The heterogeneous impact of the EU-Canada agreement with causal machine learning

Lionel Fontagné, Francesca Micocci () and Armando Rungi
Additional contact information
Francesca Micocci: IMT - School for Advanced Studies Lucca

Université Paris1 Panthéon-Sorbonne (Post-Print and Working Papers) from HAL

Abstract: This paper introduces a causal machine learning approach to investigate the effects of free trade agreements and applies it to the EU-Canada Comprehensive Economic and Trade Agreement (CETA). Previous estimates of the impact of trade liberalization have been found to be unstable and contradictory, possibly due to the presence of heterogeneous treatment effects. The matrix completion estimator computes multidimensional counterfactuals in trade data at the firm, product, and destination levels. Compared with other estimators, it relies on a weaker exogeneity assumption and a more general functional form. In the case of CETA, we obtain both positive and negative idiosyncratic treatment effects at the product-destination level, although the sales-weighted average treatment effect is 6.4% in the year after the agreement. At the same time, we can estimate idiosyncratic treatment effects for the extensive margin at the product-destination level; thus, we find product churning beyond regular entry-exit dynamics: 8.1% that were not previously exported, and about 7.3% that are no longer exported. Finally, we consider the case of multiproduct firms after ranking product portfolios. After CETA, we observe a reallocation of French exports toward the first and most exported products, possibly driven by increased competition in the local market by other European producers after trade liberalization.

Keywords: Causal Inference; Machine Learning; Matrix Completion; International Trade; Free Trade Agreements (search for similar items in EconPapers)
Date: 2026-07
New Economics Papers: this item is included in nep-big, nep-cmp, nep-eec and nep-int
Note: View the original document on HAL open archive server: https://shs.hal.science/halshs-04913313v2
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://shs.hal.science/halshs-04913313v2/document (application/pdf)

Related works:
Working Paper: The heterogeneous impact of the EU-Canada agreement with causal machine learning (2026)
Working Paper: The heterogeneous impact of the EU-Canada agreement with causal machine learning (2026)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:cesptp:halshs-04913313

Access Statistics for this paper

More papers in Université Paris1 Panthéon-Sorbonne (Post-Print and Working Papers) from HAL
Bibliographic data for series maintained by CCSD ().