Talk Title: When Contestability Leads to Concentration in Digital Markets
Abstract: Digital markets often appear highly competitive in their early phases: firms race to acquire compute, data, users, or routing flow, and consumers can switch freely across providers. Yet many of these markets rapidly concentrate. What drives this pattern? We develop a model to isolate a minimal structural mechanism for concentration that does not rely on traditional network effects, switching costs, or behavioral lock-in. Instead, centralization emerges from contestability over rival inputs combined with quality-based competition. The framework applies to digital markets such as search advertising, spectrum-based services, high-frequency trading, and emerging AI ecosystems in which access to compute, data, or routing flow determines performance. We show that markets may stay competitive (but are not guaranteed) only when rival input constraints are relaxed or multihoming is frictionless.
(https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6299698)