Building Social Early Warning System (SEWS): Predicting Social Unrest Through Economic Early Warnings

Abstract

This study develops a Social Early Warning System (SEWS) to predict unrest by analyzing economic indicators in Jordan. Using the Standardized Index of Social Unrest (SISU), it evaluates key triggers including oil prices, remittances, inflation, income growth, and unemployment. By combining data analytics with historical case validation, the model demonstrates an 83.9% accuracy in forecasting tranquil periods and certain unrest episodes. Although limited in detecting localized events, SEWS offers a practical tool for policymakers to anticipate instability. The study recommends incorporating grassroots data and refining algorithmic thresholds to enhance responsiveness.