The Ambivalent Role of Punishment in Post-Merger Coordination and the New Approach of the 2010 US Horizontal Merger Guidelines

Abstract

Threat of punishment is considered a deterrent force, which ensures the sustainability of post-merger tacit coordination. US and EU Agencies, Courts and economists, including those who extend simulation methods to quantify coordinated effects, assume that parties follow grim trigger strategies: e.g. they choose the collusive output until one firm deviates, and produce the competitive output thereafter. More complex game theory insights indicate that collusion mechanisms can never be as perfect and simple in reality. Punishment plays an ambivalent role in post-merger coordination. More realistic scenarios include long phases of collusion that are interrupted by shorter phases of price wars, which remind players of their interdependence and force them to renegotiate. Moreover, punishment is part of an interactive learning process, in which players experiment with competition parameters with a view to identifying optimal strategies. Punishment does not cause parties to abide by concrete patterns of coordination; it primarily helps them to identify focal points, realize interdependence, and fine-tune the terms of tacit coordination. It is higher payoffs and the bundle of firm and market characteristics that render sustainable collusion more likely. The defections of collusion-cheating-punishment mechanics might have induced US Agencies to introduce in their 2010 Horizontal Merger Guidelines the theory of parallel accommodating conduct, which relies on individual strategic reasoning not motivated by retaliation.