MIT Review Claims Twitter Has A White Male Problem

REUTERS/Kacper Pempel/Illustration/File Photo

Ian Miles Cheong Contributor
Font Size:

The MIT Technology Review is pushing a study that claims Twitter has a white male problem, and that it’s keeping women and ethnic minorities down with what it calls a “glass ceiling.”

In a newly published article, the MIT Technology Review claims that social media plays an important role in “reversing inequality of various kinds.” Fair play to sites like Twitter and Facebook for bridging the class gap by enabling people to openly communicate to each other, but the study claims that a preexisting inequality exists, which works against efforts to end inequality.

Collecting 341 million tweets from 50 million users over a period of three months, researchers Johnnatan Messias at the Federal University of Minas Gerais in Brazil and his colleagues, the study filtered these posts down by time zone, geolocation and profile images to 1.6 million users based in the United States. The pictures they collected were then filtered through what the MIT article calls a “state-of-the-art face recognition application called Face++,” which guesses the ethnicity (black, white, or Asian) of a person’s avatar.

The overall makeup consisted of 53 percent women, 47 percent men. Most users (68 percent) were reported to be white, followed by Asians and blacks.

Messias and his group say that of the top percentile of Twitter users with the most followers they collected, 57 percent of them were men and 43 percent were women. The 15 percent swing represents the inequality in society, which they claim exists. “The results reflect the general idea that higher positions are usually taken by males,” Messias said.

White users also come out on top with the most followers “who are overrepresented by 20 percent,” according to the research, which the MIT article claims is evidence of white privilege. White females are similarly overrepresented by 3 percent. The study claims that the most underprivileged groups are Asian and black females, who are underrepresented by 31 percent.

Messias says that his group traced the way genders and races link to each other and found that people of the same race tend to follow each other more often than they follow people outside their groups. It’s an effect called “homophily.”

The article claims: “But the magnitudes of this effect are interesting. White people tend to follow more white people than expected by a margin of 16 percent. Black people tend to follow more black people than expected by a very significant margin of over 200 percent. However, Asian people tend to follow fewer Asian people than expected by a margin of 10 percent. ‘The expected homophily was not clear for the case of Asians,’ say Messias and co.”

Asians are not a homogenous group, especially not in the United States where they’re made up of many different ethnicities, including Chinese, Japanese, Korean, Filipino, Cambodian, Indian, and so on. So there’s little reason to think why any of them would identify with each other as “Asians.” Furthermore, North American users on Twitter are more likely going to associate with white people, regardless of ethnicity, because there’s just more white people in North America.

The Twitter user Uberfeminist has taken issue with the paper’s conclusions and the results of the Face++, which wrongly identified former Reddit CEO Ellen Pao as white.

It also lists Black Lives Matter’s Shaun King as white, and Democratic congressional candidate Brianna Wu as a man.

Tags : mit twitter
Ian Miles Cheong