Harassing behavior is any repeated, unwanted conduct aimed at intimidating, degrading, or controlling another person, and it shows up everywhere: workplaces, schools, online platforms, and personal relationships. What research consistently reveals is that harassment is not random or purely situational. It tends to be driven by identifiable personality traits, reinforced by specific cognitive patterns, and amplified or dampened by the environment in which it occurs. The picture that emerges from decades of study is more layered than most people assume.
Who Becomes a Harasser
One of the strongest findings in harassment research involves a cluster of personality traits known as the Dark Triad: narcissism, Machiavellianism, and psychopathy. These are not clinical diagnoses but personality dimensions that exist on a spectrum in the general population. People who score high on all three are significantly more likely to engage in bullying and harassment. A recent meta-analysis found that the combined Dark Triad score had a strong correlation with bullying perpetration, with psychopathy showing the largest individual effect and narcissism and Machiavellianism contributing moderately.1PubMed Central. Dark Triad traits and workplace bullying: a systematic review and meta-analysis of personality, power, and psychosocial safety Another study found that Dark Triad traits accounted for roughly 42% of the variation in self-reported bullying perpetration, with Machiavellianism and psychopathy emerging as the strongest individual predictors.2Heliyon. Workplace bullies, not their victims, score high on the Dark Triad and Extraversion, and low on Agreeableness and Honesty-Humility
The picture gets more specific when you look at how different traits predict different roles. Narcissism and sadism are both linked to being a perpetrator of workplace harassment, while low agreeableness further increases the likelihood.3Personality and Individual Differences. The incremental effect of Dark personality over the Big Five in workplace bullying: Evidence from perpetrators and targets That same research found neuroticism and Machiavellianism predicted being a target of harassment, while narcissism was actually protective against becoming a victim. This means the personality profile of someone who harasses is genuinely different from the profile of someone who gets harassed. Harassers tend to be disagreeable, dominant, and willing to manipulate, sometimes with a component of outright enjoyment of others’ distress.
How Harassers Rationalize What They Do
Personality alone does not explain everything. People who harass also tend to rely on a set of mental strategies that researchers call moral disengagement. This is the process of convincing yourself that harmful behavior is acceptable. It can take the form of blaming the victim, minimizing the damage, or telling yourself the target deserved it. Research finds that moral disengagement is strongly linked to both offline aggression and online harassment, and that the connection runs partly through empathy deficits. When someone does not register the suffering they cause, it becomes much easier to keep going.4Journal of Assessment and Research in Applied Counseling. Moral Disengagement and Aggression: The Mediating Role of Empathy Deficits
This pattern holds online as well. In studies of cyberbullying among young people, moral disengagement emerged as the strongest predictor of perpetration, with dehumanization and attribution of blame identified as the most powerful specific mechanisms. The effect was not linear either: once moral disengagement crossed a certain threshold, the probability of cyberbullying jumped sharply, especially when combined with low empathy.5Journal of Adolescent and Youth Psychological Studies. Predicting Cyberbullying Perpetration via Random Forest Modeling of Moral Disengagement and Empathy Deficits What this means in practical terms is that harassing behavior is not simply about being a “bad person.” It is about having the right combination of personality, low empathy, and a well-practiced set of internal justifications.
Harassment and Power in the Workplace
Workplace harassment often has less to do with attraction or personal grievance than people assume. Research on sexual harassment in particular suggests it frequently functions as a tool for enforcing social hierarchies. A study using nationally representative data found that women in supervisory roles were more likely to report harassing behavior than women without authority, not less. This runs counter to the assumption that having power protects someone from harassment. Instead, the evidence points to harassment being used as an “equalizer,” motivated by a desire to undermine women who hold authority rather than by sexual interest.6PubMed Central. SEXUAL HARASSMENT, WORKPLACE AUTHORITY, AND THE PARADOX OF POWER
This framing helps explain why certain workplace interventions backfire. Formal grievance procedures, which are supposed to give targets a path to report harassment, have actually been followed by decreases in the number of women in management positions. The likely mechanism is retaliation: filing a complaint makes someone a target for further punishment. By contrast, training aimed at managers, encouraging them to proactively watch for warning signs, has been followed by increases in women managers. Interestingly, mandatory training aimed at all employees, which tends to lecture attendees on what not to do and implicitly frames male employees as potential perpetrators, has also been followed by declines in women’s representation in management.7PubMed Central. The promise and peril of sexual harassment programs The takeaway is counterintuitive but important: not all anti-harassment programs help, and some measurably make things worse.
Online Spaces and the Disinhibition Effect
The internet creates conditions that are nearly ideal for harassing behavior. Anonymity, the absence of face-to-face accountability, and the speed of interaction all contribute to what researchers call online disinhibition. A systematic review and meta-analysis found that both anonymity and online disinhibition had moderately strong positive correlations with cyber aggression.8PubMed. Anonymity, Online Disinhibition, and Cyber Aggression: A Systematic Review and Meta-Analysis This is not just about trolls acting out for fun. The research shows that disinhibition also operates through moral disengagement and low self-control, the same cognitive mechanisms found in offline harassment, but amplified by the digital environment.9Children and Youth Services Review. Online disinhibition and adolescent cyberbullying: A systematic review
Social dynamics on platforms compound the problem. Research on Twitter (now X) found that the tone of the very first reply to a toxic post strongly shapes what comes after. If the first response is also toxic, subsequent replies are more likely to pile on. When more people are already present in a conversation thread before toxicity appears, individuals are less likely to push back, a pattern consistent with bystander effects observed in real life.10arXiv. The Power of Social Norms: How Initial Responses to Toxicity Shape Conversations on Twitter Online harassment, in other words, is not just a collection of individual decisions. It is a social process where early signals snowball.
One specific form of online harassment that has received increasing attention is doxing, the practice of searching for and publishing someone’s private information. A study of adolescents found that social doxing (seeking out social details about someone) and hostile doxing (targeting personally identifiable information) were the two most common forms. Girls were more likely to conduct social doxing, while boys were more likely to engage in hostile doxing aimed at finding targets’ home addresses and personal data.11PubMed Central. Doxing: What Adolescents Look for and Their Intentions The gendered split in doxing behavior mirrors broader patterns in harassment research, where men and women tend to engage in different forms of aggression rather than simply differing in overall frequency.
What Harassment Does to Targets
The health consequences of sustained harassment go well beyond feeling upset. Physiological research has found that people subjected to workplace bullying show lower morning cortisol levels compared to non-bullied colleagues, a pattern that mirrors what is seen in people with post-traumatic stress disorder and chronic fatigue.12PubMed. Bullying at work, health outcomes, and physiological stress response This is significant because cortisol is the body’s main stress hormone, and chronically disrupted cortisol rhythms indicate that the stress-response system has been fundamentally altered. The body stops mounting a normal stress response and instead shifts into a pattern associated with long-term wear and exhaustion.
Mental health effects are similarly severe, and they persist long after the harassment ends. A study of military veterans found that those who experienced sexual harassment, even without accompanying physical assault, reported significantly more severe PTSD, depression, and suicidality than veterans who had no such experiences. Strikingly, veterans who experienced both harassment and assault reported the worst outcomes, but harassment alone was enough to produce lasting damage.13PubMed. The Relative Impact of Different Types of Military Sexual Trauma on Long-Term PTSD, Depression, and Suicidality People sometimes dismiss verbal or non-physical harassment as less serious than assault. The data does not support that assumption. Repeated degradation, even without a single act of violence, rewires the brain’s threat-detection system in lasting ways.
When Harassment Becomes Stalking
Stalking represents a particularly persistent and dangerous form of harassing behavior. Researchers have identified five distinct types of stalkers based on their motivations: rejected stalkers who pursue former partners, intimacy seekers who believe they are in a relationship with the target, incompetent stalkers who lack social skills and do not understand their behavior is unwelcome, resentful stalkers motivated by grievance, and predatory stalkers who plan an attack.14PubMed. Study of stalkers These categories matter because they predict very different levels of risk. Rejected and predatory stalkers are the most likely to escalate to violence, while incompetent stalkers are generally less dangerous but extremely persistent.
What pushes someone toward stalking in the first place often traces back to early attachment experiences. A study of stalkers referred to a forensic clinic found that compared to the general population, stalkers were significantly more likely to recall their parents as emotionally neglectful and to have insecure attachment styles as adults. The degree of attachment disruption varied by stalker type, but the overall pattern held: stalkers as a group had more troubled early bonding experiences than the average person.15PubMed. Parental bonding and adult attachment styles in different types of stalker This does not excuse the behavior, but it helps explain why some people develop an inability to let go of unwilling targets. Their model of relationships is built on insecurity and a desperate need for connection that overrides normal social boundaries.
How Institutions Respond, and How They Often Fail
When someone reports harassment to their employer or institution, the response itself can become a second source of harm. Researchers have a term for this: institutional betrayal. Among people who experienced workplace sexual harassment in one study, roughly 55% also reported experiencing institutional betrayal, meaning the organization failed to respond appropriately, punished them for reporting, or actively covered up the behavior. Institutional betrayal was associated with lower job satisfaction, decreased commitment to the organization, and increased physical symptoms like headaches and fatigue.16PubMed Central. Institutional courage buffers against institutional betrayal, protects employee health, and fosters organizational commitment following workplace sexual harassment
Retaliation after reporting is not rare. At one academic medical center, about 28% of people who filed formal harassment reports experienced at least one form of institutional retaliation. Being labeled a “troublemaker” was the most common form. Men who reported harassment faced especially steep retaliation: they were several times more likely than women to be denied opportunities, given worse assignments, or denied promotions after filing a report.17Academic Medicine. Formal Reporting of Identity-Based Harassment at an Academic Medical Center: Incidence, Barriers, and Institutional Responses That last finding is worth sitting with. Men who break the norm of silence around harassment face an unusually harsh backlash, which likely contributes to underreporting among male targets. The broader implication is that institutions do not simply fail to address harassment. In many cases, their response systems actively discourage people from coming forward.18Sociological Perspectives. Institutional Betrayal: Inequity, Discrimination, Bullying, and Retaliation in Academia
What Actually Works to Reduce Harassment
Given how badly many institutional responses perform, the question of what actually reduces harassment is urgent. The evidence splits into two broad categories: training and bystander intervention.
On training, a meta-analysis found that harassment training programs are quite good at increasing knowledge about what counts as harassment and what reporting procedures exist. The effect on declarative knowledge was large. But the effects get progressively weaker as you move toward the things that actually matter: changing attitudes showed a moderate effect, building skills to handle situations showed roughly the same, and the transfer of training to real behavior on the job was small.19Personnel Psychology. The effects of sexual harassment training on proximal and transfer training outcomes: A meta‐analytic investigation In plain terms, after training, people can define harassment on a quiz but are only slightly more likely to do anything differently when faced with it in real life. Training programs that focus specifically on managers, giving them responsibility and tools to intervene, show better downstream results than blanket employee training that simply lists prohibited behaviors.
Bystander intervention programs take a different approach. Instead of telling potential harassers not to harass, they teach everyone how to recognize and safely interrupt problematic behavior. A multi-campus evaluation found that campuses running bystander programs had about 17% lower rates of interpersonal violence victimization and about 21% lower rates of perpetration compared to campuses without such programs. The reductions spanned unwanted sexual contact, harassment, stalking, and psychological abuse in dating relationships.20PubMed Central. Multi–College Bystander Intervention Evaluation for Violence Prevention Research in China found similar patterns: bystander programs changed not only attitudes but actual intervention behavior, with participants reporting greater willingness to step in when they witnessed harassment after training.21PubMed. Empowering Chinese College Students to Prevent Sexual Assault in Post-MeToo Era: An Empirical Study of the Bystander Intervention Approach Results have been somewhat mixed across studies, with some showing strong gains in bystander confidence and others showing smaller effects, likely driven by differences in program design and length.22Aggression and Violent Behavior. Examining the efficacy of bystander sexual violence interventions for first- year college students: A systematic review Still, the overall pattern favors bystander approaches over traditional compliance-style training.
Childhood Roots of Harassing Behavior
Many adults who engage in harassment did not develop those patterns in a vacuum. A study of adolescent bullying perpetration found that childhood adversity, including harsh punishment, sexual abuse, and a negative home atmosphere, was significantly associated with engaging in bullying later in adolescence. Of all the factors examined, peer influence had the strongest link to bullying perpetration, outweighing any single adverse childhood experience.23Scientific Reports. Childhood adversity and peer influence in adolescent bullying perpetration This combination matters: a child who grows up in a punitive or neglectful environment is primed for aggression, but whether that potential gets activated depends heavily on who they spend time with during adolescence. Peer groups that normalize cruelty and dominance can turn an at-risk teenager into an active bully, while peer groups with different norms can buffer against it.
This developmental lens connects to the broader research on stalkers and their disrupted attachment histories, and to the finding that moral disengagement is a learned cognitive pattern rather than an inborn trait. People are not born harassers. They develop the personality traits, acquire the cognitive rationalizations, and find the environments that permit or encourage the behavior over the course of years. That does not make the behavior less harmful, but it does suggest where prevention efforts are most likely to succeed: early, and focused on the social ecology around young people rather than on punishment after the fact.