Twitter's algorithm boosts right-leaning content, internal study finds

Twitter’s timeline algorithm tends to amplify right-leaning politicians and news outlets more than their counterparts on the left, a study conducted by the platform found.

The company analyzed millions of tweets from elected officials in seven countries — the United Kingdom, U.S., Canada, France, Germany, Spain and Japan — and hundreds of millions of tweets from news outlets between April 1 and Aug. 15, 2020.

Twitter used third-party sources to categorize both politicians and news sources along the political spectrum.

ADVERTISEMENT

In every country but Germany, tweets posted by accounts on the political right received more amplification than those on the political left.

Right-leaning outlets similarly outperformed left-leaning ones, although that analysis was limited to the U.S.

The research did not seek to address the reasons why the disparities exist, but the platform plans to tackle that matter down the line.

“Algorithmic amplification is problematic if there is preferential treatment as a function of how the algorithm is constructed versus the interactions people have with it,” Rumman Chowdhury, the head of Twitter's machine learning, ethics, transparency and accountability team, and Luca Belli, a Twitter researcher, wrote in a blog post. “Further root cause analysis is required in order to determine what, if any, changes are required to reduce adverse impacts by our Home timeline algorithm.”

Twitter first started offering its algorithm-determined timeline in 2016 after starting with a purely chronological one. Users can now choose between the two ways to organize posts.

The platform is offering to share the raw data behind the study with third-party researchers upon request.