry a burden worth watching over the long run.

🔍 Comparing Remote-Capable and In-Person Jobs

Natalia Emanuel of the Federal Reserve Bank of New York, Emma Harrington of the University of Virginia, and Amanda Pallais of Harvard University. This study, published in Science by the three economists, stands out for its sheer scale.

They pulled data from five nationally representative U.S. surveys spanning 2011-2024, excluding 2020-2021 — the years hit hardest by the pandemic — from the main analysis. This was a methodological choice to reduce the distortion from that shock. But excluding that period doesn’t erase the pandemic’s longer-term effects or every other confounding factor.

The team used a difference-in-differences¹ approach, comparing jobs that can be done remotely (finance, software, and the like) with jobs that require in-person presence. After the pandemic, the frequency of remote work in remote-capable jobs tripled compared to 2019. The gap that opened up between these two groups is the heart of the study.

Let me break down the numbers one by one.

In remote-capable jobs, people spent about an hour more alone on workdays after the pandemic compared to those in in-person jobs. Post-work socializing also declined, and there were more days with zero contact with another person at all.

The drop in contact during work hours wasn’t offset by more socializing after work. That’s exactly why we need to look beyond working hours to relationships in general.

A separate commuter experiment discussed in the paper found that participants who talked to strangers reported more positive experiences than they expected. This kind of finding makes you think about the value of brief social contact, but it should be kept distinct from results that directly test the long-term effects of remote work.

Mental health indicators showed a gap too. Remote-capable workers saw their Kessler K-6 psychological distress scores rise by about 0.1 standard deviations more than in-person workers. Depression, mental health treatment, and antidepressant prescriptions moved in similar directions. The team notes that the results largely hold up even after controlling for AI exposure by occupation or looking at within-person changes. Based on this, they estimate that the rise in remote work accounts for roughly one-third of the increase in psychological distress over the study period.

🧠 Preferred Work Styles and Their Long-Term Effects

We may prefer convenience and only belatedly notice the burden of reduced social contact. The research team suggests these effects could accumulate over time. This doesn’t mean every remote worker is misjudging their own condition.

The benefit of a shorter commute is felt immediately, but the changes that come from fewer face-to-face encounters can be hard to separate from other life changes. The researchers also discuss the possibility that workers may not fully register the long-term costs of remote work.

Behavioral economics sometimes distinguishes between the expectations we hold when making a choice and the satisfaction we actually derive from the experience. With remote work, too, we need to separately examine our preference for convenience and the longer-term changes in our well-being.

There’s also a vicious cycle at play. If you go into an office where half your colleagues are working from home, all you’re left with is a nagging “why did I even bother coming in?” feeling. A half-empty office isn’t an appealing alternative. As the social value of the office declines, people prefer remote work even more, and as more people work remotely, the office grows even emptier.

In this study, the increase in psychological distress among people living alone was roughly twice as large as among those living with family. Whether or not someone lives with others was associated with the outcome, but this analysis can’t confirm that having dinner with family is the direct cause.

🇰🇷 Korea as a Structural Variable

“This is a US study — isn’t Korea different?” you might be thinking. Fair point. Korea’s remote-work rate stands at 4.4% of all workers, roughly 960,000 people. That’s far lower than the US (23%).

But what this research suggests for Korea has less to do with the remote-work rate itself and more to do with the living conditions we need to weigh alongside it.

As of 2024, Korea had 8,045,000 single-person households, 36.1% of all households — an all-time high. In Seoul, the figure is 39.9%. That means nearly 4 out of every 10 households consist of someone living alone. The nationwide rate has also climbed from 30.2% in 2019.

And in the National Data Agency’s social survey, 48.9% of single-person households reported feeling lonely on a regular basis — 10.7 percentage points above the all-household average of 38.2%. In 2024, the number of godoksa² (dying alone, unnoticed) deaths reached 3,924, up 7.2% from the previous year.

It’s worth noting that the US study found the effect was stronger among people living alone. That said, not everyone in Korea’s single-person households is a remote worker, and household ratios don’t map neatly onto worker ratios. These figures alone don’t let us calculate that Korea would see health effects of the same magnitude.

The workplace is one of the places where adults build social relationships. Many companies, in Korea and abroad, are pushing return-to-office mandates, but simply bringing people back into the same physical space doesn’t automatically increase meaningful social contact. What matters is whether there are actual opportunities to work together and talk.

If remote work expands in Korea, we need to examine its effects based on workers’ living arrangements, existing relationships, and job type. It’s also worth checking whether workers who live alone have chances to connect with colleagues or their local community.

One more point worth flagging: the study itself acknowledges a limitation — it couldn’t distinguish between fully remote work and hybrid work. The research team speculated that hybrid work might have a protective effect, cushioning the impact of isolation. That’s a detail worth considering as Korean companies design hybrid models built around 2-3 office days per week.

Oswarld’s Lens

Honestly, reading this paper, the first thing that came to mind was an old dilemma from GTM strategy.

When building tech management strategy, I’ve often seen this pattern play out. In user surveys, “I like this feature” and “this feature actually helped me” are different questions. Quite often, the feature people say they like doesn’t match the feature that actually drives retention³. When you build exactly what users say they want, short-term satisfaction goes up, but so does long-term churn — that’s the paradox.

Reading this remote-work paper, I was reminded of that exact experience. The benefit of saving commute time and the burden of reduced social contact may not register on the same timeline. That said, we can’t simply assume that user behavior in a product follows the same logic as a worker’s mental health.

So after reading this study, I came to think that when deciding where people work, organizations also need to deliberately design opportunities for colleagues to interact. For instance, they could align in-office days for hybrid workers, or create spaces for spontaneous conversation. These are suggestions worth trying — this paper doesn’t actually verify the effect of either one.

If meeting with colleagues is left entirely to individual willpower, people can easily miss the chance because of scheduling or workload. I think organizations need to actively create conditions that make these interactions possible, while also checking in on how much burden and how much benefit employees actually experience.

Closing

In multi-year U.S. survey data, isolation and psychological distress rose more among remote-capable occupations than in-person ones, and the gap was largest among people living alone. Rather than mapping those exact figures onto Korea, we can use them as a benchmark for checking each organization’s work arrangements and employees’ living conditions.

Alongside where people work, we should also look at how many chances they actually get to interact with others.

Robert Putnam⁴​ addressed the shift in social bonds in Bowling Alone. When discussing remote work, it’s worth examining not just how often people come into the office, but the actual quality of the interaction that happens there.

This week, why not be the one who suggests lunch to a colleague first? Small contact like this can help offset the cost of isolation that the research points to.

---

How has your access to colleagues changed as you’ve moved between remote work and office days? If you’ve found something that helped, share it in the comments.

---

References & Further Reading

Primary sources

• Natalia Emanuel, Emma Harrington, Amanda Pallais, “Home alone: Remote work, isolation, and mental health”, Science, Vol. 392, June 4, 2026. : If you’re curious about the research methodology, I’d recommend checking out the Supplementary Materials section.
• Natalia Emanuel, Emma Harrington, “We Liked Remote Work. Then We Looked at the Data.”, The New York Times, June 2026. : This is a summary column written by two of the paper’s authors themselves. It’s much easier to read than the paper, and it’s woven through with personal experience (working remotely while pregnant, caring for a sick child), which makes it great for grasping the context.

Background

• Robert D. Putnam, Bowling Alone: The Collapse and Revival of American Community, Simon & Schuster, 2000. : This is the classic that warned of the “collapse of social capital” 26 years ago. The final paragraph of the Science paper picks up exactly where Putnam’s concerns left off. (There’s a Korean translation too, but the title was translated so literally that it ended up… buried in obscurity.)
• National Data Office (Korea), “2025 Single-Person Households in Statistics”, December 9, 2025

The author is Oswarld (Kwangseob Ahn). Current roles: Adjunct Professor at Sejong University, Strategy Consultant at INLEVEL9. Career, research, books, and recent work are kept current on the About page. Latest · July 2026: HEMA-2: A Consolidation-Aware Tri-Memory Architecture with Multi-Channel Scheduling for Lifelong Conversational AI.

Footnotes

1. Difference-in-Differences: A research method that compares the difference in changes shown by two groups (e.g., jobs that can go remote vs. jobs requiring in-person work) before and after a specific event (the pandemic). It estimates the effect under assumptions such as that, absent the event, the two groups would have shown similar trends. If other changes occurred differently across groups, this can affect the interpretation. ↩

2. Godoksa (solitary death): Under the law, this refers to a person who was living in a state of social isolation, cut off from those around them, and who dies by suicide or illness. In Korea, the “Act on the Prevention and Management of Solitary Deaths” has been in effect since 2021. ↩

3. Retention: The rate at which customers who have used a product or service continue to use it after a certain period. Unlike short-term satisfaction, it’s a metric showing whether people are actually feeling long-term value. ↩

4. Robert D. Putnam: Professor Emeritus of Political Science at Harvard University, who warned in his 2000 book Bowling Alone that Americans’ social bonds were rapidly weakening. He is the scholar who popularized the concept of “social capital.” ↩