Gender Bias in YouTube Exposure: Allocative and Structural Inequalities in Political Information Environments

2026-04-30

AI Summary

This study investigates whether gender-based disparities exist in political content exposure within YouTube’s recommendation algorithm. By deploying male- and female-coded social bots in a controlled field experiment and integrating collaborative filtering models, co-occurrence networks, time-series analysis, and clustering techniques, the research uncovers— for the first time—the dual mechanisms of algorithmic bias: distributive (differences in the distribution of topics, ideologies, and political actors) and structural (gendered clustering in recommendation networks). The findings demonstrate that accounts associated with different genders receive significantly divergent political content, and that the recommender system’s design inherently reproduces and amplifies these inequalities, highlighting the algorithm’s potential role in exacerbating societal biases.

Abstract

Recommendation algorithms have become the dominant mechanism for information distribution on digital platforms, profoundly shaping personalized information consumption environments. However, gender bias, as a significant form of algorithmic discrimination, may cause users to experience unequal exposure within different political information environments. Taking YouTube as a case, we conduct a controlled social-bot field experiment, where male-coded and female-coded profiles are constructed. We track the exposure and click patterns of these bots to analyze their recommendation trajectories. We analyze the distribution of recommended content from two dimensions: allocative bias and structural bias. First, we find statistically significant differences in allocative bias across male-coded and female-coded profiles, particularly in terms of issue distribution, ideological orientation, and political entities. Secondly, we observe structural bias in the political information environments, characterized by distinct clustering patterns. Additionally, time-series analysis shows that exposure pathways continue to be shaped over time by both communities detected in the co-occurrence network and individual profile-level dynamics. Finally, we construct a simple collaborative-filtering model that reproduces the observed gender bias. We argue that gender bias in recommendation systems is reflected not only in the allocation of political content, but also in how community structures shape these environments, reinforcing societal inequalities and highlighting the need for algorithmic fairness.

Problem

• gender bias
• recommendation algorithms
• allocative bias
• structural bias
• political information environments

Innovation

• gender bias
• recommendation algorithms
• allocative bias
• structural bias
• social-bot experiment