Surveys and reporting collections
Social media safety statistics: what the evidence measures
Start with who or what was counted.
These sources do not produce one overall social-media safety rate. CDC measures electronic-bullying experience among U.S. high-school students. FTC describes submitted fraud reports and reported losses. Their percentages have different denominators and cannot be added or ranked as risks.
In CDC's 2025 YRBS, 14.6% of students with a valid answer reported electronic bullying in the prior 12 months (95% CI 13.5–15.8%). This includes texting and is not social-media-only. CDC national Total table.
Separately, social media was the contact method in 28% of FTC's included 2025 fraud reports indicating a monetary loss. That is a share of reports, not the share of users scammed. FTC Spotlight, footnote 1.
2025 data · CDC: U.S. public/private school grades 9–12 · FTC: reporting collection, residency filter unspecified
FTC excludes missing contact method and write-in other; it did not collect reports during the 2025 government shutdown. CDC materials are free at CDC.gov; Social Usage is not endorsed by CDC, HHS or the U.S. Government.
Contact method describes the reports in the collection
The chart answers: among the included reports of monetary loss, how was contact made? It does not follow all users of each contact method to see who was scammed. The source reports five leading methods, not a complete distribution.
Percent of loss reports after excluding missing contact method and write-in other
Source: FTC Data Spotlight, footnote 1. Collection residency filter unspecified; FTC did not collect reports during the 2025 government shutdown. These are reports, not population risk estimates.
Read the contact-method table
| Contact method | Share |
|---|---|
| Website or app | 31% |
| Social media | 28% |
| Phone call | 11% |
| 10% | |
| Text | 7% |
No residual category is calculated. The source does not supply confidence intervals or raw report counts for this selection.
A method with a larger share of these reports is not necessarily riskier per user. We would need an appropriate count of exposed people, a defined period, a consistent harm definition and a method for measuring outcomes among those people. This collection does not supply that denominator. Nor does it establish that each report represents a different person.
FTC also reports $2.1 billion in social-origin fraud losses in 2025. That aggregate amount measures money reported lost; it is neither a victim count nor an average loss per person. We do not divide it by a user total or use it to rank platforms. FTC Spotlight, footnote 2. The same collection interruption applies.
Three measures, three different questions
| Measure | Counted entity and base | What it supports |
|---|---|---|
| CDC electronic bullying | People: weighted students with a valid Yes or No response to Q24 in U.S. grades 9–12. | Prior-12-month electronic-bullying experience in this school population. Not incident counts, all teens or social-media-only harm. |
| FTC contact-method share | Fraud reports indicating monetary loss, excluding missing contact method and write-in other. | Composition of included reports by contact method. Not the chance that a social-media user suffers fraud. |
| FTC reported losses | Reported dollars; no population denominator. Selected reporting year, not necessarily when each scam began. | Aggregate reported loss amount. Not unique victims, complete incidence or individual severity. |
CDC's weighted school survey and FTC's submitted reports enter the evidence through different processes. CDC asks sampled students about experience whether or not a complaint was filed. FTC's collection describes submitted reports, including reports from contributors. A shared data-year label does not make those processes or populations comparable.
The school survey's confidence interval describes sampling uncertainty under its method. It does not turn the finding into an all-teen or all-user estimate. The absence of an FTC confidence interval does not mean its reports capture every scam or establish a precise population rate.
Methods: CDC 2025 Data User's Guide, weighting p12; Q24/QN24 p23 (PDF p26); FTC Consumer Sentinel; Spotlight footnotes 1–2.
Match the question to the evidence
- How common is electronic bullying among high-school students? Use the CDC electronic-bullying analysis, with its population, recall window and confidence intervals. Its grade and source-sex breakdowns retain the same question.
- What kinds of social-origin scams appear in reports, and how much money was reported lost? Use the FTC scams analysis. Keep reporting exclusions and the collection interruption with the results.
- Which platform is safest for an individual? Neither selection answers that question. Contact method is not a platform-specific exposure denominator, and the school item includes texting without separating named platforms.
- What share of people experience any online harm? These selections cannot supply that total. The samples, eligibility rules and outcomes differ; possible overlap is not measured. Adding, subtracting or averaging the percentages would not resolve those missing intersections.
A change in a reporting total could reflect changes in submissions, collection or underlying events. These retained records cannot separate those explanations. Do not relabel an increase in recorded dollars as an increase in the probability of harm.
Related questions need their own measures: perceived health misinformation is not an audit of false posts, and survey associations with mental health do not establish causation. This page is a guide to interpreting selected evidence, not a complete inventory of online harms.
When citing a finding, carry the counted entity, eligible base, period and limitation into the same sentence. The survey-question guide explains why response wording and denominator choices matter.
Sources, dates and reuse
Original periods and source checks
CDC: 2025 YRBS; the prior 12 months before each student's survey response. Exact fieldwork dates, item sample size and dynamic-table publication date remain unknown. Supporting guide: September 2026. Statistical retrieval and source-rights check: 2026-10-06. Selected Total table.
FTC: reporting year 2025. Spotlight displayed publication date 2026-04-27; retrieval and source-rights check 2026-10-06. No U.S.-resident-only filter is stated. Source dollar rounding and whole-percent contact shares are retained; exact counts and sampling uncertainty are not supplied. Original Spotlight.
This synthesis reuses previously reviewed records without new source retrieval. Its editorial review date does not make the statistics newer. It adds no population estimate, cross-source calculation or new source release.
Source credit and reuse limits
Source: CDC. Original materials are available free of charge on CDC.gov. Use by Social Usage does not imply endorsement by CDC, HHS or the U.S. Government.
Source: United States Federal Trade Commission (FTC). Selected U.S. Government material is in the public domain; other source material may retain its rights. Social Usage is not affiliated with or endorsed by the FTC.
Selected federal aggregate facts and original Social Usage prose/chart follow the retained source-specific rights decisions. Source logos, graphics, third-party material and individual reports are not distributed. No downloads or embeds are offered. Review was assisted factual and technical checking, not independent human review.