Media sentiment and government healthcare investment : a time-series analysis of GDELT news data

Abstract

Media play an important role in public decision-making, but empirical research on whether media tone influences actual government investment in public healthcare remains limited. This study uses news data from the Global Database of Events, Language, and Tone (GDELT) and data on current health expenditure expressed as a percentage of GDP across countries to examine the relationship between media tone and government public healthcare investment across a 20 years time-series. Using linear mixed-effects models with lagged media variables and controls for economic conditions and major public health events, we analyze convergent evidence from two independent GDELT datasets spanning 2004–2024. We find no significant media effect on healthcare expenditure. Instead, a long-term upward trend in spending and COVID-19 case burden emerge as the primary predictors, suggesting that healthcare investment decisions are driven by policy trajectories and direct public health threats rather than media framing.