Large protest events can be divisive, spurring an outpouring of both support and opposition. But new Penn State research found that the 2017 Women’s March, which championed goals in support of women and human rights, was met with mostly positive support on social media, with relatively few negative messages.
In an analysis of all geolocated tweets in the continental U.S. on the day of the march, the researchers found that not only were tweets about the march generally positive, but they were actually more positive than other geolocated tweets — those that have an attached location — on that day. Tweets about the march rose to a peak of 12% of all geolocated tweets on that day.
Diane Felmlee, professor of sociology, said research into understanding how social justice issues are reflected in social media continues to be relevant given the current protests championing racial justice happening across the country.
“The U.S. has a long history of protest movements, from Colonial times through Civil Rights and Vietnam War protests,” Felmlee said. “Recently, however, protests are accompanied by corresponding online reactions in real-time. Furthermore, controversy regarding Twitter’s recent move to increase its labeling of tweets that are misleading, glorify violence, or could cause harm, makes studying links between social media and events that advocate social causes especially important.”
According to the researcher's, the Women's March — which took place on January 21, 2017, the day after the inauguration of President Donald Trump — was at the time the largest single-day protest in American history. More than 4 million people, about 1.3% of the nation's population, participated.
But despite the widespread support and participation, Felmlee said the March also spurred some controversy from certain political and activist groups.
“The Women’s March of 2017 was an inspiring, historic event, and my coauthors and I wanted to capture public sentiment before, during, and after it occurred,” Felmlee said. “My own research focuses on sexist and racist online harassment, and we were concerned that a backlash toward the March could erupt into extensive, aggressive, harmful messaging on social media.”
For the study, the researchers used the Twitter Streaming Application Programming Interface to collect all geolocated tweets in the continental U.S. from Jan. 20 to 22, 2017, that referenced the Women’s March and its sister marches. They analyzed the content of the tweets to rank them on how positive or negative they were.
The researchers found that from Jan. 20 to 22, 3.1% of all tweets in the study contained a term about the Women’s March. On the day of the March — Jan. 21 — more than 40,000 users tweeted more than 64,000 geotagged messages about the Women’s March. These tweets made up 2.6% of all tweets.
According to the researchers, while the majority of messages were positive, they did find negative ones. Negative sentiment was mostly tweeted from seven metropolitan areas in Alabama, Colorado, New Mexico, Ohio, Oregon, South Dakota and Washington. However, these tweets made up a small portion of tweets.
Felmlee said she was surprised to find few aggressive and bullying tweets that targeted the March or its participants.
“The widespread, largely supportive and positive nature of messages across the continental U.S. was significant,” Felmlee said. “I was particularly surprised by these generally supportive reactions to the March, especially given the political divides in our country, and the all too common occurrence of sexist and racist online harassment and abuse.”
Additionally, Felmlee said the results — recently published in PLoS ONE — give insights into how social media fits into social movements and pushes for change.
“The physical and online worlds are linked — social movement rallies can be effective tools to gain online public support for social justice causes,” Felmlee said. “This is why some governments shut down social media during protests. More positively, it is why so many agencies and organizations use social media to disseminate information.”
Justine I. Blanford, associate teaching professor of geography; Stephen A. Matthews, professor of sociology, anthropology, demography, and geography; and Alan M. MacEachren, professor of geography and information science and technology, also coauthored this work.
The Penn State Social Science Research Institute and the Population Research Institute helped support this research.