Rational Withdrawal: Victim Responses to Online Hate Speech and the Limits of Institutional Protection in Norway
This mixed-methods study of 111 respondents in Norway reveals that both direct victims and witnesses of online hate speech predominantly respond with anger and avoidance rather than legal action, suggesting that their withdrawal is a rational adaptation to structural constraints and institutional fragmentation that ultimately threatens civic discourse.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the digital public square, a new kind of harm has taken root, one that travels faster than any physical threat and leaves no visible bruise. This is online hate speech: cruel, targeted messages sent to individuals or groups because of who they are, their skin color, their faith, or their background. For decades, researchers have understood that such attacks do more than hurt feelings; they can silence people, driving them out of conversations and away from public life. This phenomenon is often described as a "chilling effect," where the fear of abuse makes people choose silence over speech. Another related idea is the "spiral of silence," which suggests that when people feel their views are unpopular or dangerous, they stop speaking up, causing those views to disappear from the public eye even if many people actually hold them. The question that matters most today is not just how much hate exists, but how the people who face it respond. Do they fight back, seek help, or simply walk away? And if they walk away, is it because they are weak, or because the world around them has made it impossible to stay?
A recent study conducted in Norway offers a clear, sobering answer to these questions. Researchers surveyed 111 people, including those who had been directly targeted by hate speech and those who had simply witnessed it happening to others. The goal was to map out the emotional toll of these attacks and, more importantly, to understand the choices people make when they are under fire. The findings reveal a pattern that challenges the way we often think about resilience. Instead of seeing people who stop posting or leave social media as passive victims who have given up, the study suggests they are making a rational, calculated choice. When the systems designed to protect them—like police reporting or platform moderation tools—seem broken or inaccessible, stepping back becomes the only logical way to survive.
The study, which gathered data in early 2024, found that for the vast majority of those exposed to hate, social media was the battlefield. Among the 75 people who had either experienced or witnessed hate speech, the most common emotional reaction was anger, reported by more than half of the group. Fear and shame were also widespread, with over a third of respondents feeling violated or humiliated. Yet, despite this intense anger, the most common way people dealt with the abuse was to ignore it or avoid engaging with it. More than half of the respondents chose to simply look away. While some people blocked the offenders or reported the posts to the website administrators, the most striking finding was what did not happen: not a single person in the study took legal action.
This zero percent rate of legal involvement is the core of the paper's argument. The researchers propose a concept they call "constrained resilience." Usually, when we talk about resilience, we think of it as an inner strength, a personal trait that helps someone bounce back from trauma. This study argues that resilience is not just a personal quality but is shaped by the environment. If the environment is full of barriers—such as laws that are too hard to use, police who do not understand the specific nature of online abuse, or websites that ignore reports—then a person's ability to cope is "constrained." In this context, withdrawing from the conversation is not a sign of defeat or a lack of courage. It is a smart, adaptive strategy. When the tools for fighting back are broken, the most rational thing to do is to protect oneself by leaving the field.
The study also looked at people who were not the direct targets of hate but had seen it happen to others. These witnesses, who made up nearly half of the exposed group, showed coping patterns almost identical to the direct victims. They, too, were more likely to avoid engagement and less likely to seek help. This suggests that the "chilling effect" spreads beyond the person being attacked. When people see others being harassed and see that nothing is done to stop it, they learn that the space is unsafe for them too. This creates a ripple effect where entire groups of people, particularly those from minority backgrounds, begin to disappear from public discourse. The study notes that this is not just an individual problem but a civic one; when voices are silenced, the public conversation becomes poorer and less representative of the society it is meant to serve.
One of the most practical discoveries in the research concerns the gap between knowing about help and actually getting it. The data showed that people who were unaware of any support resources were far less likely to seek help than those who knew where to turn. In fact, those who knew of at least one resource were four times more likely to seek assistance than those who did not. This points to a simple but critical failure in the current system: the people who need help most often do not know it exists, or they do not know how to access it. The researchers argue that the solution is not necessarily more laws or new technology, but better coordination. They suggest creating a single, clear gateway that guides victims to the right support, whether that is legal advice, mental health care, or platform reporting tools.
The study also casts a skeptical eye on the idea that artificial intelligence will solve the problem of online hate. While many tech companies are building AI tools to detect and remove harmful content, the research suggests that these tools can be opaque and unreliable. If the people who are being attacked already distrust the systems in place, adding a layer of automated decision-making that they cannot understand or appeal may only make them feel more isolated. The researchers warn that if these tools are not designed with transparency and community input, they risk making mistakes that could silence innocent people or fail to catch the very hate they are meant to stop.
Ultimately, the paper paints a picture of a digital society where the rules of engagement are broken. The people being targeted are not passive; they are actively navigating a landscape where the usual defenses have failed. Their decision to withdraw is a rational response to a system that offers them no viable path to justice or safety. By understanding this behavior not as a personal failure but as a structural one, policymakers and platform designers can begin to build systems that actually work. The goal is to create an environment where people do not have to choose between their safety and their voice, where the path to support is clear, and where the promise of a free and open digital public square is kept for everyone.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.