Age Gaps and Marriage Timing: What the Data Actually Shows

Americans Marry Later Than Ever — Especially Where You Live

Start with the timing side, where the data is cleanest. The U.S. Census Bureau's historical marital status tables show the median age at first marriage in 2024 at roughly 30.2 for men and 28.6 for women. For perspective: in the 1950s those figures were about 22.8 and 20.3. The entire concept of a "late" marriage has moved by nearly a decade in two generations, while the roughly two-year gender gap has stayed almost constant.

The geographic spread matters for anyone dating in New York specifically. Analyses of 2024 data by Bowling Green's National Center for Family & Marriage Research put the highest male medians in Connecticut, Massachusetts, and New Hampshire (around 32.4) and the highest female median in Washington, D.C. (32.3), with Utah lowest (25.2 for women). The urban Northeast marries latest in the country — so an NYC professional who's 33 and unmarried isn't behind the local curve; they're on it. That's worth holding against the ambient panic covered in dating in your 30s vs. your 20s: the 30s dating pool is large partly because the median itself has moved there.

What the Age-Gap Study Actually Found

The most-quoted evidence on age gaps comes from Emory economists Andrew Francis-Tan and Hugo Mialon's study "'A Diamond is Forever' and Other Fairy Tales," published in Economic Inquiry in 2015 — a survey of 3,151 ever-married U.S. adults, analyzed with survival models predicting divorce.

On age difference specifically: each additional year of spousal age gap was associated with a modestly higher divorce hazard — hazard ratios around 1.02 per year in the full sample and about 1.03 in the recently-married subsample. The viral version of this finding — "a 5-year gap means 18% higher divorce risk, a 10-year gap 39%" — is that per-year estimate compounded across years. The arithmetic is fair; what gets lost is the caveats: it's an association in one retrospective sample, the per-year effect is small, and couples with large gaps differ from other couples in many ways the models can only partly control. The honest summary is that larger gaps correlate with somewhat higher divorce rates at the population level, and that a modest gap moves the needle far less than the headline suggests.

The Findings From the Same Study Nobody Quotes

The age-gap number became the headline, but it wasn't close to the study's strongest result. From the same models:

• Courtship length dominated. Couples who dated three or more years before the proposal showed roughly 40–50% lower divorce hazard than couples who dated less than a year. Knowing a spouse "very well" before marriage was similarly protective. Time-before-commitment was among the largest effects in the paper — worth holding next to how many dates finding a relationship actually takes.
• Cheap weddings beat expensive ones. Spending under $1,000 on the wedding was associated with lower divorce hazard; $20,000+ weddings were associated with substantially higher hazard among women in the sample. The authors traced a plausible mechanism directly: bigger wedding spending predicted much higher odds of reporting wedding-debt stress in the marriage.
• Attendance mattered more than budget. High wedding attendance was strongly associated with lower divorce hazard — couples marrying in front of 200+ people showed a fraction of the hazard of couples who eloped alone. Public commitment, not production value, was the protective ingredient. Honeymoons helped too, regardless of cost.
• Marrying for looks or money backfired, measurably. Respondents who said their partner's looks were important in the decision to marry showed elevated divorce hazard; same for partner's wealth. The unglamorous compatibility variables — the attraction vs. compatibility distinction, in effect — carried the durability.

What This Means Practically

Read together, the timing data and the age-gap study point somewhere consistent: the variables that predict lasting marriages are mostly about information — how long and how well you knew each other before committing — and fit, not about hitting an age, matching birth years, or staging the right wedding. A modest age gap is statistical noise next to a rushed courtship; marrying at 33 in New York is the local median, not a delay. For anyone dating deliberately, the actionable version is unexciting and well-supported: prioritize the compatibility variables you can actually screen for early — the direction-of-life questions — and give the knowing-each-other part the time the data says it's worth.

Sources:

• Median age at first marriage, 1890–present — U.S. Census Bureau historical tables
• Median Age at First Marriage: Geographic Variation, 2024 — National Center for Family & Marriage Research, Bowling Green State University
• Francis-Tan, A., & Mialon, H. (2015). "'A Diamond is Forever' and Other Fairy Tales: The Relationship between Wedding Expenses and Marriage Duration." Economic Inquiry, 53(4), 1919–1930.