Opening the Black Box: Uncovering the European Commission’s Cartel Fining Formula through Computational Analysis

Abstract

Is the European Commission biased against non-EU and in particular U.S. firms when enforcing competition law? This ‘protectionist hypothesis’ has been rejected in the area of merger control. In the areas of abuse of dominance and restrictive agreements, however, few researchers have empirically tested the Commission’s decisional practice for protectionism. The difficulty is that, despite increased transparency since the 2006 Fining Guidelines, the Commission’s fine-setting process remains opaque, which is why it has been likened to a ‘black box’. This article opens that black box by computationally assessing all cartel fines based on the Fining Guidelines from 2006 until 2020. For the tenures of Commissioners Almunia and Vestager, and for some steps of the fine calculation, we find a fining methodology so consistent it can be considered a ‘fining formula’. This formula shows no clear evidence of the protectionist hypothesis.

Keywords: cartel, European Commission, fine calculation, machine learning, computational antitrust

JEL Classification: C53, C63, K21, K42, L41