Recent Advances In Causal and Computational Methods for Stratification

This proposed session brings together research that develops or applies causal inference methods, including computational and machine-learning approaches, to questions of social stratification and inequality. The session invites submissions that either advance causal methods for inequality research or apply existing tools to generate new evidence on stratification processes. Recent advances in machine learning and artificial intelligence, along with the growing availability of large-scale administrative, textual, and digital-trace data, are creating new opportunities to address causal questions about how inequalities are produced and transformed. Relevant topics may include challenges in causal identification and estimation, computational and machine-learning approaches to causal inference, the use of text-as-data or digital-trace data in causal analysis, the role of generative AI in causal research, and other methodological extensions. We especially welcome research showing how these approaches can generate robust causal evidence on dimensions of inequality that are difficult to examine using existing methods alone. By bringing together methodological and substantive contributions, this session aims to foster dialogue on how emerging causal approaches can expand the empirical study of inequality while maintaining attention to theory, identification, measurement, and interpretation.