Chillin'Competition

Given today’s announcement, I suspect Alfonso has better to do than posting on this blog. Run, Alfonso (on the banks), run!

With this, it is thus my duty, and honour, to introduce the 4th edition of the Economist Corner. For this edition, Benoît Durand (RBB Economics) has sent us a good piece on a money-related issue, i.e. fines for cartel infringements. Enjoy!

In the last decade the European Commission has imposed higher fines on cartels, in particular under the helm of Neelie Kroes.  The stated purpose for this increase was that fine levels were not sufficiently high to deter the formation of cartels.

In general, the deterrence property of sanctions is a key aspect of law enforcement.  Becker (1968), who was the first to apply economic principles to crime and punishment, explains that the level of sanctions should be set so as to deter crime.  A high level of sanction in turn contributes to minimise the costs of enforcing the law.[1]

Firms consider the expected benefits and costs of participating in a cartel.  Under this logic, if the expected sanctions are higher than the collusive gains, then firms will not take the chance.  Because there is always a significant probability that cartels slip through the net, the penalties should be several times larger than the gains such that no firm would dare try fixing prices.  By way of example, consider that a cartel member expects to pocket 50 million euros extra every year for about 6 years, whilst the probability of being caught is 1 out of 5.  In this setting, it would take a massive fine of slightly more than 1.5 billion euros to convince a firm not to collude.[2]

As cartels continue to exist, it must be the case that the current level of sanctions is ineffective.  This is the conclusion that Combe and Monnier (2011) draw after reviewing the fines for 64 EC cartel decisions between 1979 and 2009.  They show that in virtually all cases fines were set below the optimal deterrence level; i.e. in spite of the sanctions, the cartels were profitable.[3]

Is it therefore necessary to raise corporate fines above the current levels to deter the formation of cartels?  It is hard to say, but to achieve full deterrence, competition authorities need not increase fines at stratospheric levels as suggested by the logic described above.  First, they could adjust sanctions to give cartel members the incentive to undercut each other, which would trigger the collapse of cartels.  Second, in complement to corporate fines, competition authorities could consider applying measures targeted at company officers who have brokered the cartel agreement.

On the first possibility, it is well known that cartel members are permanently tempted to undercut their rivals.  But they adhere to the collusive price level because they fear that once other members realise that they have cheated on the agreement, all hell will break loose, everyone will start competing aggressively, driving down future profits.  In sum, cartels are stable because its members anticipate that undercutting the collusive price level is not a profitable strategy in the long run.

Importantly, competition authorities only need to set corporate fines at a level that tips the balance and give cartel members the incentive to undercut the collusive price level.  To achieve this goal, fines do not have to be equal to a multiple of all collusive gains as in the logic developed by Becker.[4]  Furthermore, competition authorities can make deviation the more attractive option by exempting the firm that first reports the cartel from paying any fine (as in many leniency programs).  The firm understands that sticking to the collusive price level involves paying some fines if convicted, whilst deviating and reporting the cartel brings an additional benefit: a fine exemption.  Competition authorities could further strengthen the incentive to deviate by giving a reward to the first firm that submits evidence.  Although this might be politically heroic, the authority could employ part of the fines paid by the other cartel members to financially reward the firm that reports the infringement.

On the second possibility, an effective sanction policy could seek to exploit agency problems within firms.  A firm is a complex organisation, and the interests of its employees do not always match those of its shareholders.  Generally, shareholders are not involved in the decision to participate in a cartel agreement.  This decision is taken by company officers, who benefit through higher salaries, bonuses and perks, but in many jurisdictions they do not risk any sanction.

One way to discourage the formation of cartels would thus be to develop–in parallel to corporate fines–a toolbox of measures (including fines, disqualification orders, etc.) aimed directly at individuals who decide to participate in a cartel agreement.  If company officers become aware that not only fixing prices is illegal, but also that they could be held responsible, this might further strengthen deterrence.  Recently, Ginsburg and Wright (2010) proposed to expand sanctions for individuals, in particular debarment, which is not often used.[5]  Aubert, Kovacic and Rey (2006) go further and propose to reward whistleblowers in order to encourage insiders to report the existence of a cartel.[6]^,[7]

Food for thought… and possibly for future reforms…

Benoît Durand (c), 30 May 2012