In 1975, the most common way to be a 25-to-34-year-old American was to have already done four things: moved out of your parents' house, gone to work, married, and had a child. Not some of them, all of them, in a bundle. The Census Bureau counted 45 percent of that age group who had checked every box. It was the single largest group. That is what the ordinary shape of a young adult life looked like fifty years ago: the whole set, assembled early.
In 2024, fewer than a quarter had done the same. The bundle came apart. The Census's own headline, from a working paper by Paul Hemez and Jonathan Vespa, is that the four-milestone pathway stopped being how most people arrive at adulthood. The most common combination now, describing about 28 percent of young adults, is the stripped-down one: in the labor force, living independently, not married, no kids. A job and a lease and nothing else yet.
Read too fast, that is a story about people giving up on marriage and children. It is not. It is a story about order. The milestones did not disappear; they got resequenced, and the family ones went to the back of the line.
Marriage went from first brick to last
The sociologist Andrew Cherlin gave this its name two decades ago, and it has held up better than most things sociologists name. In a 2004 paper he argued that marriage had gone, in his words, from "a marker of conformity to a marker of prestige," and he called the new form the capstone: something you do to announce that you have already made it, not something you do to start making it. Five years later, in his book The Marriage-Go-Round, he put the image in plain brick. Marriage, he wrote, "used to be the first step into adulthood, but now it is the last," the "capstone of personal life," the "final brick put in after all the others are in place." The older arrangement, the one 1975 was still running on, is what researchers call the cornerstone model: you married young and laid the rest of an adult life on top of the marriage, together, from more or less nothing. Now the marriage is the last brick, set only after the education, the job, the savings, often the apartment are already standing.
The numbers underneath are not subtle. Between 2005 and 2023, the working paper finds, the share of 25-to-34-year-olds who had ever married fell from 62 percent to 44 percent, and the share living with a child fell from 55 percent to 39 percent. Over the same years the economic markers went the other way: the share who had finished their education rose nine points, to 83 percent, and labor-force participation became the single most common thing a young adult had done. The median age at first marriage is now 30.2 for men and 28.6 for women, the oldest on record and seven to eight years later than the 1960 figures of 22.8 and 20.3. The family bricks are still being laid. They are just being laid later, and increasingly outside the window the Census happens to be looking through.
That last point is the caveat I would staple to every alarmed headline this data produces. "Not married by thirty-four" is not "never married." The paper measures a ten-year slice of a life, not the life. When the median wedding now happens at thirty, a survey of people up to thirty-four is going to miss a great deal of marrying that is about to occur just past its own edge. Hemez and Vespa say so plainly: these declines "do not necessarily mean young adults will never marry or raise children," and many still will. A snapshot is not a film. What the snapshot does show, honestly, is that family formation is "more often foregone during emerging adulthood than just 18 years ago." Foregone in your twenties. Deferred, not deleted.
The apartment you cannot buy is the wedding you delay
So the interesting question is not whether young people want marriage. Most still say they do. The question is what they decided has to be true before they will allow themselves to have it, and here the story stops being about values and becomes about money, which is usually where these stories were the whole time. The capstone model has a price of admission. If marriage is the thing you do after you are economically secure, then the bar to marriage is wherever economic security now sits, and that bar has been moving up faster than young incomes can follow. The demographers Christina Gibson-Davis, Anna Gassman-Pines, and Rebecca Lehrman gave it a name: clearing the economic bar to marriage, a bundle of financial achievements couples decide they have to reach before they feel ready to wed. You wait until you clear it. For a lot of people the bar is now set at a level they may not clear during the years the Census is counting.
You can see the bar in the Census's own machinery. When Hemez and Vespa took apart what drove the decline, one of the largest single factors was housing. In states where more households were stretched by the cost of owning a home, young adults were measurably less likely to have reached the family milestones, and more likely to be parked in the economic-only bucket instead. Their conclusion is unadorned: "housing costs appear to reduce young adults' chances of achieving family formation milestones." The apartment you cannot afford to buy is, statistically, the wedding you are not having yet. This is the part the culture-war version of the story leaves out. Nobody in these figures is refusing marriage on principle. They are waiting for a foundation solid enough to build on, and the foundation keeps getting more expensive.
The capstone the poor cannot reach
And because it runs on money, it sorts by class, which is the part I cannot leave alone. If marriage has become a capstone, a thing you earn, then it follows that the people who can afford the earning get the capstone and the people who cannot do not. The data obliges. The first-marriage rate for Americans with a master's degree or more is more than double the rate for those with a high school diploma. Marriage, the institution we still narrate as the most personal and least economic choice a person makes, is quietly becoming a marker of who was already secure. The prestige Cherlin described is real, and prestige is not distributed evenly. Cornerstone marriage was something the poor and the comfortable both did, at roughly the same age, because it cost nothing to start. Capstone marriage costs a stable job, a cleared bar, a foundation to build on. Not everyone gets a foundation.
There is one function of the old cornerstone that the capstone keeps, and it is worth saying because it complicates the neat story. Cherlin called it enforceable trust: a marriage, unlike a private arrangement, is a public promise made in front of your family and the state, and that public quality lowers the risk of the other person walking away, which lets both people invest in things that only pay off over decades. A house. A child. A pension survivor benefit. That is not sentiment; it is a mechanism, and it is precisely the mechanism that becomes hardest to justify signing up for when you are not sure you will still have the job in three years. The people who most need the long-term security of enforceable trust are the ones whose economic lives are too uncertain to risk the long term at all. So the tool designed to help you build across decades goes to the people who already have decades of stability to build across.
I read census tables for pleasure, which is its own small confession, and the thing these ones keep telling me is that a milestone is only a choice for the people who can afford to reach it. We have taken an event that used to sit at the front of adult life, cheap and early and available to almost everyone, and moved it to the back, behind a wall of accumulated economic requirements, and then we describe the people who cannot get over the wall as having made a lifestyle decision. The order changed. The question that stays open is whether we built a sequence most people can actually walk, or just a nicer name for who gets left standing at the first brick.
