Prosocial Ranking Challenge

Overview
Rethinking Social Media Algorithms for a Healthier Digital Public Sphere
At Civic Health Project, we recognize that the way social media ranks and amplifies content shapes civic discourse, public trust, and social cohesion. The Prosocial Ranking Challenge (PRC) is an innovative projectthat tests alternative ranking algorithms to determine whether different content prioritization strategies can lead to healthier civic and individual outcomes. Through a browser extension and thousands of participants, PRC examines how changes in ranking affect users’ attitudes, behaviors, and engagement with pluralistic content.

The Problem

Social media platforms predominantly use engagement-driven algorithms that reward polarizing, sensational, or emotionally charged content. While effective for capturing attention, these ranking systems can also:

Amplify affective polarization by prioritizing divisive content. Reduce exposure to diverse perspectives by reinforcing echo chambers. Erode democratic norms by surfacing outrage-driven narratives over constructive dialogue. Encourage virality over veracity, leading to misinformation spread.

Despite growing concerns, we lack independent, rigorous testing of alternative ranking models that can point social media companies to solutions that better balance engagement with civic health. The PRC aims to fill this gap by imagining, implementing, and empirically testing new ranking strategies to assess whether social media can be structured for better societal outcomes.

Our Approach: Experimenting with Algorithmic Rankings

PRC is a real-world experimental initiative that recruits thousands of users to test alternative social media ranking models via a custom browser extension. The project allows participants to experience different ranking algorithms in real time, providing a direct comparison of their impact on attitudes, behaviors, and discourse quality.

How it Works: 1️⃣ Submission & Selection – PRC invited teams to submit innovative ranking algorithms that reorder, add, or remove content from social media feeds to promote healthier engagement. 2️⃣ Implementation via Browser Extension – Finalist algorithms are deployed through a browser extension that modifies social media rankings without altering content itself. 3️⃣ Live Testing with Thousands of Users – Recruited participants experience different ranking models as they browse platforms like Facebook, Twitter/X, and Reddit. 4️⃣ Measuring Civic Impact – Participants complete surveys and behavioral tracking data is collected to determine how ranking changes influence: Intergroup understanding Reduction in affective polarization Engagement with prosocial content Support of democratic norms PrePrint Release: Summer 2025 Stay tuned for early results and key findings, which will be publicly available through a PrePrint in Summer 2025, providing the first large-scale, experimental evidence on the relationship between ranking algorithms and social cohesion. To follow the latest updates, visit: Prosocial Ranking Challenge

Key Features Live Testing on Existing Platforms – Unlike AI simulations or survey studies, PRC operates in real social media environments, testing ranking effects in real time. Rigorous Experimental Framework – PRC incorporates randomized testing, pre/post surveys, and behavioral tracking to assess ranking effectiveness beyond engagement metrics. Diverse Algorithmic Solutions – Teams explored ranking models based on: Trust-based prioritization (elevating credible sources). Engagement with cross-cutting perspectives (rewarding diverse discourse). Reduced amplification of outrage-driven content. Open-Source & Transparent – PRC’s findings, methodologies, and top-performing algorithms will be made publicly available, allowing policymakers, researchers, and platforms to adapt and implement solutions.

Why This Matters Now

With AI-generated content, misinformation, and political polarization increasing, social media urgently needs alternative ranking models that prioritize democratic resilience over division.

Social media remains a dominant force in shaping public discourse, yet current ranking systems optimize for engagement over truth, connection, or civic health. Regulation and self-governance efforts are lagging, leaving an urgent need for independent, research-driven solutions. Without intervention, existing algorithms will continue to amplify polarization, increasing societal fragmentation and distrust in institutions.

PRC is one of the first large-scale, empirical tests of how ranking algorithms can influence social cohesion. By demonstrating which ranking strategies foster healthier civic discourse, PRC paves the way for evidence-based recommendations that platforms, policymakers, and researchers can use to reimagine digital public spaces for the better.

Advancing scalable solutions to bridge divides, strengthen democracy, and foster civic trust.

Contact Us