Many of the classic tying cases involved tied products that were common staples such as button fasteners, canned ink, dry ice, or salt. These products were sold in competitive markets, presumably at prices very close to cost. For most of them the most likely explanations for the tie were quality control or price discrimination, both with competitively benign results in the great majority of situations. When the tied good is sold in a noncompetitive market, however, an additional consumer welfare enhancing result is likely to obtain – namely, the elimination of double marginalization, which occurs when separate sellers of complementary products each has market power. The double marginalization result occurs only when both products are sold at prices above the competitive level. It can apply in both the vertical context, such as when a monopoly manufacturer must distribute through a retailer market that is subject to monopoly, oligopoly or collusion, but it also occurs in situations involving complements, such as printers and ink cartridges, hospitals and physicians, or many medical devices that include both durable and reusable components. Indeed, the problem is generally more serious in the complementary situations, because often firms offering complementary products are not in a good position to negotiate with one another for a joint-maximizing output increase, while the participants in a vertical chain of distribution bargain with each other all the time. In cases where both tying and tied products are subject to some market power, the profit maximizing price of a seller who offers both products in a bundle is typically lower than the individual profit-maximizing prices of two sellers, each of which sells only one of the products. Outside of the franchise context, which often involves commodity tying, most ties today involve manufactured tied products subject to at least some product differentiation, and typically nontrivial fixed costs. In all such cases the double marginalization argument presumptively applies and suggest another reason why ties should not be condemned except after a full rule of reason analysis. The typical result of eliminating double marginalization is that output of both the tying and the tied good increase and consumers pay lower prices. The result is a welfare improvement under both the total welfare and consumer welfare tests for competitive harm
