Abstract

Purpose: This study critically examines the role of big data analytics (BDA) in predicting social and economic vulnerabilities, highlighting methodological strengths, limitations, and ethical implications. While BDA offers unprecedented granularity and timeliness, its deployment for socio-economic risk prediction raises questions about data representativeness, algorithmic bias, and interpretability.

Design/Methodology: A doctrinal qualitative methodology was adopted, synthesizing insights from peer-reviewed literature, policy frameworks, and case studies across global contexts. The study critically interrogates the theoretical foundations of social vulnerability indices, the operationalization of economic vulnerability, and the efficacy of predictive models derived from mobile, administrative, and transactional big datasets.

Findings: The analysis reveals that BDA can enhance predictive accuracy for socio-economic vulnerabilities, particularly when integrating heterogeneous data sources. However, gaps remain in addressing structural biases, data scarcity in marginalized populations, and the limitations of existing computational frameworks to capture contextual socio-economic dynamics. Ethical and policy challenges are accentuated where predictive models may inadvertently exacerbate inequalities.

Originality/Value: This paper contributes a critical synthesis that bridges technological capability with social science theory, emphasizing the dual potential of BDA to inform policy and the inherent risks of reinforcing existing vulnerabilities. It advocates for a nuanced, context-sensitive approach that incorporates ethical safeguards and participatory validation in predictive modeling.