Socioeconomic Determinants of the COVID-19 Infodemic

TL;DR

The paper investigates how socioeconomic determinants shape the COVID-19 infodemic across 37 OECD countries using Twitter-derived metrics. It combines 20 indicators with dimensionality reduction (UMAP, PCA) and clustering (k-means) to map a multidimensional socioeconomic space and to identify temporal infodemic patterns. Key findings show that initial infodemic dynamics strongly relate to socioeconomic context, while later phases see stronger links between overall infodemic risk and structural factors;News media diet diversity serves as a protective mediator, and institutional stability dampens volatility. Collectively, the results highlight the importance of pluralistic information ecosystems and robust institutions for enhancing societal resilience to misinformation during health crises.

Abstract

The COVID-19 pandemic has been accompanied by an infodemic of misinformation that impedes effective public health responses. This study examines relationships between socioeconomic factors and infodemic risk patterns across 37 OECD countries using Twitter data from 2020-2022. Employing dimensionality reduction techniques on 20 socioeconomic indicators, we identify complex correlations with infodemic measures that evolve throughout the pandemic. Countries exhibit distinct clustering in their infodemic profiles that transcend conventional socioeconomic categorizations. We find that dynamic information behaviors dominate initial crisis responses, while stable socioeconomic conditions become more influential as the pandemic progresses. News media diet diversity emerges as a significant protective factor, with pluralistic information ecosystems demonstrating greater resilience against misinformation. Additionally, institutional stability correlates strongly with reduced infodemic volatility over time. These findings highlight how infodemics are embedded within broader socioeconomic contexts, providing foundations for targeted interventions to build societal resilience against misinformation during future health emergencies.

Figures (6)

• Figure 1: UMAP visualization of OECD countries and the socioeconomic indices UMAP visualization of the OECD* countries based on the selected indices (left) and of the selected socioeconomic indices (right). The employed hyperparameters are n neighbors = 5 and min dist = 0.01.
• Figure 2: UMAP projection of the IRI time series of the countries featuring at least 500’000 tweets in the relevant time frame. The selected parameters are n neighbors = 5 and min dist = 0.1. The 9 countries belonging to Cluster 1 are grouped on the left, whereas the remaining ones belong to Cluster 2.
• Figure 3: k-means clustering with Euclidean distance and UMAP preprocessing on the IRI time series for the 88 selected countries. The UMAP parameters employed were n neighbors = 5, min dist = 0.1 and n components = 102. The cluster average is plotted in red. All the time series are smoothed using a 30-day rolling average.
• Figure 4: The first panel shows the Spearman correlation of the first Principal Component and the monthly average of the infodemic metrics over time for OECD* countries. The second panel displays the p-value of the corresponding correlation with a black dashed line representing the significance level of 5%.
• Figure 5: Correlation between the news media diet and the UMAP of the socioeconomic indices. Panels A and B show the correlation between the socioeconomic indicators with the entropy of all the domains classified as low-risk (satire, clickbait and political) with a spearman correlation of 0.49 in 2020 and 0.52 in 2021. Panels C and D represent the correlation between the socioeconomic indices and the news classified as political, with a spearman correlation value respectively of 0.48 and 0.54.