Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
How do divorce rates vary by age group and education level in 2025?
Executive summary
Available reporting shows that divorce risk falls sharply with older age at marriage and with higher education: marrying after the mid‑20s is repeatedly linked to substantially lower divorce risk, and college graduates are commonly reported as roughly 25–30% less likely to divorce than people with only a high‑school education [1] [2] [3]. At the same time, “gray divorce” (age 50+) has risen in recent decades even as overall crude divorce rates have declined to about 2.3–2.5 per 1,000 in some recent 2020–2025 estimates [4] [3] [5].
1. Age at marriage: the clearest pattern — wait and your risk drops
Multiple pieces summarize the dominant, consistent finding: people who marry very young face higher divorce risk, while marrying later (mid‑ to late‑20s and beyond) reduces that risk. One summary cites that marrying before 18 increases the 10‑year divorce risk substantially, while waiting beyond age 25 reduces likelihood by roughly a quarter [1] [3]. Several consumer and legal overviews likewise report that couples who marry between about 25–32 have higher chances of long‑lasting marriages compared with those who marry earlier, and that risk increases again only slowly after the early‑30s in some accounts [2] [6]. The National Center for Family & Marriage Research profile also notes that first‑divorce rates “decline precipitously by age” even as age patterns have changed over time [7].
2. Education: college is repeatedly linked with greater marital stability
Across sources, higher educational attainment—particularly a bachelor’s degree or more—is associated with materially lower divorce rates. Several sites quantify that difference: college attendance or a college degree is linked to a ~13%–30% reduction in divorce risk in different summaries, while other reporting gives absolute comparisons such as high‑school‑only divorce rates near the mid‑40s percent versus mid‑20s percent for college graduates [1] [2] [3]. The Bureau of Labor Statistics analysis of NLSY79 data shows stark lifetime differences by education: by age 55, people with a bachelor’s degree were much less likely to have divorced from a first marriage than those with lower schooling [8].
3. Older adults and “gray divorce”: a notable exception to the overall decline
Although overall divorce measures have fallen, divorce among older adults has increased historically and remains a point of attention. Reporting highlights that gray divorce rose from 1990 to the 2000s and then leveled or stagnated recently, but overall counts and rates among age 50+ are much higher than decades prior [9] [10] [7]. The NCFMR profile and other summaries emphasize that younger adult divorce rates have declined while rates among older adults rose, producing a less age‑graded pattern than in previous generations [7].
4. How big are the overall divorce‑rate numbers in 2025?
Different outlets use different measures (crude, refined, percent ever divorced), producing varying headline figures. Several recent summaries place the crude divorce rate around 2.3–2.5 divorces per 1,000 people for the early‑to‑mid‑2020s and note refined measures around the mid‑teens per 1,000 married women in some datasets; others translate lifetime risk into percentages (e.g., roughly 36–41% of first marriages ending within a decade in some estimates) [4] [3] [5]. Available sources do not converge on a single 2025 number because of methodological differences across providers [5] [4].
5. Caveats, disagreements and methodological pitfalls
Reporting varies by data source, timeframe, and metric: CDC/NCHS vital‑statistics crude rates differ from refined rates based on married‑woman denominators; survey‑based “ever divorced” measures produce lifetime percentages that are not directly comparable to annual rates [5] [4]. Some commercial summaries present precise percent reductions (e.g., “30% less likely” for college grads) that differ in magnitude from academic estimates [1] [11]. Sources also mix national trends with state‑level or occupation breakdowns; those localized claims (for example, headline state lists or occupation rankings) are not consistently supported by the same public datasets cited elsewhere [12] [1].
6. What this means for interpreting “divorce rates by age and education” in 2025
The evidence consistently supports two robust takeaways: marrying later (after the early‑ to mid‑20s) and higher education are each associated with markedly lower divorce risk [1] [2] [8]. However, how large those gaps look depends on the metric used (annual crude rate, refined rate per married population, or lifetime probability), and older adults buck the long‑term downward trend by showing elevated divorce incidence relative to past generations [7] [3]. If you need a precise table for a specific age bins and education groups for 2025, available sources do not provide a single harmonized breakdown — you would need to pick one authoritative dataset (e.g., NCFMR/ACS or CDC/NCHS) and extract its age‑by‑education cross‑tabs [7] [5].
If you want, I can pull specific tables from any one of the cited datasets (NCFMR/ACS, CDC/NCHS, or BLS/NLSY summaries) and format age‑and‑education breakdowns using that source’s methodology — tell me which dataset you prefer [7] [5] [8].