The Larger the Tip, the Smaller the Iceberg: Reassessing Anti-cartel Policy Evaluation

Anti-cartel policy evaluation traditionally relies on quantifying the harm caused by detected cartels as a measure of its effectiveness. This approach implicitly `assumes a positive correlation between the detection effect and the societal value of antitrust policy. This paper argues that there may be a salience problem when conducting anti-cartel policy, potentially leading to biased assessments. Using a two-period model, we show that, under certain conditions likely to be met, a larger detection effect is associated with a smaller welfare gain from anti-cartel policy. This counterintuitive result arises because antitrust policy's welfare gains extend beyond the detection effect and include the deterrence effect, which is amplified by effective anti-cartel policy but can reduce the detection effect, i.e., there is usually a negative relationship between the detection and the deterrence effect. Consequently, contrary to common practice, metrics to assess anti-cartel policies must explicitly incorporate the deterrence effect. Failing to consider such effect may lead to erroneous conclusions, where an increase in the detection effect may mistakenly be associated with an improvement in anti-cartel policy. Considering these findings, this paper underscores the necessity for a comprehensive assessment framework that acknowledges the interplay between the detection and deterrence effects.