Democracy Discourse Index

The World's First Real-Time Tool for Measuring Democratic Health Through Public Discourse

Democracy lives or dies in the quality of its discourse. The Democracy Discourse Index (DDI) measures not what institutions do, but how citizens experience democracy through their daily conversations. By analyzing billions of digital interactions in real-time, we provide early warning systems for democratic decay and pathways for renewal.

The DDI is implemented in partnership with Sensika Technologies, as part of our joint Disinformation Observatory.

100M+ Daily Interactions | 20+ Languages | 300+ Researchers

Our Core Philosophy

Democracy thrives not only in institutions, but in the quality of public discourse. Language is not neutral. It constructs power relations and reveals moral choices.

Redefining How We Measure Democracy

  • What We Measure

    Traditional democracy indices measure institutions, elections, and constitutions. They tell us what happened. The DDI tells us what's happening—and what's coming.

  • Why It Matters

    Language reveals power relations and social dynamics. Trust manifests through how we speak to each other. Toxicity in discourse precedes institutional breakdown. Democratic culture lives in everyday conversation. Early warning signals appear first in public discourse.

  • AI-Powered Analysis

    Natural Language Processing analyzes billions of interactions daily, detecting patterns, sentiment, and discourse quality at unprecedented scale.

  • How AI meets Critical Discourse Analysis

    Built on cutting-edge AI and grounded in Critical Discourse Analysis, the DDI captures the lived experience of democracy as expressed by citizens in digital spaces. We measure the moral and informational health of democratic life in real-time.

  • Why the World needs a DDI

    In an era where a single viral conspiracy can topple governments and where social media shapes reality more than parliaments, measuring institutional structures alone is like taking democracy's temperature with a broken thermometer.

Six Dimensions of Democratic Discourse

Primary Research & Corpus Development

The heart of the project lies in human-coded content analysis coordinated through a consortium of universities across diverse democratic contexts. Faculty-led student research teams will analyze social media discourse (X, Facebook, YouTube, TikTok, and news comment threads) across identified themes: trust, governance, gender, migration, elections, and civic activism.

Process:

  1. Sampling: Each university will extract datasets from national discourse in using our AI platform.

  2. Coding: Student coders, under faculty supervision, will classify posts and comments against the DDI dimensions using structured coding protocols.

  3. Query Refinement: Insights from manual coding will refine AI keyword models, improving semantic accuracy and bias reduction.

  4. Model Training: Human-coded data will train our AI models, generating machine-learning classifiers capable of automated scaling.

  5. Quality Assurance: Inter-coder reliability tests, supervised review panels, and ethics audits will maintain rigor.

Click here if you want to bring the DDI to your classroom or university?

Project Objectives

  • Evidence for Trust-Building Measures

    Provide democratic leaders with real-time data on the status and quality of public discourse to inform evidence-based trust-building measures

  • Policy Intelligence and Early Warning Systems

    Detect democratic backsliding 6-12 months before institutional crises, enabling preventive intervention

  • Academic Research

    Generate unprecedented datasets for studying democratic resilience and discourse dynamics

  • Civil Society Tools

    Empower NGOs and activists with data to strengthen inclusive narratives and build social cohesion

Contact us to find out more