Our sources

We only publish numbers we can trace to a primary source: the original study, official statistics, or the organization's own reporting. Secondary retellings, influencer polls, and unattributed "studies show" claims are rejected at the fact-gate.

Primary sources currently cited

• Pew Research Center (2020)Pew analysis of censuses and surveys conducted between 2010 and 2018 across 130 countries and territories; sentence read directly on the Pew page: 'In the U.S., 27% of adults ages 60 and older live alone, compared with 16% of adults in the 130 countries and territories studied.'
• Pew Research Center (2025)Pew analysis of decennial census and ACS microdata via IPUMS, 2023. Read directly on the Pew page.
• Pew Research Center (2020)Pew American Trends Panel Wave 56, fielded October 16–28, 2019, 4,860 U.S. adults. Sentence read directly on the Pew page: 'three-quarters of those 65 and older are not looking for either a relationship or dates'.
• Pew Research Center (2023)Pew Research Center survey of 6,034 U.S. adults conducted July 5–17, 2022. Sentences read directly: 'About half of online daters ages 50 and older (47%) say they have encountered someone who they thought was trying to scam them. Online daters ages 50 to 64 are more likely than those 65 and older to have experienced this (52% vs. 39%).'
• the National Center for Family and Marriage Research at Bowling Green State University (2024)Loo & Brown (2024), Family Profiles FP-24-04, NCFMR analyses of IPUMS-USA ACS 2022. Both figures read directly in the PDF.
• the National Center for Family and Marriage Research at Bowling Green State University (2024)Loo (2024), Family Profiles FP-24-09; NCFMR analyses of U.S. Vital Statistics 1990 and ACS 1-year estimates 2022. Read directly in the PDF (men 15.0 → 10.6; women 2.0 → 2.9).
• the National Center for Family and Marriage Research at Bowling Green State University (2019)Schweizer, V. (2019). The retreat from remarriage, 1950–2017. Family Profiles, FP-19-17. NCFMR analyses of Vital Statistics of the United States 1950, Monthly Vital Statistics Reports 1960–1990 and U.S. Census Bureau ACS 2008–2017. Page opened and read.
• the U.S. Federal Trade Commission, the FBI's Internet Crime Complaint Center, and the Canadian Anti-Fraud Centre (2026)Verified live on 2026-08-13. ic3.gov opened and read: 'the central hub for reporting cyber-enabled crime', 'It is run by the FBI'. Canadian Anti-Fraud Centre reporting page opened and read (https://antifraudcentre-centreantifraude.ca/report-signalez-eng.htm): online reporting at reportcyberandfraud.canada.ca, toll-free 1-888-495-8501, Monday to Friday 10 am–4:45 pm Eastern; the centre is operated with the RCMP, the Ontario Provincial Police and the Competition Bureau. https://reportfraud.ftc.gov/ responded (HTTP 200, page titled 'ReportFraud.ftc.gov'), but it is a JavaScript application and ftc.gov / consumer.ftc.gov return HTTP 403 to automated fetching, so the FTC route is corroborated rather than read in full — hence medium confidence on the FTC leg alone.

What we rejected, and why

This list matters as much as the one above. These are figures that circulate widely in writing about widowhood and that we could not trace to a primary source, so they do not appear anywhere on this site.

• Share of people 70+ who met their current partner online (content-spec line 213). — No source publishes a 'met current partner online' share on a 70+ or 65+ base. Pew's meeting-place data is not broken out above 50, and Pew's 70+ figure (12%) measures having EVER USED a platform, which is a different claim. Slot stays empty unless the sentence is rewritten to 'ever used', in which case use online-dating-use-70plus.
• Message-response asymmetry by sex on dating platforms at 60+ — e.g. 'men send N times more messages and get a reply X% of the time' (content-spec line 460). — Nothing found with an age breakout at 60+ from a named platform or a peer-reviewed paper. The circulating figures ('men respond to 26% of messages, women 16%', from ~400,000 users of an unnamed app; OkCupid blog reply rates of 60% vs 36% for a 40-year-old woman messaging 25- vs 55-year-old men) come from an unnamed platform or a company blog reported second-hand, and neither describes people over 60. Describe the asymmetry qualitatively and give no ratio.
• Age-band composition of a dating platform's user base — e.g. 'X% of Bumble users are over 60' (content-spec line 531). — Neither Bumble nor Match Group publishes an age-band breakdown of its user base. Every figure in circulation (5.76% aged 65+, 1.1% aged 65+, 'the largest age group on Match.com is 55–64') comes from statistics-aggregator sites that do not state a methodology and contradict each other. Per the brief, this slot closes only if the platform publishes it itself. The page may say only what we observed, with the observation date.
• Share of adults 60+ (rather than 65+) who are not married or cohabiting; single men 60+ and single women 60+ on a clean 60+ base. — Not published. Pew's single/unpartnered series breaks at 65 (36% of adults 65+, 49% of women, 21% of men) and NCFMR's marital-status series is 65+. The 60+ base exists only for living alone (Pew, 27%). Writers must use the 65+ figures and say '65 and older'; a 60+ phrasing is unsourced.
• A general 'partnership-formation rate' or 'odds of finding love' for adults 60+ or 65+. — No such statistic exists. The nearest published measure is the remarriage rate (NCFMR: 5.1 per 1,000 previously married adults 65+, 2022), which excludes cohabitation, excludes every unmarried relationship, and excludes never-married people from its denominator. Approved as a remarriage rate only. Any sentence framing it as odds of finding a partner is a fabrication — this gap is exactly what /love-after-60/ section 2 is supposed to name.
• 'Median age at re-partnering' for men. — Re-partnering (marriage + cohabitation + unmarried relationships) has no published median age. Only median age at REMARRIAGE exists (NCFMR: 46 for men, 43 for women, 2017), approved under that name and no other.
• Share of older adults who are cohabiting rather than married (NCFMR FP-25-08, 'about 6% of couples were cohabiting'). — Population mismatch: NCFMR defines 'older adult couples' there as couples with at least one partner aged 50 or older. A 50+ couple base cannot carry a claim about people in their 60s or 70s on this site.
• FTC romance-scam loss totals ($547 million reported lost in 2021; $1.3 billion and ~70,000 reports in 2022), including any 'people 70+ lost the most' variant. — Two problems. (1) ftc.gov and consumer.ftc.gov return HTTP 403 to automated fetching, so no FTC figure could be read on its primary page — only in search snippets. (2) content-spec §2 bars scare statistics on /red-flags/. The page carries reporting routes, not loss totals.
• National Poll on Healthy Aging / University of Michigan loneliness and dating findings (e.g. 'one in three older adults feel lonely'). — Population is adults aged 50–80, which fails the 60+/65+/70+ bar, and the framing is health-adjacent, which /single-at-60/ explicitly excludes. Checked and not usable.
• Any 'N million singles over 60' or 'N million older adults using dating apps' headline count. — Every version traced back to aggregator sites or platform marketing, never to a census table or a named survey with a 60+ breakout. Same failure mode as the '37M singles 50+' claim rejected on the network's other site. Use shares with a named publisher instead of counts.
• Dating-platform success rates, member counts, or 'X% of relationships now start online' claims used on the review pages. — Self-reported by platforms, not comparable across platforms, and not independently verifiable. /dating-apps-for-women-over-60/ FAQ 4 answers this by refusing the premise. Platform facts may enter this file only as dated first-party observations with the URL observed.
• FTC consumer helpline number 1-877-FTC-HELP as a reporting route. — Could not be confirmed on an FTC-owned page that this checker was able to open (403). The online route ReportFraud.ftc.gov is corroborated; the phone number is not published on this site until it is read on an FTC page.

The rules

1. Named source in the same sentence as the number, with the year.
2. Primary source only — we link the origin, not an article about it.
3. Dated verification: every statistic carries its verification date on the statistics page.
4. Unverifiable → unpublished. No exceptions.