Local Government Accountability to Citizens as COVID-19 Took Hold: Sentiment Analysis of Tweets in Hispanic–American Countries

Abstract

This study examines the impact of COVID-19 on the interaction between citizens and local governments through Twitter and conducts sentiment analysis within the accountability process guided by dialogical theory. An open-source intelligence tool extracted 1,224,725 tweets from 126 Hispanic–American municipalities during the first wave. Statistical methods identified tweet publication factors, and the NRC Emotion Lexicon was used for natural language processing. The results demonstrate that Twitter activity increased during the crisis, both reflecting changes in accountability dynamics and heightened dialogue between citizens and local governments and highlighting negative tones and emotions such as trust and anticipation.