A tying arrangement is a seller’s requirement that a customer may purchase its “tying” product only by taking its “tied” product. In a variable proportion tie the purchaser can vary her purchases of the tied product. For example, a customer might purchase a single printer, but either a contract or technological design requires her to purchase varying numbers of printer cartridges from the same manufacturer. Such arrangements are widely considered to be price discrimination devices, but their economic effects have been controversial.
Tying has been attacked on the theory that price discrimination of this sort reduces consumer welfare. We show that this argument is based on a misunderstanding of the kind of price discrimination that is involved in variable proportion ties. The great majority of them almost certainly produce both welfare gains and net consumer benefits.
We also consider and reject the argument that tying produces greater welfare losses when viewed from an ex ante rather than an ex post perspective. That argument rests on a flawed premise about the sources of the increased returns to innovations whose distribution requires tying. Further, it ignores the important role of fixed costs in producing innovation incentives. We also show that tying in concentrated markets produces significant benefits from the elimination of double marginalization. Then we extend our analysis to bundled discounts, focusing on the possibility of increased harm that can occur if the monopolist increases the standalone price of one good when inaugurating the bundled discount.
Antitrust’s per se rule is reserved for practices that are so likely to cause antitrust harm and have so little to defend them that detailed case-by-case assessment is thought to be unnecessary. They can be condemned categorically simply upon a showing that a few basic conditions are satisfied and that they belong in a particular class of restraints. No kind of unilaterally imposed tying arrangement, even by a monopolist, falls into that category.