Research

If a retailer invests in alternative sourcing for a product otherwise provided by a dominant supplier, it may gain leverage to pressure the supplier into lowering the wholesale price (bargaining effect). The reduced wholesale price, in turn, strengthens the retailer's competitive position, enabling it to capture additional market shares (business-stealing effect). To counter the latter effect, and thus reduce the retailer's incentive to invest in alternative sourcing, the supplier might commit to uniform wholesale pricing. However, we demonstrate that this strategy is unprofitable when retailers are sufficiently differentiated. Instead, if retailers differ in size, supplier profitability is maximized through size-based price discrimination. This might harm consumers. Interestingly, when substitutability is at an intermediate level, supplier and consumer preferences align.

The authors present two classroom experiments on technology licensing. The first classroom experiment introduces the concept of royalty stacking. Students learn that non-cooperative pricing of royalties for complementary intellectual property rights leads to a double-marginalization effect. Cooperation solves the problem and is welfare-improving. The second classroom experiment introduces students to cross-licensing. It shows that reciprocal royalty payments dampen competition. The classroom experiments stimulate discussions of technology licensing, intellectual property rights, different royalty structures, patent pools and technology standards. The authors present the experimental procedures, and suggests routes for the discussion.

Digital platforms are technology intensive and often require access to intellectual property rights. As those rights are often owned by competing platforms, we observe cross-licensing contracts between them. Cross-licensing facilitates the sharing of technologies, leading to less duplication of resources and equal access to technology---facilitating competition on equal terms. Yet the economic literature has traditionally been worried for the anti-competitive effects of cross-licensing in the presence of monetary transfers in traditional (one-sided) markets, since a positive royalty essentially raises the rival's costs, leading to higher prices. Through a two-sided platform model, I show that positive royalties in cross-licensing contracts in two-sided markets have two opposing effects: the competition effect and the two-sided effect. The two-sided effect alleviates the anti-competitive effects, and may even flip the outcome. Under strong network effects, the two-sided effect of cross-licensing contracts may in fact dominate, leading to lower prices. The results suggest that policymakers should be less concerned about competition harm associated with cross-licensing contracts in markets with strong network effects.

We address the impact of targeting through the use of first-party data and consumer multi-homing on platform competition and market equilibria in two-sided markets. We analyze platforms that are financed by both advertising and subscription fees, and let them adopt a targeting technology with increasing performance in audience size: a larger audience generates more consumer data, which improves the platforms' targeting ability and allows them to extract more advertising revenue. Targeting therefore increases the importance of attracting consumers. Previous literature has shown that targeting could result in fierce price competition if consumers subscribe to only one platform (i.e., single-home). Surprisingly, we find that pure single-homing possibly does not constitute a subgame-perfect Nash equilibrium. Instead, platforms might rationally set prices that induce consumers to subscribe to more than one platform (i.e., multi-home). With multi-homing, a platform's audience size is not restricted by the number of subscribers on rival platforms. Hence, targeting softens the competition over consumers. We show that this result could imply that equilibrium profit is higher with than without targeting, in sharp contrast to the predictions given in previous literature.