We develop a leverage theory of tying in markets with network effects. When a monopolist in one market cannot perfectly extract surplus from consumers, tying can be a mechanism through which unexploited consumer surplus is used as a demand-side leverage to create a “quasi-installed base” advantage in another market characterized by network effects. Our mechanism does not require any precommitment to tying; rather, tying emerges as a best response that lowers the quality of tied-market rivals. While tying can lead to exclusion of tied-market rivals, it can also expand use of the tying product, leading to ambiguous welfare effects.

We consider competition between sellers selling multiple distinct products to a buyer having k slots. Under independent pricing, a pure strategy equilibrium often does not exist, and equilibrium in mixed strategy is never efficient. When bundling is allowed, each seller has an incentive to bundle his products, and an efficient “technology-renting” equilibrium always exists. Furthermore, in the case of digital goods or when sales below marginal cost are banned, all equilibria are efficient. Comparing the mixed-strategy equilibrium with the technology-renting equilibrium reveals that bundling often increases the buyer's surplus. Finally, we derive clear-cut policy implications.