[Paper Review] White, Man, and Highly Followed: Gender and Race Inequalities in Twitter

Johnnatan Messias, Pantelis Vikatos+1 more|arXiv (Cornell University)|Jun 26, 2017

Social Media and Politics39 references21 citations

TL;DR

This study uses Face++ to identify gender and race in 1.67 million U.S.-based Twitter users and analyzes their network connections and interactions. It reveals that White and male users achieve significantly higher visibility and follower counts, indicating systemic racial and gender inequalities in Twitter’s social structure, with White males benefiting most from network advantages.

ABSTRACT

Social media is considered a democratic space in which people connect and interact with each other regardless of their gender, race, or any other demographic factor. Despite numerous efforts that explore demographic factors in social media, it is still unclear whether social media perpetuates old inequalities from the offline world. In this paper, we attempt to identify gender and race of Twitter users located in U.S. using advanced image processing algorithms from Face++. Then, we investigate how different demographic groups (i.e. male/female, Asian/Black/White) connect with other. We quantify to what extent one group follow and interact with each other and the extent to which these connections and interactions reflect in inequalities in Twitter. Our analysis shows that users identified as White and male tend to attain higher positions in Twitter, in terms of the number of followers and number of times in user's lists. We hope our effort can stimulate the development of new theories of demographic information in the online space.

Motivation & Objective

• To investigate whether gender and race influence social network visibility and connectivity on Twitter.
• To quantify how demographic groups (male/female, White/Black/Asian) connect and interact with each other on Twitter.
• To examine whether online social networks perpetuate offline social inequalities, particularly the 'glass ceiling' effect.
• To provide empirical evidence on how demographic factors shape visibility and influence in online social media.

Proposed method

• Collected a large-scale sample of 1,670,863 active Twitter users located in the U.S. with profile images.
• Used Face++ face recognition API to infer gender and race from profile pictures with high accuracy.
• Crawled user friendship lists and interaction data (e.g., mentions, retweets) to analyze network structure.
• Computed relative probabilities of connections and interactions between demographic groups compared to expected demographic distributions.
• Applied statistical modeling to assess deviations from expected proportions in intergroup connections.
• Used visualization techniques to represent intergroup connection patterns, highlighting over- and under-representation.

Experimental results

Research questions

• RQ1 Do gender and race affect a user’s visibility and follower count on Twitter?
• RQ2 How do different demographic groups (male/female, White/Black/Asian) connect and interact with one another on Twitter?
• RQ3 To what extent do network connections and interactions reflect systemic inequalities in online social media?
• RQ4 Is there evidence of a 'glass ceiling' effect for non-White and non-male users on Twitter?
• RQ5 How do self-loops and endogenous connections among demographic groups contribute to observed inequalities?

Key findings

• White male users have significantly higher follower counts and appear more frequently in user lists than other demographic groups, indicating structural advantage.
• Black and Asian male users experience a 'glass ceiling' effect, with lower visibility and follower counts compared to White males despite similar activity levels.
• White males receive disproportionately more positive connections (friendships and interactions) from White and Asian users, while Black users are underrepresented as friends of White males.
• Black males and females exhibit the highest self-loop rates (130% and 86% respectively), indicating strong in-group connectivity and potential network segregation.
• Asian users are underrepresented in both friendships and interactions across all groups, with negative relative probabilities for incoming links from other demographics.
• Interactions among Asian users are relatively higher than expected, but their connections with White users are below expected levels, suggesting limited integration into broader networks.