Simulating the (Unilateral) Effects of Mergers: Implications of the Oracle/Peoplesoft Case

Abstract

Unilateral effects analysis has been the single most important innovation in merger control driven by economic theory in recent years. In practice, this requires the use of sophisticated economic tools such as simulation models, which facilitate the quantification of short-term merger effects. In this paper, we study the application of this new device in the landmark merger case between software firms Oracle and PeopleSoft. Despite serious initial concerns the transaction was not blocked, nor even required to be modified in the EU and US. Besides some case-specific issues we highlight the fundamental problems associated with the use of sophisticated economic tools in merger control. The basic questions are: (i) What can merger simulation models actually demonstrate? (ii) What is an adequate standard of proof for their sensible use in merger review? Based on the nature of predictive economic evidence we then develop an adequate standard of proof, i.e. the evidentiary threshold that must be met for a prohibition decision. For a more general treatment we also present a simple analytical model in order to illustrate the effects of different thresholds on legal and economic quality. Finally, our recommendation is to adjust the standard of proof in the direction of a moderate threshold or, more preferably, to switch to a mixed burden of proof. Both suggestions are geared to guaranteeing the fruitful contribution of sophisticated economic evidence to case assessments and, thus, increasing decision quality.

Keywords: Unilateral effects, simulation models, merger control, standard of proof, dynamic competition

JEL Classification: F02, K21, L41