Grievance Politics vs. Policy Debates: A Cross-Platform Analysis of Conservative Discourse on Truth Social and Reddit

A team of six researchers from institutions including Indiana University and University College London conducted the first large-scale comparative analysis of conservative discourse on Truth Social versus Reddit communities like r/Conservative. Using advanced AI techniques—specifically FASTopic for topic modeling combined with LLM-assisted refinement—they analyzed content from the first eight months of Truth Social's existence. Their methodology allowed them to track topic prevalence, measure toxicity levels, and observe temporal dynamics across these parallel online ecosystems.

The findings reveal a stark platform-driven divergence in political discourse. Truth Social consistently centers on grievance-oriented, narrative-driven content, functioning more as an amplifier of shared political identity. In contrast, Reddit's conservative communities engage more substantively with policy debates and legislative details. Interestingly, while both platforms showed similar event-driven participation spikes, Reddit demonstrated 23% higher toxicity overall, with peaks occurring specifically in cultural and leader-focused discussions. The research team's LLM-assisted analysis proved crucial for accurately categorizing nuanced political topics that traditional algorithms might miss.

This study provides concrete, data-driven evidence that platform architecture—from algorithms to community norms—fundamentally reshapes political conversation. The researchers note that Truth Social users showed 57% higher baseline engagement with political topics compared to their Reddit counterparts, suggesting different user intentions and platform affordances. Accepted for publication at ICWSM 2026, this work establishes a new methodological standard for cross-platform political analysis using contemporary AI tools.