Pinned on vice president and Democratic presidential nominee Kamala Harris’ official Instagram page is a post featuring her alongside her running mate, Minnesota governor Tim Walz. In the comments, along with praise, criticism, and more than one “Trump 2024,” are several comments asking if Harris had offered Walz oral sex, with one calling her “Kamel toe.”
Harris has long been the subject of online abuse, which is likely to intensify as her campaign wears on. But a new report from the Center for Countering Digital Hate (CCDH), a nonprofit that tracks hate speech and misinformation online, found that Instagram failed to remove 93 percent of the 1,000 hateful and violent comments it flagged to the platform targeting both Republican and Democratic female politicians, including Harris.
In doing so, Imran Ahmed, CEO of CCDH, says that the platform is helping to create an environment that discourages women from seeking political office. “It’s an unconscionable, regressive barrier to women’s participation in politics,” he says.
Researchers monitored the accounts of 10 incumbent female politicians in the US for six months. These included five Democrats (Harris, Senator Elizabeth Warren, and Representatives Nancy Pelosi, Alexandria Ocasio-Cortez, and Jasmine Crockett) and five Republicans (Representatives Marjorie Taylor Greene, Anna Paulina Luna, Lauren Boebert, and Maria Elvira Salazar, and Senator Marsha Blackburn). The abuse the researchers observed ranged from death and rape threats to racial slurs and more generally toxic comments.
In one comment directed at Senator Blackburn, a user posted, “Hope someone leaves you for a dead in a ditch.” Another targeting Representative Crockett read, “All these black women trolling her should spend more time not being single mothers, raising the trash that’s destroying your shitty country …” Yet another, this time directed at Representative Pelosi, said, “hope whoever attacked your husband has more people ❤️❤️❤️❤️ so they can finish the job.”
Researchers collected more than half a million comments from 877 Instagram posts between January 1 and June 7, 2024, and, using Google Jigsaw’s Perspective API, analyzed them for content that appeared to violate the platform’s community standards. (Meta’s policies prohibit attacks based on “race, ethnicity, national origin, disability, religious affiliation, caste, sexual orientation, sex, gender identity, and serious disease” as well as threats of violence, calls for self-harm, or “severe sexualized commentary.”) The research team then flagged 1,000 abusive comments to the company using its reporting function to see whether they would be removed from the platform.
Some comments, like one that used a racial slur to refer to Representative Crockett, seem to clearly violate Meta’s community standards. Others, like one directed at Vice President Harris saying “GO TO THE BORDER YOU USELESS PIECE OF SHIT !” are what researchers defined as “toxic”—not necessarily a direct threat or slur but a “rude, disrespectful, or unreasonable comment that is likely to make someone leave a discussion.” Though they may not cross the line to using sexualized or racialized language that would warrant removal, toxic comments are part of what researchers say creates an overall hostile environment for women politicians online. According to CCDH’s analysis, about one in every 25 comments contained toxic content.
Ahmed says this creates an environment where “any woman that wants to enter public office needs to be able to cope with extreme abuse” creating an “ever increasing barrier to entry for women” in politics.
Around the world, women politicians consistently face threats and harassment, both online and off. A report from the Institute for Strategic Dialogue that analyzed online abuse directed toward politicians on Twitter and Facebook during the 2020 US elections found that women politicians were more likely to face abuse on Twitter. On Facebook specifically, the report found that “female Democrats received ten times more abusive comments than their male counterparts, while Republican women received twice as many abusive comments as Republican male peers.”
This abuse has concrete effects. In 2019, women’s groups in the UK highlighted that several women members of Parliament decided not to stand for reelection, citing abuse. A 2020 study of online abuse against Canadian politicians found that “some women indicated gendertrolling might keep them from running (again) in the future.” And in a 2021 study, female politicians in the UK reported feeling “significantly more unsafe” than their male counterparts, due to varying forms of harassment, of which abuse on social media was one of the most common.
Female politicians of color, like Harris, are particularly likely to face abuse online, and do so at the highest rate. One comment identified by CCDH on Harris’ Instagram read, “We don't want blacks around us no matter who they are.”
A 2022 report from the Center for Democracy and Technology found that during the 2020 elections, women of color running for office were more likely to get targeted with disinformation than any other group. “We're seeing this with Harris and with other women of color politicians, but Harris in particular now,” says Dhanaraj Thakur, research director at CDT.
Meta is aware that its platforms can be used as a way to abuse and target female politicians. Since 2021, it has had a stand-alone section on its website entitled “Online Safety for Women in Government,” which encourages users to report hateful or abusive content. The page also includes a link to the company’s #SheLeads guide, which it says was “created to provide women leaders with safety tips and best practices to protect themselves on our platform.” The link directs to a page reading “URL signature expired.”
Cindy Southworth, head of women’s safety at Meta, says, “We provide tools so that anyone can control who can comment on their posts, automatically filter out offensive comments, phrases, or emojis, and automatically hide comments from people who don't follow them. We work with hundreds of safety partners around the world to continually improve our policies, tools, detection and enforcement, and we will review the CCDH report and take action on any content that violates our policies.”
But Ahmed says that CCDH “served up on a silver platter content that meets the criteria for hateful abuse prohibited on that platform. And even when someone else is doing the work for them, nine out of 10 times, they fail to take action.”
Thakur says that the female politicians his team spoke to while conducting their research complained that platforms’ reporting processes were often unclear and that it “took forever” to get a response from companies. “The protocols fell short in many ways,” he says. “And so they didn’t feel as supported as they should.” It is also not clear how Meta determines whether to leave up a post or not or how it determines whether a comment is violative.
But Thakur says the bigger issue is that platforms don’t necessarily differentiate between those at whom hateful or abusive content is directed and how much more damaging that may make it. “As a society we do not treat all politicians the same because of race and gender and other factors,” says Thakur. “But platforms want to treat everyone the same.”
