Presenting posts and other content on social media feeds in the order they are posted, without the use of an engagement algorithm.

By reducing exposure to content that is likely to fuel toxic division, reverse chronological feeds would reduce toxicity and conflict.

Guess et. al (2023) tested the effects of reverse chronological feeds on attitudes and behaviors using a three-month field experiment on Facebook and Instagram. A group of users on Facebook and Instagram were randomly assigned to either a reverse-chronological feed (treatment) or the platforms' default feed (control group). After three months, the researchers found that assignment to a reverse chronological feed reduced engagement as would be expected, but had no discernible effect on participant’s levels of polarization or self-reported political participation.

Social media feeds which feature engagement-based algorithms have been criticized for fueling toxicity and polarization by upranking content that sparks conflict and outrage. By not prioritizing content that elicits reactions, it is hypothesized that feeds with reverse chronological algorithms would cause less toxic division.

The engagement algorithm tested by Guess et. al. (2023) was Meta’s “default” algorithm for Instagram and Facebook during the time of the experiment (September-December of 2020). Given that companies all use different ranking algorithms, it is hard to judge definitively the utility of reverse chronological feeds at nudging users towards more congenial content online.

We should also point out that, even though this is a well designed study with a large sample size (>20k participants) and the researchers found no effect on attitudes, platform field experiments rarely observe effects on attitudes measured by surveys (whereas observing behavioral changes is more common); so the absence of an observed effect may be due to the choice to use survey outcomes.