Online dating's long shadow on commitment

Michael Rosenfeld's "How Couples Meet and Stay Together" survey, which he has run since 2009 with successive waves, gives the cleanest empirical picture of the shift. In 1995, fewer than two percent of US heterosexual couples met online. By 2017, that figure crossed forty percent and made online dating the modal pathway. Same-sex couples crossed the threshold a decade earlier — the apps solved a thinner local market first. The line is not a slow drift; it is a hockey stick whose hinge sits around 2013, when smartphone penetration and Tinder's swipe interface fused. Rosenfeld's contribution is not opinion about whether this is good or bad; it is the dating of when the transition actually happened, against which every other claim about the era has to be calibrated.

Eli Finkel's team, writing in Psychological Science in the Public Interest, argued that online dating's three core features — access, communication, and matching — each carry a trade-off, and that the access feature in particular shifts users into what they called an "assessment mindset." Once a partner is rendered as a profile next to other profiles, the cognitive operation becomes comparison rather than commitment. Comparison is hard to switch off once it starts. People who report being "in love" after meeting on an app still describe the early phase as a winnowing process; people who met without apps more often describe it as a discovery. The frame matters because the frame structures what counts as evidence about the relationship.

In economic terms, apps drove the search cost of finding a partner toward zero while leaving the matching cost — the work of actually pairing well — roughly constant. The predictable result is more searching. Median age at first marriage in the US rose from 26.1 for men and 23.9 for women in 1990 to 30.5 and 28.6 in 2022. The trend predates apps, but the apps did not arrest it; they accelerated it. When the marginal cost of looking at one more profile is a thumb movement, the rational stopping rule is later than when the marginal cost was a Saturday night out.

Barry Schwartz's "paradox of choice" applies with unusual force to mating because mate selection has a built-in irreversibility that consumer choices lack. When users are shown a dozen profiles, satisfaction with the chosen partner is generally higher than when shown several hundred — but the platforms are economically incentivized to show several hundred, because engagement is their revenue model. The user's interest in resolving the search and the platform's interest in extending the search are structurally opposed. This is not a side effect; it is the business model, and it shapes every UI decision the platforms make.

Pre-app commitment had a default-on character: continuing was easier than ending. App-era commitment has a default-off character: each week is, in a small way, a re-decision. The shift is partly technological — the option to return to the market is one tap away — and partly cultural, in that the technology has normalized the framing. Couples now talk explicitly about "checking in" with each other, "redefining the relationship," and "reassessing" — language that would have been alien in 1985. The active form is more honest, but it loads more weight onto the participants' communicative competence than most couples actually possess.

In pre-app pairing, meeting a partner meant meeting their network — friends, family, workplace — at roughly the same time. The network functioned as both a vetting mechanism and a commitment scaffold: leaving the partner meant leaving the network, which raised the cost of leaving. App pairing decouples these. Couples can date for months before either party meets the other's friends, and either party can exit without social consequences. The vetting function migrates to the partners themselves, who are poorly positioned to do it, and the scaffolding function disappears, replaced by nothing.

A subtlety in Rosenfeld's data: couples that meet online and marry are not, on average, less stable than couples that meet offline and marry. They are, if anything, slightly more stable, controlling for age and education. The selection effect is doing real work — these are people who searched deliberately, often across larger pools, and chose with eyes open. The pessimistic story about apps is not "the marriages are worse." It is "fewer marriages happen, and the people who don't marry are increasingly invisible to themselves and to each other."

Robert Putnam's Bowling Alone argued that American social capital — the dense weave of weak ties through clubs, churches, leagues, and civic groups — had been declining for decades. Apps both fed on and accelerated this thinning. When the church social and the bowling league no longer function as marriage markets, an app fills the gap, but it fills it with a different kind of tie: dyadic, transactional, low-context. The replacement is not equivalent. A church social produced not only marriages but also the networks that supported those marriages once formed.

The app market is famously skewed: men send most messages, women receive most, and the distribution of attention is steeply unequal on the receiving side. A widely cited Hinge analysis found that the top twenty percent of men receive the majority of likes; the corresponding figure for women is much less skewed. The result is that the median male user has a worse app experience than the median female user, but the top decile of male users has a far better experience than the top decile of female users. This asymmetry distorts perceptions of the market and feeds resentment that leaks back into how both genders enter relationships when they do form.

Ray Oldenburg's "third place" — the bar, café, hair salon, corner store where people who are not family or coworkers actually encounter each other — has been thinning since the 1980s. Apps did not cause this, but they made the loss bearable. A young adult in 2024 can find a date without a third place. They cannot, however, find the casual ambient sociability that third places used to provide for free, and that sociability turns out to have been an input to commitment that nobody priced. Without it, partners depend more heavily on each other for ordinary social need, which raises the temperature of every relationship.

In every country with high app penetration and high smartphone adoption, fertility has fallen below replacement. The causation runs through many channels — economic, cultural, educational — but the romantic-pairing channel is real. Couples who do not form, or who form late, do not have children, or have fewer. South Korea's fertility rate of 0.72 (2023) is the extreme case, but Italy, Spain, Japan, and increasingly the US are on the same curve. Whatever else apps have done, they have changed the demographic arithmetic of the developed world, and the change is not reversing.

The 2nd Law asks for sustained attention to one thing at a time. In the app environment, sustained attention is the rarest commodity. A user who wants to take commitment seriously has to actively reverse the platform's incentives: delete the apps once a relationship begins, refuse the comparative frame when it intrudes, and rebuild the network-and-scaffold functions that apps stripped out. None of this is what the apps want. The platforms profit from churn. The user who wants commitment has to be a worse customer for the platform — and most users are not willing or able to make that trade until they have already accumulated years of evidence about its costs.

1. Rosenfeld, Michael J., Reuben J. Thomas, and Sonia Hausen. "Disintermediating Your Friends: How Online Dating in the United States Displaces Other Ways of Meeting." Proceedings of the National Academy of Sciences 116, no. 36 (2019): 17753–58.

2. Finkel, Eli J., Paul W. Eastwick, Benjamin R. Karney, Harry T. Reis, and Susan Sprecher. "Online Dating: A Critical Analysis from the Perspective of Psychological Science." Psychological Science in the Public Interest 13, no. 1 (2012): 3–66.

3. Twenge, Jean M. iGen: Why Today's Super-Connected Kids Are Growing Up Less Rebellious, More Tolerant, Less Happy — and Completely Unprepared for Adulthood. New York: Atria Books, 2017.

4. Twenge, Jean M. Generations: The Real Differences Between Gen Z, Millennials, Gen X, Boomers, and Silents — and What They Mean for America's Future. New York: Atria Books, 2023.

5. Julian, Kate. "Why Are Young People Having So Little Sex?" The Atlantic, December 2018.

6. Haidt, Jonathan. The Anxious Generation: How the Great Rewiring of Childhood Is Causing an Epidemic of Mental Illness. New York: Penguin Press, 2024.

7. Murthy, Vivek H. Together: The Healing Power of Human Connection in a Sometimes Lonely World. New York: Harper Wave, 2020.

8. Putnam, Robert D. Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster, 2000.

9. Cherlin, Andrew J. The Marriage-Go-Round: The State of Marriage and the Family in America Today. New York: Vintage Books, 2010.

10. Cohen, Philip N. "The Coming Divorce Decline." Socius: Sociological Research for a Dynamic World 5 (2019): 1–6.

11. Galician, Mary-Lou. Sex, Love, and Romance in the Mass Media: Analysis and Criticism of Unrealistic Portrayals and Their Influence. Mahwah, NJ: Lawrence Erlbaum, 2004.

12. Schwartz, Barry. The Paradox of Choice: Why More Is Less. New York: Ecco, 2004.