Governments and institutions are learning how to respond to platform power. The response is slower than the power it is addressing, and the delay can deepen the power it is meant to constrain.
The Institutional Response Lag
Platform power is more than market size. It is the ability to become a gateway and then govern the conditions under which businesses and users meet. Network effects can reinforce scale; data-driven advantages can raise barriers to entry; control of operating systems, app stores, search distribution, marketplaces, and interfaces can create dependence. These effects are not automatic or identical in every market. But the Digital Markets Act identifies their combination—network effects, multi-sidedness, lock-in, limited multi-homing, vertical integration, and data advantages—as a basis for an entrenched and durable gatekeeper position.
The asymmetry is temporal. Platforms can change products, contracts, and technical rules continuously. Institutions must identify harm, collect evidence, establish a legal theory, enact or apply authority, survive challenge, and build enforcement capacity. The sequence protects due process, but it also consumes time. During that interval, installed bases, contractual defaults, ecosystem dependencies, and data advantages may continue to accumulate.
A Chronology, Not a Blank Decade
The 2010s were not a period of institutional silence. The EU adopted the General Data Protection Regulation in 2016, and it applied from 25 May 2018. The European Commission issued major competition decisions concerning Google Shopping in 2017 and Android in 2018. It is therefore too broad to say that the decade passed with minimal institutional constraint.
The narrower claim is stronger: institutional activity and effective structural constraint are not the same thing. Privacy rules, ex post competition cases, and ex ante gatekeeper obligations address different mechanisms. An investigation is not a judgment; a judgment is not a remedy; a rule on paper is not implementation; implementation is not demonstrated competitive effect.
The later chronology makes the lag visible. The US Department of Justice filed its Google search complaint on 20 October 2020, and the Federal Trade Commission filed its Facebook complaint on 9 December 2020. At filing, both contained allegations requiring adjudication, not established findings. The European Commission proposed the DMA in December 2020; the European Parliament and Council adopted it on 14 September 2022; it became applicable on 2 May 2023; and the Commission designated the first six gatekeepers in September 2023, giving them six months to comply. Almost every date marks a real advance, but none by itself establishes market effect.
What Effective Response Requires
First, analysis must fit market structure. Price and current market share remain useful, but they are incomplete where one user group pays no monetary price, demand on one side affects demand on another, and quality, switching costs, interoperability, or access govern competition. The European Commission’s 2024 market-definition notice expressly treats multi-sided and zero-price services as markets requiring attention to indirect network effects and non-price parameters. The US agencies’ 2023 Merger Guidelines likewise direct analysis toward competition between platforms, competition on a platform, and competition to displace a platform.
Second, the remedy must reach the mechanism. If default placement forecloses distribution, a remedy confined to headline price misses the bottleneck. If lock-in or limited multi-homing protects a position, switching and interoperability matter. If data generated through the platform creates dependence, access and portability may matter. If the operator competes with firms that depend on its gateway, self-preferencing and discriminatory access become central. This is why the DMA specifies conduct obligations rather than requiring every platform problem to be reconstructed from first principles after entrenchment.
How to Measure Institutional Learning
Institutional response should be measured on three clocks: recognition, when authorities identify the problem; authority, when law and doctrine permit intervention; and effect, when remedies alter conduct, entry conditions, user choice, or bargaining power. Reports and lawsuits may advance the first clock. Statutes and judgments may advance the second. Only implementation and observed outcomes advance the third. Confusing the clocks makes the response look earlier—or more effective—than it is.
When power compounds faster than governing institutions can learn and act, institutional delay becomes part of the power being regulated. By the time the rule is ready, it must address not only the original conduct but the position accumulated while the rule was being made.
Discussion