# Algorithmic Amplification of Politics on Twitter

arXiv: https://arxiv.org/abs/2110.11010 · HTML: https://mlsanity.com/paper/2110.11010/

## Key facts

- Cited by 38 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Oct 2021.

## Built on by (most cited first, 30 of 38)

- [Simulating Social Media Using Large Language Models to Evaluate Alternative News Feed Algorithms](https://mlsanity.com/paper/2310.05984/) (2023)
- [Handling Bias in Toxic Speech Detection: A Survey](https://mlsanity.com/paper/2202.00126/) (2022)
- [Diminished Diversity-of-Thought in a Standard Large Language Model](https://mlsanity.com/paper/2302.07267/) (2023)
- [Modeling Content Creator Incentives on Algorithm-Curated Platforms](https://mlsanity.com/paper/2206.13102/) (2022)
- [AI Alignment with Changing and Influenceable Reward Functions](https://mlsanity.com/paper/2405.17713/) (2024)
- [Linear Representations of Political Perspective Emerge in Large Language Models](https://mlsanity.com/paper/2503.02080/) (2025)
- [Human-AI Coevolution](https://mlsanity.com/paper/2306.13723/) (2023)
- [Embedding Democratic Values into Social Media AIs via Societal Objective Functions](https://mlsanity.com/paper/2307.13912/) (2023)
- [From Transparency to Accountability and Back: A Discussion of Access and Evidence in AI Auditing](https://mlsanity.com/paper/2410.04772/) (2024)
- [Algorithmic audits of algorithms, and the law](https://mlsanity.com/paper/2203.03711/) (2022)
- [The Accountability Paradox: How Platform API Restrictions Undermine AI Transparency Mandates](https://mlsanity.com/paper/2505.11577/) (2025)
- [Evolving linguistic divergence on polarizing social media](https://mlsanity.com/paper/2309.01659/) (2023)
- [Down the Rabbit Hole: Detecting Online Extremism, Radicalisation, and Politicised Hate Speech](https://mlsanity.com/paper/2301.11579/) (2023)
- [On the rise of fear speech in online social media](https://mlsanity.com/paper/2303.10311/) (2023)
- [Harm Mitigation in Recommender Systems under User Preference Dynamics](https://mlsanity.com/paper/2406.09882/) (2024)
- [Hatemongers ride on echo chambers to escalate hate speech diffusion](https://mlsanity.com/paper/2302.02479/) (2023)
- [Upvotes? Downvotes? No Votes? Understanding the relationship between reaction mechanisms and political discourse on Reddit](https://mlsanity.com/paper/2302.09540/) (2023)
- [Large Language Models Hack Rewards, and Society](https://mlsanity.com/paper/2606.04075/) (2026)
- [Dynamics of Algorithmic Content Amplification on TikTok](https://mlsanity.com/paper/2503.20231/) (2025)
- [Datasets for Navigating Sensitive Topics in Recommendation Systems](https://mlsanity.com/paper/2509.07269/) (2025)
- [A User-Driven Framework for Regulating and Auditing Social Media](https://mlsanity.com/paper/2304.10525/) (2023)
- [Learning Unbiased News Article Representations: A Knowledge-Infused Approach](https://mlsanity.com/paper/2309.05981/) (2023)
- [Drift and selection in LLM text ecosystems](https://mlsanity.com/paper/2604.08554/) (2026)
- [Constructive Alignment: Governing Preference Dynamics in Human-AI Interaction](https://mlsanity.com/paper/2607.00001/) (2026)
- [Fair Algorithm Design: Fair and Efficacious Machine Scheduling](https://mlsanity.com/paper/2204.06438/) (2022)
- [Error in the Euclidean Preference Model](https://mlsanity.com/paper/2208.08160/) (2022)
- [The Meme Is the Message: Generative Memesis and AI Visuals in the 2024 USA Presidential Elections](https://mlsanity.com/paper/2411.00934/) (2024)
- [Fairness-in-the-Workflow: How Machine Learning Practitioners at Big Tech Companies Approach Fairness in Recommender Systems](https://mlsanity.com/paper/2505.19441/) (2025)
- [AI-Mediated Communication Reshapes Social Structure in Opinion-Diverse Groups](https://mlsanity.com/paper/2510.21984/) (2025)
- [Metanetworks as Regulatory Operators: Learning to Edit for Requirement Compliance](https://mlsanity.com/paper/2512.15469/) (2025)

## Abstract (arXiv)

Content on Twitter's home timeline is selected and ordered by personalization algorithms. By consistently ranking certain content higher, these algorithms may amplify some messages while reducing the visibility of others. There's been intense public and scholarly debate about the possibility that some political groups benefit more from algorithmic amplification than others. We provide quantitative evidence from a long-running, massive-scale randomized experiment on the Twitter platform that committed a randomized control group including nearly 2M daily active accounts to a reverse-chronological content feed free of algorithmic personalization. We present two sets of findings. First, we studied Tweets by elected legislators from major political parties in 7 countries. Our results reveal a remarkably consistent trend: In 6 out of 7 countries studied, the mainstream political right enjoys higher algorithmic amplification than the mainstream political left. Consistent with this overall trend, our second set of findings studying the U.S. media landscape revealed that algorithmic amplification favours right-leaning news sources. We further looked at whether algorithms amplify far-left and far-right political groups more than moderate ones: contrary to prevailing public belief, we did not find evidence to support this hypothesis. We hope our findings will contribute to an evidence-based debate on the role personalization algorithms play in shaping political content consumption.