The '57% Never Dated' Number Was Wrong From the Start · Issue #174

Opening

Dear reader, around Valentine’s Day two years ago, a headline circulated across every community online: “57.3% of people in their 20s and 30s (the ‘2030’ generation) have never dated.” Reactions split roughly into two camps — those who found it reassuring, and those who thought something was off.

The skeptics were right. The denominator in that number was wrong.

Then on the 21st, KBS’s investigative program Sisa Gihoek Chang (“Current Affairs Planning: Window”) aired an episode titled “The Disappearance of Dating.” The report combined a five-country comparative survey with a fresh reanalysis of raw government statistics. So today I want to do two things: first, strip away the inflated number, and then look at what’s left standing in its place. Here’s the conclusion up front — the real signal lies outside the dating statistics.

How Was That 57.3% Number Built?

The survey in question was published in February 2024 by PMI, a data consulting firm. Among 1,174 unmarried men and women surveyed nationwide, 57.3% of those who said they’d never dated were in their 20s or 30s.

Two things got conflated here. First, this isn’t “57.3% of people in their 20s and 30s” — it’s “the share of the never-dated group that happens to be in their 20s and 30s.” Second, and more importantly, the survey pool was restricted to unmarried men and women from the start. Married people weren’t in the sample at all. That means everyone who dated and then got married was excluded entirely before anyone even asked “have you ever dated?” The sampling frame¹ had already decided the conclusion in advance.

Segye Ilbo, a Korean newspaper, plugged in the marriage rate from Statistics Korea and recalculated the figure against the entire population in their 20s and 30s. The result came out to roughly 21%. Broken down by age, it’s 33% for people in their 20s and 9% for those in their 30s. The sharp drop to 9% in the 30s means most people do end up dating as they get older. That’s a completely different picture from 57.3%.

But even after stripping away the exaggeration, something remains: one in three people in their 20s. In a separate survey by the same company, this one of 3,000 people aged 20 to 69 nationwide, only 44.0% of respondents in their 20s said they were currently dating. The top reason cited for not dating was “being alone is comfortable,” at 33.6%.

That last response is where today’s story really begins. I don’t take that 33.6% at face value.

The Real Number Isn’t in the Dating Statistics

The same debate is raging in the US. In a conversation published by The New York Times Opinion section on the 22nd, the host cited a survey finding that one in four Gen Z adults has never had sex, alongside another figure: the share of young people who hadn’t had sex in the past year doubled from 12% in 2010 to 24% in 2024.

An analysis of General Social Survey² data by the Institute for Family Studies tells a similar story. Among adults aged 18 to 64, the share having sex at least once a week fell from 55% in 1990 to 37% in 2024.

So far, this is a familiar story. But the same report contains a number that gets cited far less often.

It’s how many hours per week young Americans spend with friends. That figure fell from 12.8 hours in 2010 to 6.5 hours in 2019. It dropped as low as 4.2 hours during the pandemic and, as of 2024, is still stuck at 5.1 hours. In 15 years, 60% of it simply vanished.

What’s worth noting is the timing. Most of that decline had already happened before the pandemic. COVID didn’t create this phenomenon — it just cemented one already underway. Over the same period, the share of 18-to-29-year-olds living with a spouse or partner fell from 42% in 2014 to 32% in 2024. According to psychologist Jean Twenge’s analysis, the share of American high school seniors who’d been on a date in the past year fell from 80% in 2000 to 45% in 2023.

Why around 2010? That timing lines up exactly with the spread of smartphones. Still, I’m cautious about the one-line explanation of “it’s the smartphones.” Over the same period, young people’s living arrangements, working hours, and urban structures were all shifting too. What’s certain is the sequence: time spent meeting people declined first, and the drop in dating came after.

Thinking about the sequence changes the story. Meeting people has to happen before relationships can form, and relationships have to exist before dating can. Dating is the last link in this chain. But we’ve been counting only the numbers at the last link while missing that the first link has collapsed.

Leisure Time Grew While Face Time Shrank

What about Korea? Statistics Korea’s Time Use Survey³ gives an interesting answer.

In the 2024 survey, Koreans’ average daily leisure time was 5 hours 8 minutes — 21 minutes more than five years earlier. That directly contradicts the common assumption that people can’t meet others because they have no time.

But the picture sharpens once you see where those extra 21 minutes went. Over the same period, media consumption time rose 19 minutes, from 2 hours 26 minutes to 2 hours 45 minutes — nearly the entire increase. In particular, the share of people watching video content on weekdays jumped from 15.8% to 40.4%. Nearly every extra minute of leisure went straight into screen time.

Time spent alone rose too. Among people in their 20s who ate breakfast, 58.5% ate it alone. In Statistics Korea’s Social Survey, the share saying they had “no one to talk to when feeling depressed or discouraged” climbed from 21.8% in 2019 to 30.6% in 2021 and 31.6% in 2023. The Korea Institute for Health and Social Affairs has estimated that as many as 540,000 young people may be socially isolated or in a state of self-imposed reclusion.

And KBS Sisa Gihoek Chang’s reporting adds a decisive piece here. When the production team reanalyzed raw government data from a survey of roughly 4,800 unmarried men and women, they found that the desire to date was similar regardless of income level. But the rate of that desire actually turning into dating varied sharply. Among unmarried men, the dating rate gap between the lowest and highest income brackets was roughly twofold.

The team’s five-country comparison pointed the same way. The correlation between income and dating was pronounced in Korea, China, and Japan, but relatively weak in the US and Spain. Regional gaps showed up too. In rural areas, a skewed sex ratio combined with young women migrating to the greater Seoul area, reducing the very opportunity to meet peers — to the point where local governments have started organizing group meetups themselves.

The regional story is worth spelling out. The reporting team argues that the skewed sex ratio at birth around 1990 is now showing its full effect as that generation reaches marriageable age. On top of that, young women moving to the greater Seoul area for jobs and schools has widened the sex-ratio gap in the provinces even further. There are simply fewer people to meet, physically. That’s not a problem that individual initiative can fix.

If the intent is the same but the outcome differs, that’s not a matter of preference. It’s a matter of conditions.

Why We Shouldn’t Call This a Dating Problem

Let’s go back to that 33.6% who said “being alone is comfortable.”

In the New York Times conversation, journalist Carter Sherman makes a point along these lines: being powerless isn’t cool, but being in control is. So people relabel circumstances they can’t change themselves as if they’d chosen them. Columnist Christine Emba touches the same nerve. Declaring “I chose celibacy” or “I enjoy solitude,” she argues, is a way of reclaiming a sense of control. Editor Nadja Spiegelman summed it up by saying that what separates loneliness from solitude ultimately comes down to whether it was chosen.

Seen through this lens, the answer “being alone is comfortable” reads differently. For a good many people, it isn’t a statement of preference — it may be an adaptation to circumstances. In an environment where chances to meet people are shrinking, dating costs feel burdensome, and the odds of rejection have climbed, people choose the least painful explanation available. The conversation also notes that the scale of rejection on dating apps — approaching 50 people and being turned down by 49 — is unprecedented in the history of dating.

That’s exactly why policy misfires. Local government meetups and matchmaking-app subsidies aim directly at the dating rate as a number. But the dating rate is a lagging indicator⁴. If the structures of time, space, and cost sitting further up the chain remain unchanged, shaking only the last link produces a brief blip that snaps right back.

Let me be more precise. This doesn’t mean these programs are pointless — organizing occasions to meet people does touch an earlier link in the chain. The problem is scale and persistence. An event held a few times a year can’t make up for social time that’s vanished at a rate of seven hours a week. And the very fact that meeting people now requires a special occasion to happen at all is itself a sign that everyday social networks have already grown thin.

Doesn’t this structure look familiar? It’s the same story we covered in Issue 131. Korea cut childcare costs to among the lowest in the world, yet the birth rate didn’t rise. When people said “it’s expensive,” they didn’t really mean daycare tuition. When young people today say “being alone is comfortable,” there’s a good chance that phrase isn’t really pointing at dating either.

Oz’s Lens

In years of teaching data and building GTM (go-to-market) strategy, this is the mistake I’ve run into most often: mistaking a lagging indicator for a leading one.

There’s a scene I’ve seen play out over and over. A client comes in worried about high churn. So they blast discount coupons at customers about to leave. The number improves briefly, then reverts a few months later. Rewind the data far enough and the real problem sits in onboarding: users who never touched a core feature in their first week end up, months later, right on the churn list. Churn was the outcome. Activation was the cause.

The dating rate is churn. Social time is the activation metric.

So I’m skeptical of explaining this phenomenon as a generational shift in values. Values have of course changed. But as the KBS reporting showed, intent was similar regardless of income. If the desire hasn’t changed but the conditions have, this isn’t a story about changing hearts.

So what should we watch instead? In place of the dating rate, I’d track two things: how many hours a week young people spend with others face-to-face, and in which income brackets that time is shrinking fastest. But Korea has essentially no time series tracking this consistently by age group. While the US has been charting 15 years of data, we’re checking only the overall average once every five years.

One more thing worth adding: this isn’t a problem exclusive to young people. The decline in social time shows up across every age group. It’s just visible earlier, and more sharply, among the young.

Closing

“Half of people in their 20s and 30s have never dated” is an illusion created by a flawed sampling frame — the real, properly calculated figure is roughly 21%. Even so, one in three people in their 20s remains, and behind that number sits a far quieter, more thoroughly collapsed indicator: the time young Americans spend with friends has lost 60% of itself in 15 years.

Dating didn’t decline first — time spent meeting people did.

If how often you see friends has noticeably dropped in the past few years, tell me in the comments what took that time. Was it work, distance, or has simply making plans become exhausting? I’ll gather reader responses and put together a follow-up in the next issue.

If you’ve been feeling disconnected from people for a while, you can reach out for support through your local youth center or Korea’s Health and Welfare Counseling Center (dial 129).

💬 Tell me in the comments what’s taken away your time with friends. I’ll fold your answers into the next issue. 📨 If you know someone who’s been wondering why meeting people feels so hard these days, pass this along.

References & Further Reading

Primary sources

Institute for Family Studies, “The Sex Recession: The Share of Americans Having Regular Sex Keeps Dropping”, 2025. Link ··· This is the core evidence behind today’s piece. I’d recommend looking at the social-time graph before the sex-frequency statistics.
Nadja Spiegelman, Carter Sherman, Christine Emba, “Gen Z-ers Are Ditching Sex. They Might Be Onto Something.”, The New York Times Opinion, 2026. Link ··· The passage distinguishing “choice” from “powerlessness” became the backbone of Chapter 4 in this issue.
KBS Sisa Gihoek Chang, Episode 556, “The Disappearance of Dating,” July 21, 2026. Link ··· Likely the first domestic analysis to separate income from the intent to date. I’d recommend watching the full broadcast.
Statistics Korea, “2024 Time Use Survey Results,” 2025. Link ··· Placing leisure time and media-use time side by side makes this issue’s argument clear at a glance.

Background

Segye Ilbo, “[Exclusive] 57.3% of People in Their 20s and 30s Have Never Dated? A Look at a Strange Statistic,” 2024. Link ··· The article that flagged the denominator problem behind that 57.3% figure. Good practice material for reading statistics critically.
Ministry of Health and Welfare, “Support Measures for Isolated and Reclusive Youth,” 2023. Link ··· Contains the source and calculation behind the 540,000 estimate.

A Related Past Issue

Empty Strollers and ₩380 Trillion (Issue 131) ··· This is the next link in today’s chain. It covered, through the same structure, why the birth rate didn’t rise even after childcare costs fell.

📝 Glossary

The author, Kwangseob Ahn, is a professor of business administration at Sejong University and lead consultant at OBF (Oswarld Boutique Consulting Firm). He teaches statistics and data analysis, including business data management and business analytics, while leading GTM and AI strategy consulting in the field, designing the seam between technology and business. He has published academic research on a memory architecture for AI dialogue systems (HEMA) and runs Daily Arxiv, a daily curation of global AI papers. He holds a master's from Korea University's Graduate School of Technology Management and a KMBA. He is the author of Homo Brainless: The People Who Outsource Their Thinking.

Footnotes

Sampling frame: The list of the population from which a survey draws its sample. If this list doesn’t match who you’re actually trying to study, the conclusion tilts entirely off — no matter how precisely the survey itself is conducted. ↩
General Social Survey (GSS): A landmark American social survey run by researchers at the University of Chicago since 1972. By asking the same questions decade after decade, it serves as a baseline for comparing generational change. ↩
Time Use Survey: A survey Statistics Korea conducts every five years, recording how much of each 24-hour day people spend on which activities. Its value lies in capturing actual time allocation rather than the self-reported “perceptions” that typical survey questions elicit. ↩
Lagging indicator: A number that shows a result only after it has already happened. Trying to move this kind of metric directly tends to produce only short-lived effects. To create lasting change, you need to find the leading indicator sitting upstream, on the causal side. ↩