Carnegie Mellon University

Causal model search applied to economics: gains, pitfalls and challenges: Alessio Moneta

Abstract: This talk assesses the contribution that causal model search brings to empirical economics by examining a case study in industrial economics. In particular we investigate the link between exporting activity and firm performance. Here a debated question is whether trade increases firm productivity or it is only more productive firms that enter and remain in the export market. Settling such a question would contribute to the understanding of the causes and consequences of international trade. We tackle this issue by applying causal model search methods to a data set on exporting firms in Chile. These methods consist of a combination between graphical causal models, or, alternatively, independent component analysis, and structural vector autoregressive models. The presentation assesses the usefulness and adequateness of these methods to provide concrete answers to the question at hand and the potentiality to address similar issues. At the same time it tries to identify what are the drawbacks and pitfalls of these methods when they are applied to disentangle typical causality issues in economics. This opens up challenges for future research.