College Women Aren't Running Out of Partners. The Squeeze Moved.
A new Opportunity Insights paper finds college-educated women kept marrying by pairing with high-earning men without degrees. The thinner partner pool hits women without college hardest.
The think-piece version of the marriage market is familiar: college women can't find anyone. The quieter finding from a new economics paper is almost the opposite. College-educated women's marriage rates held up. The sharper squeeze landed on women without degrees.
In May 2026, Clara Chambers, Benjamin Goldman, and Joseph Winkelmann released Bachelors Without Bachelor's as NBER Working Paper 35179, through Harvard's Opportunity Insights lab. They tracked cohorts born from 1930 to 1980 using survey data plus linked census and tax records. The headline campus fact is already hard to ignore. About 1.6 million more women than men now attend four-year colleges in the U.S. The ratio of college men to college women fell from roughly 1.8 for the 1930 birth cohort to about 0.8 for the 1980 cohort.
If everyone still preferred to marry within the same education lane, you would expect college women's marriage rates to tumble. They did not tumble much. For college women, marriage by age 45 slipped from about 77.7% to 71.0%. For women without college, it fell from about 78.7% to 52.4%. That gap is the story relatives should hear before they lecture one niece about "standards" and another about "settling."
What the paper actually shows
Marriages where the wife has a four-year degree and the husband does not rose from 2.3% of Americans in the 1930 cohort to 9.6% in the 1980 cohort. Those husbands are often economically strong. At age 45, non-college men married to college women earned about $68,400 on average (2024 dollars), while other non-college men averaged about $46,100. The share of non-college men who earn above the median and are not already married to a college woman fell from about 73% to about 35%.
Family pressure often treats marriage delay as a character flaw. This research says the market moved: college women kept marrying by widening the education match, and women without degrees absorbed more of the shortage. That should change the lecture, not only the headline.
Who felt the shortage
Benjamin Goldman, an economist at Cornell and a research principal at Opportunity Insights, has said the team got curious after reading essays by highly educated women about how impossible dating had become. The data did not match the panic. College women's marriage rates stayed relatively stable because many of them married men without bachelor's degrees who were still doing well economically: small-business owners, skilled technicians, and other high earners without the diploma. That choice has a downstream effect. Women without college degrees historically married men with similar schooling. As the best-earning men in that group paired upward into college households, the remaining pool of "economically stable" non-college partners thinned. Goldman put it plainly in a Harvard write-up: if your goal is a spouse who is relatively financially steady, that person got harder to find if you yourself do not have a degree. Place matters too. Where non-college men face more joblessness and incarceration, marriage rates for non-college women fall further. In the paper's lowest-employment areas, those rates average about 44.5%, versus about 66% in the highest-employment areas. This is less a personality story than a labor-market story with wedding invitations attached.
What diaspora dinner tables do with this
In many Asian and multicultural households, education is treated like a marriage credential. Parents may bless a partner with a good job and no degree, or they may treat the missing diploma as a character flaw. This paper does not settle that fight. It does explain why the fight feels louder now. The campus gender gap is real. So is the sorting by money. Relatives who only track degrees will misread both the college niece who dates a high-earning tradesperson and the cousin without a BA who is dating in a thinner market. A useful dinner sentence: national research finds college women's marriage rates held up by expanding who counts as a match, while women without degrees lost access to many of the most economically stable men in their education group. That is not a license to shame anyone. It is a reason to stop pretending every timeline delay is about "being picky" in the same way. The authors also project that Americans born in the mid-1990s are on track to become the first cohort where fewer than half of non-college women are married by age 45. Class gaps in family formation are not abstract. They shape who raises kids with two incomes and two sets of hands in the house.
What to do with the finding
If you are dating seriously, talk about money and work early without turning the first coffee into a résumé audit. A degree is one signal. Earnings stability, debt, remittances, and whether someone can share a household are others. Our premarital money guide exists for the spreadsheet version of that conversation. If parents are pushing a marriage timeline while also policing education status, separate the questions. Do you want partnership. What does financial readiness mean in this city. Which status rules are about safety, and which are about face. Our guide on parents asking for a marriage timeline helps keep those threads from tangling into one insult. And if someone you love is marrying across the old education line, resist the cheap joke. The data says that pattern is how a large share of college women's marriages still happen. The harder question is how communities support the people left with fewer stable options, not how loudly we police diplomas at the reception.
Keep reading: Dating When Parents Keep Asking When You Will Marry, Premarital Money Talk: Debts, Remittances, and the Spreadsheet You Keep Postponing, When Do Asian Americans Marry? Median Age by Group, and Living Together Before Marriage When Your Family Would Be Ashamed.
Early dating and marriage pressure hub · When Gen Z says marriage can wait