Rural‐Urban Differences in the Determinants of Subjective Well‐Being Among X/Twitter Users in the United States

Abstract

Twitter Sentiment Geographical Index (TSGI) has been proposed to complement traditional surveys to measure subjective well‐being (SWB) at the US county level. Our study aims to investigate determinants of TSGI‐measured SWB in rural and urban US counties. Using the Classification and Regression Tree, we identified phenotypes or county‐level characteristics associated with high SWB. Counties with newer homes were the top characteristic of high SWB in both urban and rural areas. Counties of the identical phenotypes tend to concentrate geographically, with the most favorable phenotypes clustered in the South. Random Forest analysis identified additional characteristics of high SWB, including higher population density in rural areas and lower real estate tax ratio in urban areas. Our results yield a comprehensive understanding of determinants of SWB at the local level, guiding evidence‐based policy decisions and community initiatives to improve well‐being in target populations.