§ Claim under review · Mixed
"Married couples with different last names have a 50% higher divorce rate than couples who share a surname, according to an Institute for Family Studies analysis. Couples with different surnames divorced after an average of eight to 10 years, compared with about 12 years for couples who shared a name." Image headline: "MARRIED COUPLES WITH DIFFERENT LAST NAMES DIVORCE 50% MORE"
Verdict
Partially accurate but misleading
Confidence
HighSummary
The 50% figure is real and does come from an Institute for Family Studies analysis by Lyman Stone, and the 8-to-10-years versus 12-years comparison is quoted correctly. But the analysis is not a measured divorce rate. The author states outright that the data gave him no way to calculate a real divorce rate by surname, so he estimated one by comparing Texas divorce filings from 2016 and 2017 against a 2023 national Pew survey of married couples. It is one state, two years, roughly a decade old, published as a think-tank blog post rather than a peer-reviewed study, with no adjustment for age at marriage, income, religiosity, or how recently couples married. The author himself says part of the gap is certainly not causal and that requiring shared surnames would probably not save many marriages. His own stress test also shows the gap nearly vanishes for Asian couples and reverses for Hispanic couples, which the single pooled number hides. Treat it as a suggestive correlation from limited data, not as evidence that keeping your name raises your risk of divorce.
The readings
key figures from the evidencehigher divorce rate for couples with different surnames
marriage duration before divorce, different surnames
longer marriage duration when wife changed surname (Canadian study)
Why this verdict
Evidence
The Institute for Family Studies analysis exists and it does contain the 50% figure. Its own summary line states that spouses who do not share a surname divorce at about a 50% higher rate than those who do, and that their divorces come about 30% earlier in their marriages. On duration, the analysis states that couples who share a name stay together about 12 years before divorce, while couples with other naming practices tend to make it 8 to 10 years. Both numbers in the post therefore trace correctly to the named source.
The critical detail is how the 50% was produced. The analysis states plainly that Texas's marriage index records bride and groom names before marriage and says nothing about post-marital surname choice, so, in its own words, there is no denominator available to create an actual divorce rate by surname. Instead the author builds what he calls a loose approximation: assume Texas resembles the national average, where roughly 75 to 80% of married couples share a surname (77% in the Pew microdata, 79% in Pew's published estimate), and compare that against the finding that only 64% of Texas divorces involved a shared surname. The gap between those two percentages is the basis of the inferred higher divorce rate. It is not a measured divorce rate for either group.
The author also states directly that the evidence does not definitively answer whether sharing a surname saves a marriage, that some of the divorce difference is certain to be due to non-causal correlations, and that simply imposing shared surnames on couples probably would not save many marriages. The causal mechanisms he offers, including the administrative hassle of reverting a legal name and the idea that name-sharing signals a "team-first" orientation, are presented as hypotheses, one of which he labels speculative.
Findings
✓ What's accurate 5
- The Institute for Family Studies analysis is real, is authored by Lyman Stone, and does state an approximately 50% higher divorce rate for spouses who do not share a surname.
- The 8-to-10-years versus about-12-years duration comparison is accurately reproduced from the source.
- The analysis does draw on 2023 Pew Research Center data and Texas divorce records, as the caption says.
- The post attributes the figure to IFS rather than presenting it as unsourced.
- An association between surname sharing and marital duration is not novel; a small peer-reviewed Canadian study pointed in the same direction.
≈ What's misleading 6
- Omitted qualifier: the post says couples "have a 50% higher divorce rate." The source says no actual divorce rate by surname could be calculated and that the figure is a loose approximation built by comparing one state's divorce records against a national survey. Presenting an inferred ratio as a measured rate materially overstates what was computed.
- Unsupported causal inference: the image headline "MARRIED COUPLES WITH DIFFERENT LAST NAMES DIVORCE 50% MORE" strips attribution and hedging and reads as an established causal fact. The author himself says some of the difference is certainly non-causal and that the evidence does not definitively answer the causal question.
- Scope and date mismatch: nothing in the post signals that the divorce data is Texas-only and from 2016 and 2017, or that the comparison denominator comes from a different population measured in 2023. A reasonable reader would take this as a current, nationwide finding.
- Subgroup concealment: the pooled 50% masks a wide spread across racial and ethnic groups, including a near-zero gap for Asian couples and a reversed direction for Hispanic couples under the author's own sensitivity scenario.
- Marketing of authority: aggregators upgraded a non-peer-reviewed blog analysis into a "new study," and the Polymarket-branded "TRENDING" treatment adds a veneer of institutional data reporting that the underlying evidence does not carry.
- Duration figure slightly broadened: the source's 8-to-10-year range covers couples with "other naming practices," which includes hyphenated and partially shared names, not solely couples with entirely different surnames.
? What's uncertain 5
- The exact arithmetic producing "about 50%" is not spelled out in the public post, and the New York Post reported the group-specific rates were not clear from the report. It cannot be independently reproduced from what is published.
- Whether the estimate would survive adjustment for age at marriage, marriage cohort, education, religiosity, or income. This is not reported and is the single largest open question, since non-shared-surname couples are likely to have married more recently and recent marriages sit in the higher-risk early years.
- Whether Texas in 2016 and 2017 actually resembled the 2023 national surname distribution, which is the load-bearing assumption of the whole estimate.
- Whether the finding replicates outside Texas. No independent replication using a different state or national dataset was found.
- Whether any causal component exists. The bureaucratic-friction and commitment-signal mechanisms are hypotheses, and the author describes the first as speculative.
Sources
5 of 8 linked to recordsInstitute for Family Studies blog analysis by Lyman Stone, "Can Sharing a Last Name Save Your Marriage? It Depends" / "His or Her Last Name: Shared Surnames and Divorce Risk"
Lyman Stone, "Can Sharing a Surname Save Your Marriage?" (Substack, July 2026)
Texas Department of State Health Services marriage/divorce indexes (2016, 2017 files)
Pew Research Center 2023 survey on surname choices (79% of women take husband's name)
MacEacheron, M. (2021), "Marital Surname Change and Marital Duration Among Divorcées in a Canadian County," Frontiers in Psychology
New York Post / Yahoo Lifestyle coverage, mid-September 2026
Not the Bee (Sep 17, 2026) and Conservative Institute (Sep 17, 2026)
Polymarket-branded "TRENDING" graphic in the post itself