People Censor Themselves Online for Fear of Being Harassed

Christian Bertrand/

New research reveals when tech companies don't police abuse, it can put a damper on free speech.

When somebody famous gets harassed online, it usually makes a stir. The comedian Leslie Jones logged off Twitter this summer after being barraged with racist and sexist tweets. Last year, Zelda Williams walked away from her account when she was brutally harassed about the death of her father, the actor Robin Williams.

But abuse is a consistent feature of online life, even when it isn’t highly publicized. Nearly half of American internet users have been harassed or abused online, according to a new study published Monday by Data & Society, a technology-focused think tank.

Some groups are more often targeted than others. Lesbian, gay and bisexual users are more than twice as likely than straight users to experience abuse online, the study found, and although men and women are subject to similar levels of abuse, the attacks on women were often of a more serious nature.

Of the 20 categories of harassment the researchers looked at, men were more likely to report being called names and being embarrassed online, while women were more likely to be stalked, sexually harassed, or have false rumors spread about them.

But a person doesn’t have to be the target of abuse for it to color their experience online. More than 70 percent of Americans say they’ve seen others harassed on the internet. For black users, that percentage rose to 78; among younger users and lesbian, gay and bisexual Americans, the proportion is close 90 percent. Groups more likely to come into contact with online abuse were also more likely to say people on the internet are mostly unkind.

The researchers’ definitions of harassment and abuse didn’t always line up with internet users’ definitions, though. Of respondents who said they’d had at least one of a subset of the negative experiences being studied, fewer than half said they had experienced “online harassment or abuse.” Men were far less likely than women to consider their bad online experiences to be harassment.

The fact that straight, white men are the least likely group to be the targets of abuse online—and that they’re less likely to interpret negative online experiences as abuse—might help explain the attitudes of tech executives in charge of online spaces like Twitter and Facebook, says Amanda Lenhart, the report’s lead author.

“The people who are building these technological platforms aren’t the people who this data would suggest would be at risk of witnessing or experiencing harassment as much as other groups,” Lenhart said. “So what does that mean in terms of the kinds of things that you think are important to build into the technology that you’re creating?”

For a long time, Twitter had a habit of calling itself the “free speech wing of the free-speech party.” Unlike Facebook, which has more restrictive rules that ban hate speech and abuse, Twitter prides itself on providing an open platform for political discourse—but its open door lets in a lot of vitriol, too. While the platform is still a place for breaking news, thoughtful analysis and chance encounters, much of it is toxic.

Last week, Twitter announced new features that will make it easier for users to report abuse when they see it online. Product changes like these can help make it easier for users to protect themselves—or to stand up for others. Already, 65 percent of American internet users say they’ve responded to harassment that targeted someone else, either by saying something to the abuser or the victim, or by reporting the behavior through the platform.

But design tweaks might not be enough.

“It’d be nice to think that we could find a technological solution that would fix this,” Lenhart said. “But I suspect it will take a combination of technological solutions and conversations around social norms, beliefs and behaviors.”

Getting harassed online has real-life consequences beyond feeling attacked. The researchers found 43 percent of harassment victims changed their contact information—their email addresses, telephone numbers, or social-media handles—and that 26 percent stopped using their social-media account, cellphone, or the internet altogether. One in 10 victims said they felt disconnected from information, family and friends because they avoided online platforms because of harassment.

And more than one in four internet users—not just those who faced abuse online—decided not to post something online for fear of harassment. Among women between the ages of 15 and 29, 41 percent of respondents said they’d censored themselves to avoid being harassed.

These numbers reveal a paradox in online speech: Twitter, in an effort to avoid having to act as a censor, has been slow to tamp down on hate speech and abuse. But as a result, users are censoring their own free speech.

“I certainly think a lot of platforms have defaulted to a deeply held position that freedom of speech is the be-all, end-all that they want to enforce,” Lenhart said. “That worked for a long time—but it worked at the expense of people whose voices are suppressed in a different way, because of the kinds of speech that are possible on those platforms.”

Since Donald Trump was elected earlier this month, there have been increased reports of hate crimes from around the United States: from racist to sexist to homophobic to xenophobic. Last month, a study from the Anti-Defamation League concluded an uptick in anti-Semitic harassment directed at journalists online had been “driven by rhetoric in the 2016 presidential campaign.”

The din of online abuse is showing no signs of abating. Twitter has shown it may be willing to go further than just making it easier to report harassment: Last week, it suspended the accounts of several users connected to the alt-right movement, a loosely connected group of agitators who reject mainstream conservatism in favor of more extreme, insular views. One of the accounts belonged to the head of a white nationalist organization, who led supporters in a Nazi salute this week.

But as Twitter considers how far it’s willing to go to quiet abuse, it should consider a platform without rules poses its own dangers to free speech.