Evaluating Citizen Participation in Local Public Meetings: Exploring a Large Language Model Approach Using Transcripts from the United States

Abstract

Citizen participation in local public meetings is crucial for planning decision-making processes. This study employed large language models (LLMs), specifically ChatGPT models, to analyze more than 4,000 transcripts of local planning public meetings in the United States between 2006 and 2023. We quantify citizen participation levels and explore their relationship with public meeting topics. Findings align with previous scholarship that local public meetings do not consistently result in citizen empowerment. We also identify actionable strategies—such as fostering solution-oriented discussions and increasing civic organization involvement—that can enhance participatory planning. These insights suggest data-driven approaches for inclusive and equitable planning processes.