Discussion of online harassment tends to focus on its emotional effect, which is real. It understates a second cost that is easier to measure and harder to dismiss.
Sustained harassment generates work. Blocking accounts. Reporting content. Screenshotting for evidence. Adjusting privacy settings. Deciding whether to respond. Explaining to platforms, employers or police. Managing the practical consequences of doxxing.
This is time, and it is time not spent on anything else.
The distribution
Surveys of online harassment consistently find that overall rates of experiencing some form of harassment are broadly similar across genders, and that the composition differs substantially.
Women, and particularly young women, report higher rates of sexual harassment online, of stalking, and of sustained harassment by the same person or group over extended periods.
The sustained category is the one that generates workload. A single insult requires no management. A campaign requires ongoing administration.
Studies of harassment directed at people with public professional profiles — journalists, academics, politicians — have found substantially higher volumes directed at women, with the disparity greatest for women who are also members of minority groups.
The behavioural consequences
The most consequential effect is not distress but withdrawal, and it has been documented repeatedly.
Surveys of people who have experienced online harassment find substantial proportions reporting that they subsequently self-censored, reduced their online presence, declined public-facing opportunities, or left a platform.
Studies of journalists in particular have found that harassment influences what topics are covered and by whom, with some reporting avoidance of particular subjects.
This is the mechanism by which harassment produces effects beyond the individual. A field in which certain contributors are subject to substantially higher costs for participating will, over time, have fewer of them.
Why platform reporting works badly
The reporting mechanisms available on most platforms are designed around individual pieces of content, and this mismatch is the core problem.
Harassment is frequently distributed across many accounts, each of which posts content that individually falls below the threshold for action. The pattern is the harm; the individual items are not.
Reporting therefore requires the affected person to submit each item separately, and each is assessed in isolation against a policy that considers only its own content.
Coordinated harassment exploits this deliberately. Several platforms have introduced pattern-based and bulk reporting tools in response, with uneven implementation.
The cross-platform problem compounds it. Campaigns typically operate across several services, and no reporting mechanism spans them, so the affected person is the only entity with visibility of the whole pattern.
What actually reduces the burden
Delegation. Where possible, having someone else monitor and manage reporting removes the requirement to read the material, which is where most of the psychological cost sits. Some organisations now provide this for staff, and it is one of the more effective employer interventions.
Automated filtering. Keyword filters, restricting replies, and pre-approval settings reduce exposure substantially and are underused because they are buried in settings menus.
Documentation systems set up in advance. Deciding how to record incidents before they occur is considerably easier than doing it during, and consistent records are what police and platforms require.
Separating professional and personal accounts, with different exposure settings.
And a decision made in advance about the response policy — whether to engage at all — removes a recurring decision that itself consumes attention.
The legal position
Provisions vary widely. Several jurisdictions have offences covering communications intended to cause distress, stalking including online conduct, and threats.
Enforcement is the weak point. Reporting rates are low and prosecution rates lower, and the reasons given by police in research on this include jurisdiction problems, difficulty identifying anonymous perpetrators and resource constraints.
Where the harassment includes threats, doxxing or a persistent identified individual, the position is considerably stronger than for diffuse abuse, and these categories are worth distinguishing when reporting.
What institutions should do and mostly do not
Employers of public-facing staff have a duty of care that most have not operationalised for this.
Measures that have been implemented by some organisations include monitoring and reporting handled centrally, legal support, security assessment where doxxing occurs, and explicit statements that the organisation will not require staff to maintain a public presence that generates this.
The absence of such provision effectively means the cost is borne privately by the individual, which is a distributional decision an organisation is making whether or not it recognises it.
The framing worth adopting
Treating harassment as an emotional matter invites advice about resilience. Treating it as a workload invites a different question: who is doing this work, why is it unpaid, and what is being displaced by it.
That question has better answers, and it does not require anyone to demonstrate that she was upset enough to warrant help.