Facebook Political Ad Delivery Algorithms Inferred Users' Political Alignment, Inhibiting Political Campaigns' Reach

Researchers conducted a study investigating Facebook's ad delivery algorithms and their impact on political advertising during the lead-up to the 2020 U.S. presidential elections. The study involved running a series of political ads on Facebook to measure how the platform's algorithms delivered those ads to different demographic groups based on ad content and targeting criteria. The researchers found that Facebook's ad delivery algorithms effectively differentiated the price of reaching users based on their inferred political alignment with the advertised content. This algorithmic behavior inhibited political campaigns' ability to reach voters with diverse political views, with the effect being most pronounced when advertisers used small budgets. The algorithms tended to preferentially deliver ads to users who Facebook estimated were most relevant to the content, even when opposing viewpoint advertisers were actively trying to reach them. Due to Facebook's current ad performance reporting methods, this effect was largely invisible to political campaigns. The findings suggest that advertising platforms may play a role in political polarization by creating informational filter bubbles, where users are primarily exposed to political content that aligns with their existing beliefs.