An in-depth look at Facebook's role in eroding the informational underpinnings of democracy during the 2016 election, and why it was so hard to see it coming
Alexis C. Madrigal / The Atlantic :
Context & Ripple Effects
This Atlantic feature was the early, systematic account of what Facebook's targeting and distribution machinery had done to the 2016 information environment — written before the company itself conceded anything. Within months that changed: Facebook made an admission with little precedent among tech firms that its product hurts democracy, an acknowledgment that handed arguments to the very critics this piece catalogued (the company's own democracy-harm admission).
The arc since has run through Cambridge Analytica's voter-profiling models built on 2014 data, a midterm-eve analysis of 250K posts showing how little had been fixed by 2018 (Jonathan Albright's post-and-ad audit), and insider testimony — first from a former head of election integrity operations blaming the business model itself (Facebook's own election-integrity chief on why it can't stop harming democracy), then from Biden staffers who left the 2020 cycle convinced the platform threatens elections (Biden staffers' concerns after dealing with Facebook). This piece is the origin point those later accounts keep returning to.
First-order effects
- Facebook's ad-targeting and distribution tools are reframed from neutral infrastructure as active vectors for political misinformation, putting the platform's core product decisions — not just content moderation — at the center of the criticism.
- Political operatives and campaigns gain evidence that microtargeted advertising can shape voter perception faster than journalists or fact-checkers can respond, changing how they evaluate the channel.
Second-order effects
- Advertisers and publishers face pressure as brands weigh association with a platform whose civic externalities are now documented, while regulators acquire a concrete record of harms to justify oversight of targeted political ads.
- Rival platforms inherit the scrutiny: any social network selling granular political targeting must now answer for the same failure mode Facebook exposed.
Third-order effects
- If the pattern holds, the industry moves toward structural constraints on political ad targeting — disclosure regimes, audience-size limits, or outright restrictions — because the harm traces to the business model rather than to individual bad actors.
- Platform governance shifts from reactive takedowns to pre-election institutional review, with companies building standing election-integrity operations and outside researchers auditing them, as the recurring cycle of 2016-to-2020 coverage shows.
The trend: Social platforms are being redefined from neutral distribution pipes to accountable political actors, with each election cycle producing deeper documentation that their business models themselves carry democratic risk.