Is the French mobile phone cartel really a cartel?

France Telecom (FT), SFR and Bouygues Telecom (BT) have been fined by France's Conseil de la Concurrence (CC) for organizing a mobile phone cartel with stable market shares (one-half, one-third and one-sixth, respectively) and for directly exchanging commercial information. While not contesting the legal decision, it is argued here that the economic reasoning is flawed. (1) As the CC made much of the firms' stable market shares, we have first followed this line of reasoning by considering that the market shares are quotas under uniform costs. Even if there is a general incentive to form a monopolistic cartel, BT was too small for it to be worth its while to join it; it is not necessary to exchange information directly to coordinate market shares and prices effectively; all partial cartels are unlikely. (2) We then departed from the CC by considering that the non-uniform market shares are explained by the costs in Cournot competition: this allows costs to be deduced from the observed market shares by assuming that they are kept the same when switching from Cournot competition to any form of cartel. We deduced that market shares cannot be other than stable and non-uniform; any monopoly is unlikely to come about, because FT has negative incentives to form a monopolistic cartel; no partial cartels of two operators are viable because at least one member would lose out. The paper also shows that Stackelberg competition is unlikely as well as Bertrand-Edgeworth competition. In conclusion, Cournot competition is the only arrangement that guarantees no losses to all operators.