Bundling has been regarded as a highly ambiguous method for price discrimination or vertical control.  Barry Nalebuff has recently proposed an alternative model of bundling as a highly suspect exclusionary tactic.  A virtue of the model is that its exclusionary implications do not appeal to strategic considerations, e.g., threatening to charge a predatory price for the bundle now and recoup losses later. It involves two goods, A and B, each initially supplied by monopolies.  If the A-monopolist sells an A-B bundle at the sum of those monopoly prices, the B-monopolist loses half of its profits and consumer welfare falls.  Because those prices are not an equilibrium, we focus on three possibilities: sequential pricing, simultaneous pricing?both of which involve the B-monopolist remaining?and monopoly, i.e., where the bundler is the only seller of A or B. 

In all cases, including B?s departure followed by a monopoly price, total welfare and consumer welfare are greater after bundling than before.  We cannot guarantee that bundling always increases welfare, but the results have intuitive support.  By bundling, the A-monopolist provides previously non-existent competition with the B-monopolist in the latter?s market.  The B-monopolist?s exclusion comes from the lost profits due to the A-monopolist?s entry.  Bundling is an entry barrier because competition reduces profits.  Were bundling harmful, the per se proscription against market allocation should be lifted.  It would presumably be beneficial for an erstwhile bundler and a single product firm to cut a deal in which the former stops bundling in order to preserve a monopoly for the latter.  Our findings suggest, however, that the present antitrust rules promote welfare and, thus, that generic opposition to bundling remains unwarranted.