Daily interactions with right-leaning pages on Facebook fell only 6% in the 20 days post-election vs. 20 days before despite reports of an adjusted algorithm
After the election, the platform had reportedly decided to elevate established news outlets and reduce visibility of right and left-leaning pages. Tweets: @peltzmadeline See also Mediagazer Tweets: Madeline Peltz / @peltzmadeline : We hear a lot about the right's dominance of Facebook's daily top tens. @ohhkaygo is doing groundbreaking research showing that so far Zuckerberg's algorithm adjustment has failed to mitigate this harm. https://www.mediamatters.org/ ... See also Mediagazer
Context & Ripple Effects
This measurement lands on top of two threads in the related coverage: former Facebook employees had already described News Feed tests whose tweaks favored right-leaning sites, and Media Matters' own 2018-19 data showed right-leaning pages with over 500,000 likes consistently outperforming left-leaning rivals. The post-election adjustment — elevating established outlets, demoting partisan pages — was the company's visible response to exactly that documented imbalance.
The finding that daily interactions fell only 6% echoes an earlier pattern: in the [[a:933472|study tracking fake-news publishers' engagement dropping by half on Facebook between 2016 and mid-2018]], the platform-side decline coexisted with continued sharing elsewhere. Announced curation changes and measured reach are again diverging, now with a watchdog quantifying the gap within weeks.
First-order effects
- Right-leaning page operators retain essentially their pre-election audience on Facebook, since a 6% dip leaves the daily-interaction base Zuckerberg's adjustment was reportedly meant to shrink largely intact.
- Facebook's stated fix is directly contradicted by its own platform data, handing Media Matters and researcher Madeline Peltz a concrete benchmark against which any future algorithm claims will be checked.
Second-order effects
- The gap between announcement and outcome raises the cost for Facebook of self-reported moderation claims, pushing journalists and academics toward continuous third-party measurement rather than relying on company statements.
- Established news outlets that were supposed to be elevated under the same adjustment see little evidence the redistribution actually happened, weakening the incentive to treat News Feed placement as a reliable partner channel.
Third-order effects
- If near-real-time external audits keep showing announced changes failing to move distribution, algorithmic curation shifts from a private governance tool toward a regulated accountability surface — the distribution layer itself becoming the liability.
- The recurring pattern across 2018's fake-news engagement decline and 2020's post-election adjustment points to structural inertia: whatever the policy intent, the engagement economics of high-participation partisan pages resist algorithmic suppression.
The trend: Platform algorithm adjustments aimed at rebalancing political news distribution are increasingly judged by independent, real-time measurement — and so far the measured reach of right-leaning pages has outlasted every announced correction.