Sex, marriage, and children used to come as a complete package. You got all three or you got none. Biology made it that way. If you had sex, pregnancy followed. If pregnancy happened outside marriage, social consequences were severe enough that marriage typically followed too. And children needed two parents' resources to survive, so the whole thing held together through a combination of hormones, economics, and shame.
Between roughly 1960 and 1990, every link in that chain was severed. The pill arrived in 1960. No-fault divorce spread through state legislatures in the 1970s. Women entered the workforce. Welfare programs expanded. Each change made sense on its own terms. Taken together, they disassembled something that had been load-bearing for the entire species.
What replaced the old package was a market. A matching market where you had to find a partner, negotiate what the relationship would look like, and accept that either person could walk away at any time. Some people turned out to be very good at this market. A lot of people turned out to be terrible at it. And at the population level, something that had been almost universal, forming a family, started becoming uncommon.
This is part 3 of the What Went Wrong series
Part 1: We Forgot We Were Animals
Part 2: The Thirty Months That Broke Everything
US fertility sits at about 1.62 children per woman as of 2023. Replacement is 2.1.
Marriage rates have been cut roughly in half since 1970. Among young adults, the share who have never married has climbed past 30%. The thing that humans have done since before we were fully human, pair up and have kids, is now failing at scale in every developed country on earth.
I want to understand why. Understanding mechanisms is the only way to address them. Moralizing about it does nothing.
We are pair-bonding animals
I don't mean that as a metaphor or as some kind of poetic shorthand. Unlike most mammals, humans developed stable partnerships between males and females, held together by specific adaptations: hormones like oxytocin and vasopressin that make you want to stay with someone, jealousy that keeps you watching for threats to the bond, and the fact that both parents invest in offspring who take years to become self-sufficient. Our children are helpless for so long that a single parent was, for most of our evolutionary history, basically a death sentence for the kid. Two parents splitting the work was the arrangement that natural selection landed on.
But these partnerships always needed something outside of themselves to hold them together. People needed economic reasons to stay paired up. Laws and customs that made it hard to walk away from your commitments. Social pressure that pushed sexual energy toward marriage rather than letting it go wherever it wanted. And on top of all that, they needed meaning, stories about what family was for, that made all the sacrifice feel like it was worth it. Nobody sat down at a round table and designed any of this on purpose. It evolved through selection for thousands of years. The arrangements that produced grandchildren who survived are the ones that spread, and everything else just disappeared. Imperfect, obviously. It constrained people, created genuine injustices. But it also kept the species going.
Modern life destroyed the scaffolding piece by piece. Each was removed for defensible reasons. More autonomy, less gender-based injustice, adapting to how the economy was changing, more "rights". The problem is that scaffolding does things even when it constrains you. Take it away, and what it was holding up doesn't simply stay there floating.
Quick definition before we go further: replacement fertility means the birth rate needed to keep a population stable without immigration. That's 2.1 children per woman. The US is at 1.62 and still dropping.
Family formation collapsed because the scaffolding was removed from every direction at once
Economically, marriage became optional. Legally, commitment became reversible. Technologically, sex got separated from reproduction. And culturally, family went from being something you just did to a lifestyle option you could take or leave. Each of these fed into the others. What emerged was a mating market where fewer people pair up and even fewer of the people who pair up bother having kids.
It all starts with economics (wages went flat for non-college men, housing costs exploded, credential requirements stretched education into the late twenties). That feeds into time pressure (both partners have to work, every milestone gets pushed back by years). Which erodes trust (when half of marriages end in divorce, getting married starts to feel like a bet you'll probably lose). Which creates mating market dysfunction (people sorting by class, women's preference for marrying up slamming into the reality that there are fewer men to marry up to, dating apps turning partner selection into a scrolling past time). And all of that produces fertility collapse.
Each rung makes the one below it worse. You can't fix this by grabbing one rung.
How It Came Apart
For most of human history, sex meant pregnancy was a real possibility. That was the deal and everyone knew that. And pregnancy outside marriage carried penalties harsh enough that marriage was, functionally, the price of regular sexual access. Not a pretty system if you value autonomy. But it was extremely effective at converting sexual desire into family formation.
The pill broke the first link. By 1973, over 10 million American women were on contraceptives. For the first time in the history of the species, sex could happen with zero reproductive consequence. The entire system had been built on the assumption that sex and pregnancy were linked. They no longer were.
Legal abortion became the plan B. Even when contraception failed, reproduction could be prevented Roe v. Wade came in 1973, same year the pill hit mass adoption, and I don't think that timing is a coincidence. Shotgun marriages, the mechanism that had historically converted premarital pregnancy into marital childbirth, started vanishing. Which makes sense. Why marry to legitimize a pregnancy you can terminate?
No-fault divorce broke the meaning of marriage itself. Marriage had been a binding legal contract, permanent unless you could prove cause. California started it in 1969, and by 1985 every state had some version. Suddenly marriage was an at-will arrangement, dissoluble by either party for any reason at all. The marriage contract now had less legal force than a standard employment agreement, which is a weird outcome when you think about what's actually at stake.
Female labor force participation broke the economic logic. When women could earn their own money, marriage stopped being a survival requirement. Women's labor force participation jumped from less than 40% in 1960 to over 60% by 2000. Options expanded and the economic pressure that had been pushing people into marriage contracted. When you don't depend on someone's paycheck, you can afford to hold out for the perfect partner. Or just not bother.
Welfare and transfer programs took care of the rest. For women with children and limited income, the state could partially replace what a father's paycheck had provided. Imperfectly, sure. But enough that partnership wasn't the only way to keep a roof over your head and food on the table.
I want to be honest about something. Every single one of these changes could be defended individually. Contraception gave women control over their own reproduction. No-fault divorce freed people trapped in abusive or dead marriages. Female employment expanded human potential. Welfare prevented children from going hungry. I am not arguing any of these were wrong on their own terms.
The question is what the combined, system-level effect was.
And the combined effect was unbundling. Sex, marriage, and reproduction, which had been linked by biology and economics and law all at once, became separate transactions. You could have sex without being married, or be married without it having meaning, or have children on your own, or be married and just never get around to kids. Every combination was suddenly on the table. Freedom expanded massively.
But freedom to do what? The system had channeled behavior in a particular direction. Without it, behavior just... became chaotic.
When Partnership Became Optional
Economics drives mate selection more than people are comfortable admitting.
In virtually every human society that we have data on, what a man was worth in the mating market had a lot to do with what he could provide. Women evolved to prefer partners who would invest resources in their children. Men competed for status because status was the thing that got you reproductive access. I'm not pushing an ideology here. It's just what people actually do when you look at the choices they make rather than what they say they want.
Several things reduced what men had to offer as providers, and the thing that made it so damaging is that they all hit at more or less the same time.
Real wages for men who don't have college degrees have been declining since the early 1970s. A guy without a degree in 2020 can buy less with his paycheck than a guy in the same position in 1970. That sounds like it can't be right, but it's what the BLS data shows. And what he's worth to someone who might marry him went down with it.
Meanwhile women's income was going in the other direction. Labor force participation went from under 40% to over 60%. What they earned rose relative to what men earned. A woman who is making her own money does not need a partner's income the way that her grandmother did. And that changes the whole calculation about whether you need to be with someone or not.
And then government transfers filled in whatever was left of the gap for some women. Food assistance, housing vouchers, medicaid, cash. It's not the same as having a partner who works. But it's enough to get by, and it's enough that partnership is no longer a matter of whether or not you and your kids survive.
Add all of that up and marriage became economically optional for a lot of women. People still wanted to be with someone for companionship, for love, for status, for all the reasons that have nothing to do with money. But the raw economic force that had been pushing people toward marriage lost a lot of its strength. And when something that used to be necessary becomes merely something that would be nice to have, the rate at which people actually do it starts to drop. First it goes from universal to common. Then from common to "something most people get around to eventually." Then to "something that a lot of people just never quite manage."
Meanwhile, marriage got riskier for men. No-fault divorce meant that everything you'd invested in the relationship, career sacrifices, shared assets, the house, your relationship with your kids, could evaporate when your partner decided to leave. Any rational person doing the math on that is going to hesitate. You don't put everything into something that can be taken apart unilaterally.
Both sides recalculated. Women could afford to wait for someone better. Men faced more downside from committing. Both of those effects push marriage later and make it less likely. Which is exactly what happened.
Preferences That Didn't Adapt
Here is where the conversation gets uncomfortable, and people tend to stop being honest with you.
Mate preferences are not infinitely flexible. They show the same patterns in study after study, across dozens of societies, in exactly the configuration that evolutionary psychology predicts. These are averages across populations and plenty of individual couples break the pattern, but the averages are remarkably stable.
Women prefer partners with equal or greater education, income, and social status. Hypergamy shows up in basically every society that's been studied. The evolutionary logic is straightforward: women who chose high-investing partners left more surviving offspring.
Men are much less interested in a partner's socioeconomic status. What they respond to is youth and fertility cues. Also makes evolutionary sense, though it's not the kind of thing anyone says at dinner parties.
The thing is, these preferences evolved in a world where the gap between men and women in status was enormous. For most of recorded history, men had more education, more income and power. So it was easy to satisfy women's preference for marrying up, because the pool of men who were higher-status was large enough that most women could find one.
Then everything flipped.
Women now earn 57% of bachelor's degrees. The gap is widening at every level: associate's, bachelor's, master's, professional, doctoral. A gender gap in education that had persisted for centuries reversed completely within about two generations.
Women's earnings have risen relative to men's, especially among the young and college-educated. In several major US cities, young women now out-earn young men. But preferences didn't shift to match. Women still prefer partners of equal or higher status. Men still don't care much about status. Supply changed. Demand didn't.
The arithmetic on this is brutal and there is no way around it. If women with bachelor's degrees want men who also have at least a bachelor's degree, and women are the ones earning 57% of those degrees, then there just are not enough men who clear the bar. At every level of female education, the pool of men who would have qualified under the old rules is getting smaller. And the gap gets wider every year that goes by.
I'm not saying that anyone is wrong for wanting what they want. Preferences are what they are. You cannot talk someone out of what they find attractive, and it would be weird to try. What I'm saying is that the matching market is now lopsided in ways that no one person's choices can fix. Either the preferences shift at the population level, which could take generations if it happens at all, or big chunks of both sexes end up alone.
We're getting the second outcome.
The Apps Made It Worse
Dating apps were supposed to be the thing that fixed this. You would have access to way more people than you could ever meet in your daily life. Better information about who they were and what they wanted. Broader geographic reach. In theory it should have made the matching problem much easier to solve.
In practice the apps broke things further, and it matters to understand the mechanics.
There's a body of research on choice overload, the finding that past a certain number of options, people get worse at choosing and less satisfied with what they pick. We did not evolve to evaluate hundreds of potential mates simultaneously. When the options feel infinite, committing to any one of them gets hard. There is always someone who might be better one swipe away, and that thought is genuinely difficult to override once it's in your head.
The apps also convert partner selection into something that feels a lot like shopping. You set your filters, you compare across profiles, you keep browsing even after you've already matched with someone decent. All of which runs exactly counter to what building an actual relationship requires, which is picking one person and going deep.
And there's a distributional problem underneath everything else. Local mating markets, the kind that existed before the internet, kept most people choosing from roughly the same pool. Apps blew that open. Now everyone compares themselves against everyone. Attention concentrates at the top. Data from the platforms themselves shows the bottom 80% of men receive almost zero attention. That's not rhetorical, that's their own numbers.
For women, the apps create a distorted picture of what's actually out there. You match with men who are willing to sleep with you but who are not going to date you, and without realizing it you start calibrating your expectations to that level. You hold out for someone like that who also wants a real relationship. He doesn't show up, because he was never going to.
For men, the experience is the other kind of brutal. You get almost no responses. Over time you internalize the futility of it and just stop trying. And honestly, the rational response to a game that you can't win is to stop playing it.
And then there's the business model, which is the part I find evil. The apps use variable ratio reinforcement, the same reward schedule as slot machines. Maybe the next swipe is the one. The product is designed to keep you swiping, and keeping you swiping is a very different objective from helping you find someone. Every hour on the app is an hour not spent with a real person in a physical place where something might actually happen.
We're over a decade into the app era and relationship formation hasn't improved. Sex among young adults has actually declined. The apps captured the matching process and made it worse. I don't know how you undo that.
What This Means for Children
Family structure predicts child outcomes. This is among the most replicated findings in developmental psychology, and also one of the most politically fraught things you can say out loud, which tells you something about how ideology warps what we're allowed to discuss.
The pattern is consistent: family structure predicts school achievement, psychological adjustment, and behavioral problems. Children not living with both biological parents show an average disadvantage of about 0.15 to 0.20 standard deviations across outcomes.
Sounds small. Isn't.
A 0.15 SD shift is not going to do much to the average kid. But look at what it does at the tails of the distribution, which is where the damage actually shows up. More children cross the line into clinical levels of behavioral problems. Fewer of them finish college. More of them end up in the criminal justice system. And when you multiply that across millions of children and then compound it over generations, these "small" effects are not small at all. They add up to something enormous.
Causation is a real concern here and it deserves more than a hand-wave. Parents who end up divorced are different from the ones who stay married in all sorts of ways. Personality, mental health, how much money they have, how good they are at being in a relationship. Some of what we see in the children is probably just those pre-existing differences showing through (what researchers call the selection effect) rather than the family disruption itself doing the damage (the causal effect).
The best study designs we have for disentangling this, the ones that compare siblings with different exposure to instability or use genetically-informed designs, suggest roughly 40 to 60% of the observed association is selection. Which leaves 40 to 60% that appears to be genuine causal harm from the instability itself. Fixed-effects models, the kind where you compare a child to themselves before and after a family transition, give similar numbers after controlling for everything that doesn't change over time.
So it's not all selection. The instability itself does damage. Modest at the individual level, maybe. But 40% of American children experience some form of family instability, and modest individual harm multiplied by 40% of the population is massive aggregate harm.
What is it that actually does the harm, when you look at the mechanisms? There's less money in the house after a split. There's less supervision when one person is trying to do everything on their own. Conflict between the parents, both while the marriage is falling apart and after the split, is one of the strongest predictors of how badly the kids turn out. And then there's the modeling. Children learn what a relationship is supposed to look like from what they see at home. And what a lot of them see is that relationships are something that ends, that the people who were supposed to love each other can end up being the ones who hurt each other the most.
And then the loop closes on itself. Kids who grew up in unstable families are more likely to go on and form unstable ones of their own. The pattern gets passed down from one generation to the next. Each cohort enters the mating market with a slightly worse idea of what a partnership is even supposed to look like, and so they end up reproducing the same dysfunction they grew up with. It ratchets.
Fertility Collapse
Everything I've described feeds into this.
Marriage declines, and most births still happen within marriages, so fewer marriages means fewer children. Marriage keeps getting pushed later, and biology doesn't care about your career timeline, so you lose fertility to delay alone. Housing absorbs more and more of what people earn, and you can't raise kids in a studio apartment, so cost squeezes out births. Education extends into the late twenties and career establishment eats the early thirties, and the realistic window for starting a family shrinks to something that feels absurdly narrow.
The numbers are just the numbers. US total fertility rate: 3.7 at the baby boom peak in 1957. Hovered around 1.8 from the mid-1970s through the 2000s. Dropped to 1.62 in 2023 and is still falling. Median age at first marriage: 20.3 for women in 1950, now 28.6. Among women 40 to 44, the share who never married went from under 5% in 1980 to over 15% now. Childlessness in that same age bracket roughly doubled in the same period.
The international picture makes the pattern even clearer. South Korea recorded a TFR of 0.72 in 2023, the lowest ever measured for a country not in the middle of a war or a famine. Japan, Germany, Italy, Spain, most of Eastern Europe, all far below replacement. Every developed country on earth is failing to produce enough children to survive, let alone increase.
The exceptions are where it gets interesting. Israel sustains about 3.0 TFR. Even secular Israeli Jews reproduce at or near replacement, which is something you basically cannot say about any other secular population in a developed economy anywhere in the world. What they have that others don't is the whole package at once: a genuine feeling that your nation matters, strong community structures, institutions that actively support people who have kids, and a culture where having children still feels like you're participating in something that matters beyond your own life. Meaning and community and institutions all working together. The lesson, I think, is that this was never really about technology or about economics by themselves, but it's about social conditions. The societies that kept the scaffolding in place reproduce. The ones that pulled it out don't.
The Economic Counterargument
The strongest pushback on everything above is simple: people would have families if they could afford to.
There's evidence. Fertility drops as housing costs rise, and you can see the relationship across metro areas. The 2008 financial crisis produced an immediate, measurable dip in births. Fertility is somewhat higher among the wealthy. Hungary's pro-natalist policies did produce a modest but real increase in births. You can't dismiss this.
The economic constraints are real. Young people buried in student debt, locked out of the housing market, working unstable jobs, they aren't going to find starting a family easy. Obviously true.
But the pure economics explanation has holes in it that are hard to paper over. Fertility went down during the prosperous 1990s, and it wasn't just something that happened during recessions. Israel maintains 3.0 TFR even though housing there is expensive and the economy is intensely competitive. The relationship between income and fertility in the US turns out to be J-shaped, with fertility higher at the bottom, lower in the middle, and then somewhat higher again at the top, which is a weird pattern if affordability is the thing that's really driving this. And then there's the survey data, which consistently shows a gap between how many children people say they want and how many they actually end up having. That gap points to something other than money.
The honest answer, as far as I can tell, is that economics is necessary but it's not the whole thing. You have to be able to afford to have children. You also have to want them badly enough that you're willing to sacrifice for them. And you have to find someone to have them with who is willing to do the same thing. All three of those conditions have to be met at the same time. And fixing only the money part is not going to do much if the other two are still broken, which right now they are.
There's another objection that comes up a lot: why should any of this matter? If people are freely choosing not to have children, isn't that just their preference and shouldn't we respect it?
Of course. Nobody should be coerced into having kids. But is what we're actually observing really free choice?
Most people claim to want to have 2-3 children. The actual number is 1.62. That difference between what people say they want and what they actually end up with looks a lot more like constrained choice than revealed preference.
And then, at the civilizational scale, populations that don't reproduce get replaced by the ones that do. That's just obvious. It doesn't wait for your permission and it doesn't care about your reasons.
If some rich, secular, and individualist society with a strong safety net saw fertility recover to replacement without any cultural shift, i'd be willing to change my mind. If surveys showed people actually wanted fewer children than they're having, that would prove the preference data is wrong. If family structure stopped predicting child outcomes across multiple studies, that would undermine the urgency. But none of these have happened yet.
What to Do About It
Figure out whether you want children and by when. This sounds obvious but most people don't actually do it. Female fertility drops after 35, male fertility drops more gradually. Biology doesn't negotiate. Work backward from biology rather than forward from career milestones. Hint: you should want them, and as soon as possible.
Who you partner with is probably the most consequential decision you will ever make. Take it more seriously than anything else you do in life. Character, shared values, whether they actually want a family and aren't just saying so, (mental) health, intelligence, these are the things that predict whether the thing lasts. Physical attraction and chemistry matter but they predict almost nothing about long-term stability. Evaluate for the twenty-year version of the person, not the Friday-night version.
Get off the apps. They are not designed to help you find a partner. They are designed to monetize your loneliness. Look for people through communities, activities, networks, places where you see the same people repeatedly and where being a terrible person carries reputational cost. The kind of context where you can't disappear from.
Plan the economics on purpose. Housing is the single biggest material barrier to family formation. Career paths that front-load income help. Geographic arbitrage helps. Two incomes help, though they create coordination problems. Build the economic foundation deliberately, because the drift-and-hope strategy produces people who wake up at 37 wondering why they can't afford a second bedroom.
Do not wait for perfect conditions. They will never arrive. There's always going to be some problem. The perfect partner doesn't exist. The perfect moment definitely doesn't exist. At some point committing to a good enough situation beats optimizing forever. Your great-grandparents understood this. We unlearned it somewhere.
If you have children, make stability the non-negotiable. The evidence is clear: instability harms kids more than any specific family configuration. If you can't stay together, keep the conflict low and the transitions minimal. What your kid becomes depends on what you do day after day, year after year, not on what your intentions were.
Build community around family. People who are isolated struggle to form and maintain families. People embedded in something, a church, a neighborhood, a group of friends who are all doing the same thing, have support structures, role models, built-in childcare, potential partners. Community is infrastructure for family formation and most modern people are running a deficit.
Accept that the old scaffolding is gone and it isn't coming back. The constraints that pushed your grandparents toward family formation have been removed, and no political movement is going to reassemble them. What they got by default, you have to build deliberately. That's harder. It's also the situation.
The bundle came apart. Sex, marriage, children. Linked for hundreds of thousands of years of human history, now available separately or not at all. And this was called progress.
What we got produces fewer marriages, more instability, fewer children, and the people who manage to build families anyway do it through a combination of unusual clarity, deliberate effort, and honestly a fair amount of luck. Everyone else drifts. Fertile years close. Options narrow. And then they're just... gone.
I keep coming back to Israel. The sense of purpose, the community, the institutions that actually help you if you have kids, the feeling that it all means something. But I also keep thinking about the fact that nobody designed their system so that other countries could copy it. It grew out of a very specific set of historical conditions. Whether any of it can be transplanted is something I genuinely don't know.
Civilizations that don't reproduce get inherited by the ones that do. And scaffolding that's been pulled out doesn't rebuild itself.
So what do you build?
