Thursday, April 26, 2007
Posted by D. Daniel Sokol
This week we had an excellent discussion in my class of international cartels. John Connor of Purdue's Agricultural Economics Department has been at the forefront of cartel research and I assigned one of his recent articles on the topic. John has a new working paper that suggests that cartel fines are under-deterring cartel behavior entitled A Critique of Cartel Fine Discounting by the U.S. Department of Justice.
ABSTRACT: This paper surveys declared Department of Justice policies on cooperation discounts on fines for corporate criminal price-fixing violations and develops and tests a statistical model to explain the actual discounting practices of the DOJ. The sample consists of 56 corporations that were fined for hard-core cartel behavior between 1996 and 2006 and for which reasonably accurate data on recommended fines can be found. In most respects there is consistency between principle and practice, but in three respects a divergence is observed. As promised, the DOJ does reward the second-in, third-in, and successive firms that agree to plead guilty with progressively smaller cooperation discounts. Discounts are larger for low-ranking (early-to-plead) firms, but for companies with the same rank discounts are larger when the number of cartel participants is small. A second reasonable finding is that the longer a guilty firm delays in coming to terms with the DOJ, the smaller will be its cooperation discount. Delay and moving down in the queue act independently and additively. If a firm delays its guilty plea by seven months and at the same time moves to fourth place and loses second place, the empirical model predicts that the fine tends to increase by 12 percentage points of affected sales. Three additional findings seem to point to inconsistent application of the DOJ's public policy on rewarding cooperation of guilty cartel participants. First, neither the size of a defendant's affected commerce nor the duration of its collusion raises or lowers cooperation discounts. Second, Asian and European defendants receive distinctly lower cooperation discounts than corporate defendants from North America; the effect averages 14 to 19 percentage points of the Guidelines' maximum recommended fine. Third, two measures of price-fixing recidivism are unrelated to the cooperation discounts of the sampled cartelists. The DOJ seems to be disregarding its avowed policy of rewarding lower discounts to cartel recidivists. Most evidence points to under deterrence of current penalties on cartels, particularly the international ones that comprise the bulk of this study's sample. It is likely that the level of U.S. fines contributes to under deterrence by building in expectations on the part of would-be cartelists for large cooperation discounts. Moreover, excessive discounting of cartel fines undermines the effectiveness of corporate leniency programs by reducing the monetary value of early cooperation. The DOJ and the USSC should re-examine the guidelines and toughen the fines for defendants from durable cartels and histories of recidivism.