The political effects of X's feed algorithm

Authors

Germain Gauthier (Bocconi University)
Roland Hodler (University of St.Gallen)
Philine Widmer (Paris School of Economics)
Ekaterina Zhuravskaya (École des Hautes Études en Sciences Sociales)

Abstract

A randomized field experiment conducted on X (formerly Twitter) in 2023 examined the political effects of the platform’s feed algorithm. Researchers assigned active US-based users randomly to either an algorithmic or chronological feed for 7 weeks, measuring political attitudes and online behaviour. The study found that switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump, and views on the war in Ukraine. The algorithm promotes conservative content and demotes posts by traditional media. Participants exposed to the algorithmic feed were more likely to follow conservative political activist accounts, which they continued to follow even after switching off the algorithm. In contrast, switching from the algorithmic to the chronological feed had no comparable effects on political attitudes, suggesting that initial exposure to the algorithm has persistent effects on users’ political attitudes and account-following behaviour.

Key Highlights

25%

US adults who report social media as their primary news source

One half say they at least sometimes get news from these platforms

Main · 2024
76%

X users initially on the algorithmic feed

The algorithmic feed ('For you') is the default on X; 24% were using the chronological feed

Design · 2023
4.7percentage points

More likely to prioritize Republican policy issues (inflation, immigration, crime)

Among participants initially on the chronological feed switched to the algorithmic feed.

Impact of feed algorithm on attitudes and behaviour · 2023
5.5percentage points

More likely to believe that the investigations into Trump are unacceptable

Among participants initially on the chronological feed switched to the algorithmic feed.

Impact of feed algorithm on attitudes and behaviour · 2023
7.4percentage points

Less likely to hold a positive view of Ukrainian President Volodymyr Zelensky

Among participants initially on the chronological feed switched to the algorithmic feed.

Impact of feed algorithm on attitudes and behaviour · 2023
3.7percentage points

More likely to follow any conservative account on X

95% CI: 0.5, 7.0; P = 0.025. Among participants initially on the chronological feed switched to the algorithmic feed.

Impact of feed algorithm on attitudes and behaviour · 2023
2.9percentage points

Conservative posts more likely to appear in the algorithmic feed

19.9% relative increase;

Impact of feed algorithm on attitudes and behaviour · 2023
9.3 vs 5.4%

Share of conservative content among political content for Democrats — algorithmic vs chronological feed

The algorithm amplifies conservative content even in the feeds of Democratic users.

Impact of feed algorithm on attitudes and behaviour · 2023
15.5percentage points

Posts from news organizations less frequent in the algorithmic feed

58.1% relative decrease;

Impact of feed algorithm on attitudes and behaviour · 2023
5.9percentage points

Posts from political activists more frequent in the algorithmic feed

27.4% relative increase;

Impact of feed algorithm on attitudes and behaviour · 2023
0.18standard deviations (s.d.)

More likely to follow conservative political activist accounts

Effect size expressed in standard deviations (s.d.), a measure of the magnitude of the effect relative to the variability across participants. The entire effect on political activist accounts was driven by conservative political activists.

Impact of feed algorithm on attitudes and behaviour · 2023
9.0percentage points

More posts from conservative accounts in the chronological feed after algorithmic exposure

60% relative increase; The effect persists in the chronological feeds of users who were exposed to the algorithmic feed.

Impact of feed algorithm on attitudes and behaviour · 2023
"

Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects.

Abstract
"

Switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump and views on the war in Ukraine.

Abstract
"

We found that the algorithm promotes conservative content and demotes posts by traditional media. Exposure to algorithmic content leads users to follow conservative political activist accounts, which they continue to follow even after switching off the algorithm, helping explain the asymmetry in effects.

Abstract
"

These results suggest that initial exposure to X's algorithm has persistent effects on users' current political attitudes and account-following behaviour, even in the absence of a detectable effect on partisanship.

Abstract
"

A quarter of US adults report social media as their primary news source, and one half say they at least sometimes get news from these platforms.

Main
"

Yet, the fact that switching off a feed algorithm does not affect users' political attitudes does not mean that algorithms have no political impact. If the initial exposure to the algorithm has a persistent effect on political outcomes, switching off the algorithm might show no effects despite its importance.

Main
"

The algorithm demotes accounts of traditional news media and promotes those of political activists. Posts from news organizations appear 15.5 percentage points (58.1%) less often in the algorithmic feed, whereas posts from political activists appear 5.9 percentage points more often (27.4%).

Impact of feed algorithm on attitudes and behaviour
"

In addition to promoting entertainment, X's feed algorithm tends to push more conservative content to users' feeds. Seven weeks of exposure to such content in 2023 shifted users' political opinions in a more conservative direction, particularly with regard to policy priorities, perceptions of the criminal investigations into Trump and views on the war in Ukraine.

Discussion
"

The effect is asymmetric: switching the algorithm on influenced political views, but switching it off did not reverse users' perspectives on policy priorities or current political issues.

Discussion
"

We show that exposure to algorithmically curated content led users to follow conservative activist accounts. In contrast, when the algorithmic feed was switched off, users continued to follow the accounts they had engaged with previously. This indicates that exposure to feed algorithms has a lasting impact on users' feeds and their political attitudes.

Discussion
"

Overall, we conclude that social media feed algorithms can play an important role in shaping political opinions and online behaviour.

Discussion

Sources

Notice something missing or incorrect?
Suggest changes on GitHub