Within two generations, the path into marriage in most rich democracies has been quietly rerouted. Where the median couple in 1970 went from dating to engagement to marriage to shared address, the median couple today goes from dating to shared address to either drift, breakup, or marriage — usually after several years of cohabitation. The shared apartment, not the wedding, is now the threshold event. The wedding, when it happens, often ratifies a household that has already existed for years and may already include children. This is a collective revision of the marital pipeline so thorough that the older sequence now reads as exotic in most peer groups, even as the rhetoric of marriage continues to be written as if the older sequence were still standard.

Sharon Sassler's longitudinal interview work with cohabiting couples in the United States documents what the demographic curves only outline. Couples slide into cohabitation more often than they decide on it — one toothbrush left over, a lease renewed jointly, a pet acquired, a winter survived. The transition is rarely marked by an explicit conversation about what cohabitation means or what would have to be true for it to lead to marriage. Sassler calls this "sliding rather than deciding," and her data suggests it produces a particular kind of marriage when marriage does follow: a marriage in which the constitutive choice was never actually made aloud, only ratified retrospectively. Such marriages function, often well, but they begin from a different starting point than marriages preceded by a deliberate transition.

Wendy Manning's research adds the class dimension. Cohabitation in the United States is now nearly universal as a prelude to marriage among the college-educated, but for them it tends to be a relatively short stepping-stone — two to three years, often with an engagement during the cohabiting period — before a marriage that is broadly stable. For couples without college degrees, cohabitation has become an alternative to marriage rather than a step toward it: longer durations, higher rates of children born within the cohabitation, lower rates of eventual marriage, higher rates of dissolution. The same household form, the cohabiting couple with children, plays different roles in different social classes — pipeline for some, terminus for others. Phil Cohen's demographic work makes the same point at the population level: the marriage premium has not disappeared, but access to it has become increasingly stratified.

The European data complicates the American story. In the Nordic countries and France, long-term cohabitation with children is a stable institutional form supported by legal frameworks (the French PACS, Swedish sambo arrangements) that grant most of the rights of marriage without the ceremony. Cohabitation there is not a pipeline to marriage; it is a parallel track that many couples never leave, and the children of these unions show outcomes statistically similar to those of married couples. The pipeline framing is partly an American artifact — a function of the United States' weak legal recognition of cohabiting partnerships, which leaves marriage as the only way to secure most of the benefits of joint life. Where the legal asymmetry between cohabiting and married couples is small, the pipeline structure dissolves. Where it is large, the pipeline persists.

The collective revision Law 5 names is the transition from marriage-as-entry-point to marriage-as-ratification. The wedding has shifted from the moment a couple's joint life begins to the moment a joint life that already exists is publicly named. This has consequences for what the wedding can mean. A wedding that begins a household carries information — vows about what the couple is committing to build. A wedding that ratifies a household carries different information — vows about what the couple is committing to continue. Neither is hollow; they are simply different speech acts. The cultural script for weddings, however, still leans on the older grammar, which is one reason contemporary weddings often feel ceremonially overcooked relative to the actual change they mark.

There is also a quieter revision underway in the meaning of the cohabitation itself. For many couples, the cohabiting years are now the most consequential years of the relationship — the period in which compatibility is actually tested, household economics are negotiated, fertility decisions are made, and the question of whether to marry is decided. The wedding follows. By the time it arrives, the costly information has already been exchanged. The romantic implication is significant: the cohabiting period has absorbed much of the function the engagement once performed, and the engagement has shrunk to a ceremonial marker between two phases of an already-shared life. Couples who skip the cohabitation step and marry directly are now the statistical anomaly, and the assumptions built into older marriage advice — that couples discover each other primarily after the wedding — no longer fit the modal case.

The pipeline metaphor is finally useful and misleading. Useful because it captures the sequential structure of contemporary partnering: dating, cohabitation, possible marriage, possible parenthood. Misleading because pipelines imply forward flow, and the contemporary pipeline includes substantial reverse flow — cohabitations that dissolve, marriages preceded by multiple prior cohabitations with other partners, second-marriage cohabitations that look nothing like first-marriage ones. The honest description is less a pipeline than a sequence of optional thresholds, each of which a given couple may or may not cross, in an order that has loosened considerably from the postwar template. The collective task is to develop language and law adequate to the looser structure.