The single most consequential fact about contemporary American romance is that marriage no longer happens evenly across the population. It has retreated to the top third. Among Americans with a four-year college degree, marriage rates remain high, divorce has fallen back near 1970 levels, and over 80 percent of children are born to married parents. Among Americans without a four-year degree, marriage has collapsed, cohabitation has proliferated, churn has replaced stability, and the majority of children are born outside marriage. These are not two ends of a smooth distribution. They are two systems, increasingly disconnected, each producing different outcomes for children, different patterns of adult well-being, and different political coalitions.
Charles Murray's Coming Apart and Sara McLanahan's "diverging destinies" research, approaching the data from opposite political starting points, arrived at the same descriptive map. The upper-middle class, broadly defined, lives in a world where marriage is still load-bearing: people meet through education and work networks, partner with someone of similar background, marry in their late twenties or early thirties, raise children together, and divorce relatively rarely. The working class lives in a world where the marriage script has decomposed: people pair up younger, often cohabit, often have children, often split, often re-partner. The middle—however defined—is being absorbed into the working-class pattern faster than the upper.
Why? The answers cluster around three mechanisms. First, economic. The collapse of stable, well-paying, non-college male jobs—manufacturing, building trades at scale, mid-tier clerical work—removed the material foundation on which working-class marriage rested. Andrew Cherlin's Labor's Love Lost makes this case in detail: working-class marriage was historically held together by a male-earner platform that no longer exists at scale. When the platform collapsed, the pair bond collapsed with it. Second, cultural. The upper-middle class internalized a delayed-marriage, capstone model that fits its economic timeline (long education, delayed earnings peak). The working class adopted the delayed-marriage piece but without the educational and economic platform that makes the capstone model work. They got the cultural script without the structural conditions. Third, institutional. The norms, kin networks, churches, and civic organizations that historically held working-class marriages together have weakened—not vanished, but thinned—while their upper-middle-class equivalents (still-intact networks, expensive marriage rituals, family-of-origin investment) have held.
Edin's interviews with low-income mothers provide the human texture. The women want marriage. They view it as sacred. They will not enter it lightly, and they will not enter it with men they consider unreliable. The men in their lives—shaped by labor-market collapse, mass incarceration, and substance crises—often fail their reliability tests. So the women have children with these men, raise the children largely alone, and wait for marriage as a future achievement. The marriage often never arrives. The men, meanwhile, want children, want fatherhood, and want to be the kind of provider their grandfathers were—but the economic conditions for that role have evaporated. The result is mutual failure to coordinate on a form everyone says they want.
Putnam's Our Kids documents the downstream effect on children. In college-educated households, children are raised in two-parent homes with high cognitive investment, dense extracurricular networks, and consistent supervision. In non-college households, children are raised in unstable households with rotating adults, lower cognitive investment, weaker networks. The gap is not new—class always shaped childhood—but the gap has widened sharply since 1980. Putnam frames it as a collapse of cross-class community: the kid networks, civic organizations, and informal mentoring that used to spill across class lines have retreated into class-specific bubbles. The marriage gap is downstream of, and reinforces, this larger social-capital gap.
Law 3 (Connect) is the law most visibly broken in this data. The pair bond, historically the most universal human connection, has become a class marker. What used to be a default has become a credential. The connection capacity of the bottom two-thirds of the population—measured by stable partnerships, intact families, dense kin networks—is in serious decline. The connection capacity of the top third has held, in some ways improved, and increasingly exists inside a sealed-off social world.
Law 2 (Think) requires honest framing. Both the political left and right tell self-serving stories about this data. The left emphasizes economics—if wages and jobs returned, marriage would return—and downplays cultural drift. The right emphasizes culture—if values returned, marriage would return—and downplays economic structure. The data shows both are partially right and neither is sufficient. Economic conditions are necessary but not alone determinative; the upper class kept marriage partly through cultural commitments that its economic position allowed. Cultural conditions are necessary but not alone determinative; without the material platform, cultural commitments don't hold.
Law 4 (Plan) is violated by the policy environment. Tax codes, housing policy, childcare policy, and labor policy are still designed around an assumed standard household that exists primarily in the upper third. The bottom two-thirds navigate an institutional environment that does not match their actual family structures, while the upper third receives policies optimized for their form. The mismatch compounds the gap.
Law 0 (Humility) cautions against confident solutions. Many proposed interventions—marriage promotion programs, fatherhood initiatives, wage subsidies, expanded child credits—have been tried in various forms. Effect sizes are small. The collapse of working-class marriage is not a problem with an obvious lever. It is a system-level reorganization with deep economic, cultural, and institutional roots, and a generation or more of recovery time, if recovery is even the right frame.