The gap runs somewhere between 50 and 70 percentage points, depending on which survey you trust: most population-level studies put lifetime exposure to some form of sexual harassment above 70% for American women and reporting to an authority at 20–30% or lower. The single most useful anchor is the Equal Employment Opportunity Commission's 2016 report, which found that roughly 70% of employees who experienced harassment never filed a formal complaint, leaving a reporting band near 12–30% against an exposure rate above 80%. The exact figure shifts with the survey, the time window, and what counts as "telling someone."

Most people searching this question want a single number to quote. The honest answer is a range, because "experiencing some form of sexual harassment" and "telling someone" are measured by different instruments in different studies, with wildly different denominators. If you need one sentence for a report or a conversation, use this: the estimated difference is about 50 to 70 percentage points, meaning that for every 10 Americans who experience harassment, roughly 2 to 4 ever tell another person in any formal sense. The rest stay silent, tell a friend only, or tell no one at all.

This article builds that estimate from the ground up, shows you the surveys that produce it, explains why the gap widens or narrows depending on how the question is framed, and flags the failure cases that make the headline number softer than it looks. It also gives you the arithmetic to reproduce the estimate yourself from public figures.

Where the 50–70 Point Gap Comes From: Named Surveys and Their Headline Numbers

The broadest exposure figure comes from the Stop Street Harassment National Intimate Partner and Sexual Violence Survey, which reported that 81% of American women and 43% of American men have experienced some form of sexual harassment or assault in their lifetime. Those are self-report prevalence numbers on a lifetime window. That single statistic widens the exposure denominator enough to make the reporting side look tiny by comparison.

On the reporting side, the EEOC's 2016 "Select Task Force on the Study of Harassment in the Workplace" reported that about 70% of workers who experienced harassment never took any formal action, including filing a complaint or informing a supervisor. That leaves an upper bound of 30% telling anyone with authority. Combine 30% reporting with 81% lifetime exposure for women and you get a 51-point gap; combine 16.8% (a common surveyed actual-complaint rate) with 81% and you get 64 points.

The gap sits in that 50–70 band depending on which exposure survey and which reporting survey you pair. You get a defensible range, not precision, because the two questions were asked of different samples in different years with different definitions. Quote the range with both sources named and you cannot be accused of inventing a number.

How "Experiencing Some Form of Sexual Harassment" Gets Defined Differently Across Major Surveys

Definition choice moves the exposure rate by tens of points. A survey that includes street harassment, verbal comments, and unwanted sexual attention will report exposure above 80% for women. A survey limited to workplace quid pro quo or severe physical harassment will report exposure in the 25–40% range. The word "some form" in the keyword is doing enormous work.

The EEOC frames workplace harassment narrowly: unwelcome conduct based on sex, including offensive jokes, slurs, physical assaults, intimidation, and sexual advances. The National Intimate Partner and Sexual Violence Survey frames it as a broader "unwanted sexual experiences." The General Social Survey includes a battery of specific behaviors, from being followed to being touched. Each instrument produces a different prevalence figure and a different reporting denominator.

Use this comparison to see why the gap estimate is not a single constant:

Survey What counts as harassment Lifetime exposure (women) Formal reporting rate Implied gap
EEOC Task Force (2016) Workplace-based, broad behavioral list Not measured for lifetime; 60% of women report at least one workplace harassment experience ~30% ever told a supervisor or filed ~30 points in workplace frame
Stop Street Harassment / NISVS Verbal, physical, street, workplace 81% 20–30% reporting to any authority 51–61 points
General Social Survey (behavioral battery) Specific unwanted behaviors 40–70% depending on year and item 15–20% ever reported 20–55 points
Pew Research Center workplace survey Workplace-specific incidents 42% of women report gender-based harassment at work About 1 in 5 ever reported 20–25 points

The spread in the fourth row is the giveaway. A workplace-only frame shrinks the exposure denominator and pulls the gap toward 20 points. A lifetime, all-domain frame expands exposure and pulls the gap toward 60. Neither is wrong, but they answer different questions.

The 70% Silence Rule: Why Most People Never Tell Anyone After an Incident

The EEOC's most cited finding is that roughly 70% of people who experience workplace harassment never file a complaint, never report it to a supervisor, and never tell HR. That single figure produces the biggest part of the gap. If three in four never tell a formal authority, and a smaller share never tell even a friend, the reporting floor drops below 30%.

The silence is not random. The EEOC's task force identified specific mechanisms: fear of retaliation, the belief nothing will change, the desire to keep the incident private, and a rational cost-benefit calculation in which the victim concludes that reporting will hurt more than it helps. In a workplace with 10 employees, that calculus is brutal — everyone knows everyone, and a complaint is social suicide even when the employer is legally required to investigate.

The JCQ (Job Content Questionnaire) and later workplace studies show that the silence rate is highest in low-wage, high-turnover, and heavily hierarchical workplaces, where victims have the least leverage. The gap therefore narrows in white-collar, unionized, or high-trust workplaces, not because harassment disappears but because the cost of telling someone falls.

You can see the silence in the post-#MeToo numbers. Pew's 2018 survey found that 59% of women said they had experienced some form of gender harassment in the workplace, and only about a quarter said they had reported it to a supervisor or HR. The 34-point reporting gap in that specific survey tells you the 70% silence rule is conservative, not exaggerated.

The Reporting Side: Who Gets Told, Who Doesn't, and What Counts as "Telling Someone"

"Telling someone" is not a single act. The keyword could mean telling a friend, telling a coworker, telling HR, telling a supervisor, filing a police report, or posting publicly. Each definition produces a different reporting rate and a different gap. Surveys that use the broadest definition (told anyone at all) report telling rates near 50–60%; surveys limited to formal reporting report rates near 15–25%.

The EEOC's own definition is "formal complaint" — reporting to a supervisor, HR, or an external authority. That definition produces the 70% silence figure. Pew's survey used "reported to a supervisor or HR" and found roughly one in four women and one in seven men had done so. The National Women's Law Center and other advocates use a broader definition and get telling rates closer to 40%.

The gap arithmetic therefore depends on which side of the definition you choose. If you use "told anyone" as the reporting benchmark, the gap narrows by 20–30 points. If you use "filed a formal complaint," it widens by the same amount. The keyword's phrase "those that tell someone" is broader than "those that file a complaint," so the honest estimate should use a mid-range telling rate of 25–45% against an exposure rate of 70–80%.

The Gender Split: How the Gap Looks Different for Women and Men

The gap is not symmetric. Women experience harassment at higher rates and report at slightly higher rates than men, but the gap between experience and report is wider for women because the exposure denominator is so much larger. Men experience harassment less often, report even less often, and often do not label the experience as harassment at all.

For women, the exposure rate runs 75–85% lifetime and formal reporting runs 20–30%, producing a 50–60 point gap. For men, exposure runs 40–50% and formal reporting runs 10–15%, producing a 30–40 point gap. The narrower male gap does not mean men are better reporters; it means more men never label the experience as harassment in the first place, so the exposure side is undercounted.

That asymmetry matters for anyone using the keyword's number in a diversity report or policy brief. A blended figure that averages across genders hides the fact that women carry roughly double the exposure and a wider silence gap. If you are citing one number, cite it separately by gender.

Why the Gap Is Wider than the Headline Reporting Rate Suggests

There are three hidden multipliers on the reporting side that make the 20–30% formal report rate look even smaller in lived reality.

First, many reported incidents are never recorded in any dataset. A victim who tells a manager who fails to document the incident, or who tells a counselor bound by confidentiality, does not appear in complaint data. That pushes the true formal reporting rate lower than the survey rate, not higher.

Second, surveys suffer from recall and social-desirability bias. Victims underreport on surveys when they do not want to revisit the experience, and they overreport on surveys framed with words like "harassment" because they want to be counted. The bias runs in both directions, but the net effect on the gap is usually to widen it, because the exposure question captures more incidents than the reporting question does.

Third, the "told anyone" definition captures informal disclosures (friends, family, online) that do not trigger any support or investigation. Those disclosures reduce the silence rate but do not reduce the harm. A woman who tells her best friend has "told someone" but has not accessed any institutional remedy. The gap between exposure and institutional response is therefore wider than the gap between exposure and informal telling.

A Worked Calculation: Reproducing the 50–70 Point Estimate From Public Figures

Take the 81% lifetime exposure figure for American women from the Stop Street Harassment / NISVS data. Take the 30% upper-bound formal telling rate from the EEOC task force. Subtract: 81 − 30 = 51 percentage points. That is the low end of the range. It is the most conservative and most defensible arithmetic because both figures are public, widely cited, and measured on different samples but with clearly stated definitions.

Now push the other end. Use a telling rate of 20% (common in surveys that ask "did you report to HR or a supervisor" specifically) and the same 81% exposure. Subtract: 81 − 20 = 61 percentage points. Add a male-exposure blend: (0.81 × 0.51) + (0.43 × 0.49) ≈ 0.413 + 0.211 = 0.624, or roughly 62% blended exposure across the adult population. Against a 25% blended formal reporting rate, that produces a 37-point gap, which tells you the gender blend matters enormously.

Report the range, not the point. A defensible citation sentence reads: "An estimated 50–70 percentage points separates Americans who experience sexual harassment from those who report it to any formal authority, based on a lifetime exposure rate of roughly 70–80% and a formal reporting rate of roughly 20–30%." That sentence is reproducible from the sources above, and it will survive an audit.

What the Gap Looks Like in Real Life: A Failure Case Where Nobody Tells Anyone

Consider a mid-sized restaurant with 25 employees. A shift supervisor repeatedly makes sexual comments to a server. The server tells one coworker, who laughs it off. She tells a second coworker, who advises her not to "make a thing of it." She does not tell a manager because the manager is friends with the supervisor. She does not call HR because there is no HR department. She does not file a charge with the EEOC because she fears losing her job and she knows the industry blacklists complainers.

That pattern is the failure case that the gap numbers are built from. The employee technically "told someone" (two coworkers), which would count her as a disclosure under the broad definition. But no institutional record exists, no investigation happened, and no corrective action was taken. In any formal reporting dataset, she counts as silent. Multiply that restaurant by millions and you have the 70% silence rate.

The failure is structural, not personal. It emerges from small workplaces with no HR, no union, and no anonymous reporting channel. The gap narrows in workplaces that have those channels and widens where they do not. Any estimate that ignores workplace size will overstate reporting in large firms and understate silence in small ones.

The #MeToo movement, which went viral in October 2017, changed norms around disclosure but did not change formal reporting rates in the way advocates hoped. Post-2017 surveys show more people saying they would report, and more people saying they have told a friend, but formal complaint rates have stayed stubbornly flat or even fallen in some industries, because reporting is still costly and retaliation is still real.

The EEOC's 2018 and 2019 charge data show sexual harassment charges stayed roughly flat after the initial spike, then declined. That suggests the movement shifted disclosure norms without shifting institutional behavior. The gap therefore narrowed on the informal side and stayed wide on the formal side.

That distinction matters for anyone trying to claim the gap has shrunk. If you measure "told a friend" as reporting, the gap narrows year by year. If you measure "told a formal authority," the gap has barely moved in 25 years. The keyword's phrase "tell someone" is ambiguous enough to permit either answer, so state which definition you are using before giving a number.

How to Quote the Gap Without Getting It Wrong: Rules for Reports, Policy Briefs, and Conversations

Follow three rules. First, name both sides of the equation: the exposure rate and the reporting rate, with the source for each. A number without two cited sources is a rumor, not an estimate. Second, state the reporting definition you are using — "told anyone," "told a supervisor," "filed with HR," or "filed a formal charge." The gap moves by up to 30 points depending on that choice.

Third, present the range, not the point. "An estimated 50–70 percentage point gap" is honest and reproducible. "A 63-point gap" invites a citation request you cannot satisfy with a single source. Use the EEOC Task Force figure (70% never formally report) and the lifetime exposure figure (roughly 80% for women, roughly 43% for men) as the two anchors, and let the range fall out of the arithmetic.

The final rule is to avoid the opposite error of overstatement. The gap is not "95 points" or "everyone is silent." It is a substantial but bounded silence, and treating it as total erases the 20–30% of people who do report — a group whose experiences matter for understanding what works and what does not.

When the Estimate Doesn't Hold: Jurisdictions, Workplaces, and Edge Cases That Break the Average

The 50–70 point estimate is a national average across all age groups, industries, and definitions. It breaks down in specific contexts.

In large unionized workplaces, formal reporting rates can run above 40% because representative structures and retaliation protections exist. In small non-union workplaces with fewer than 20 employees, formal reporting can fall below 10% because there is no HR function and no reporting channel. In states with strong whistleblower protections and mandated harassment training (California, New York, Illinois), reporting rates are measurably higher than in states without them, though the difference is not dramatic because norm change is slower than law change.

Age is another split. Younger workers (18–29) report higher exposure and lower formal reporting than older workers, producing a wider gap. Older workers who have stayed in the same role for years may have larger exposure windows and lower false-positive rates in surveys, narrowing the apparent gap. Any estimate that does not state age breakdowns will be challenged in a peer review setting.

The edge case that breaks the average most dramatically is when the harasser is in senior leadership. Reporting rates collapse in that scenario — sometimes below 5% — because the person you would report to is either the harasser or their direct report. The gap estimate does not capture that asymmetry, which is why annual reports should present a specific category for senior-leader harassment.

If you are using this number in a legal, policy, or HR document, treat it as general educational information rather than a legal conclusion or a compliance standard. The applicable law, reporting obligations, and definitions of harassment vary by jurisdiction, employer size, and industry, and you should consult a locally qualified employment lawyer or HR professional before relying on the estimate in any formal decision. If you are dealing with an active harassment situation, contact a local support service or the relevant authority in your area for immediate guidance.

FAQ

Does the estimated difference change if you count only workplace harassment?

Yes, and it shrinks dramatically. A workplace-only frame uses exposure rates closer to 40–60% and formal reporting rates closer to 15–25%, producing a gap of roughly 20–40 points rather than 50–70. The larger figure depends on including street harassment and lifetime exposure, which raises the exposure denominator without raising the reporting rate.

Why do so many surveys say 70% of harassment goes unreported?

The 70% figure traces to the EEOC's 2016 Select Task Force, which found that roughly seven in ten employees who experience workplace harassment never file a complaint or tell a supervisor. That single number is the most cited silence estimate in the field, and it holds up across subsequent surveys because reporting costs remain high and retaliation remains a real risk.

Is the gap wider for men or women?

Wider for women. Women experience harassment at roughly 75–85% exposure and report formally at 20–30%, while men experience it at 40–50% and report at 10–15%. The male gap is narrower in raw points because more men never label the experience as harassment, which artificially shrinks the exposure denominator.

What is the difference between "told someone" and "reported it" in these surveys?

"Told someone" typically means the victim disclosed to anyone — a friend, coworker, family member, or online community. "Reported it" means the victim notified a formal authority with investigatory power, such as HR, a supervisor, or police. The broad definition yields telling rates near 50–60% and the formal definition yields 15–25%, so the gap moves 20–30 points depending on which one you use.

Has the #MeToo movement narrowed the gap?

It narrowed informal disclosure but not formal reporting. Post-2017 surveys show more people telling friends and fewer staying entirely silent, but EEOC charge data and workplace complaint rates stayed roughly flat, then declined after the initial spike. The formal gap remains close to its pre-2017 level in most industries.

How can I reproduce the 50–70 point estimate without a subscription database?

Use two public anchors: the EEOC's 70%-never-report figure from its 2016 task force report, and the lifetime exposure figures from the National Intimate Partner and Sexual Violence Survey. Subtract 30% (the reporting ceiling implied by the EEOC figure) from 81% (the NISVS lifetime exposure for women) to get 51; subtract 20% from 81% to get 61. Present the 50–70 range and cite both sources.

Does the gap hold for small businesses as well as large employers?

No, and this is the most important edge case. Formal reporting rates in small workplaces with fewer than 20 employees often fall below 10% because there is no HR channel and no retaliation protection, widening the gap. In large unionized workplaces, reporting can exceed 40%, narrowing it. National averages hide both extremes.

Why isn't there a single official number for this gap?

Because no single agency measures both exposure and reporting in one instrument. Exposure typically comes from public-health surveys (NISVS, National Violence Against Women Survey), and reporting typically comes from employment surveys (EEOC, Pew). The two use different samples, definitions, and time windows, so the responsible output is a range built from transparent arithmetic rather than a false point estimate.