Multimarket Contact and Platform Competition: Reassessing the Mutual Forbearance Hypothesis

2019 
Antitrust authorities are particularly concerned with the dominant market position of tech giants such as Google, Facebook, and Amazon. These digital conglomerates are characterized by platform-based business models. However, despite their dominance, they are competing with each other to attract the same groups of users (developers, advertisers, end users, third party sellers, etc). They therefore have not only overlapping users (or sides) but also multimarket contact (MMC). In traditional one-sided markets, theory and empirical evidence show that MMC tends to relax competition. However, it is unclear whether this result holds under platform competition. This paper examines how MMC a ects pricing behaviour and profits of two-sided platforms. We develop a model of platform competition with two distinct markets. We assume that platforms only charge one group of users and provide free access to the other group. We argue that multimarket platforms also generate cross-market externalities that favour their users, in addition to well-known cross-group externalities. We found that when cross-market externalities benefit the side that has free access, price competition is fiercer and total welfare increases under MMC. However, when they benefit the side that pays to access the platform, the same result only holds if the cross-group externality and/or cross-market externality are suciently high. Finally, we show that a single-market platform competing with a multimarket platform may be deterred from entering the second market if cross-market or cross-group externalities are high. Our findings contrast with the mutual forbearance hypothesis which claims that MMC relaxes competition in traditional (one-sided) industries. From a competition policy perspective, our paper provides an insight into how antitrust authorities should review conglomerate mergers and assess the e ects of diversification strategies of digital platforms.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []