Arc II #10: Regulating Coordination Infrastructure

Tenth Law: Coordination infrastructure that achieves systemic importance must be regulated as infrastructure — not as a competitive product market.

When Infrastructure Is Infrastructure

The tenth structural law of the coordination economy states that when coordination infrastructure achieves the level of scale and dependency that makes it systemic — when a significant portion of the economy depends on it for the coordination that economic activity requires — it must be governed as infrastructure rather than as a product in a competitive market. The historical precedent is real, though it is more contested and more gradual than a clean story of automatic transition suggests. Congress passed the Mann-Elkins Act in June 1910, amending the Interstate Commerce Act of 1887 to expand the Interstate Commerce Commission's responsibilities to include the regulation of telephone, telegraph, and cable companies, declaring such companies common carriers subject to non-discrimination and rate oversight, following on decades of prior extension of common-carrier logic to the railroads themselves. But this transition was not a mechanical response to scale alone. Throughout the later 1800s, smaller farming communities were often served by only one rail line, and the companies serving them were able to extract rates far higher than in more competitive markets — and it was the resulting political pressure, not the scale itself, that produced the regulatory response. Infrastructure status was fought for and legislated, not triggered automatically by systemic importance.

It is also worth being honest about how that story continued. By 1970 it had become clear that federal regulation of interstate commerce had created problems of its own, and Congress began deregulating the industries the ICC had governed; in 1996 Congress eliminated the Interstate Commerce Commission altogether, after years of criticism that the agency had favored the railroad industry through high rates that discouraged lower-priced competition. The historical parallel this law depends on is genuine, but it is not a simple success story to be imported wholesale — it is a case study in both the necessity and the eventual cost of treating coordination infrastructure as infrastructure. The lesson is not that regulation fails; it is that the toolkit itself needs maintenance, sunset review, and correction, not a one-time legislative act followed by permanent trust.

The digital platforms that have achieved comparable systemic importance in the coordination economy — the dominant search engine, the dominant social network, the dominant e-commerce marketplace, the dominant cloud infrastructure provider — have, in the United States, mostly not been governed as infrastructure at equivalent scale; the primary tools remain competition law and merger review, frameworks built for a different structural problem, and the political economy of platform regulation has generally favoured the incumbent positions of the platforms being regulated. The exception, and the place where this law is now being tested in real time rather than in theory, is the European Union. The Digital Markets Act establishes a set of clearly defined objective criteria to identify "gatekeepers" — large digital platforms providing core platform services such as online search engines, app stores, and messenger services — and imposes on them obligations that sit squarely within the infrastructure-governance category this essay describes: allowing third-party interoperability, providing access to data and enabling data portability, and prohibiting unfair business practices such as self-preferencing. This is no longer a proposal on paper. In April 2025 the Commission fined Apple €500 million and Meta €200 million, the first non-compliance decisions under the regime, and enforcement has continued since. Whether this constitutes full infrastructure governance, or a partial and still-contested first draft of it, remains an open question — enforcement is uneven across gatekeepers and actively resisted by the platforms it targets, and no comparable regime yet exists in the platforms' home jurisdiction.

The Infrastructure Governance Toolkit

Governing coordination infrastructure as infrastructure requires a toolkit that goes beyond the competition law prohibitions and merger control that have been the primary tools of digital platform regulation to date. Common carrier obligations that require systemic coordination infrastructure to provide access on non-discriminatory terms are no longer hypothetical: the DMA already requires that gatekeepers provide access on fair, reasonable, and non-discriminatory terms to search data. Interoperability mandates that prevent lock-in at scale are likewise in force — the DMA mandates that gatekeepers ensure interoperability across their services, including messaging services, operating systems, and virtual assistants. Data governance requirements that prevent the accumulation of data advantages that competition cannot address — data portability obligations and restrictions on cross-service data combination — fall in the same category. And platform governance accountability requirements that bring the rule-making functions of systemic platforms within a framework of due process and appeal that their quasi-regulatory role requires remain the least developed part of the toolkit, even in the most advanced regulatory jurisdictions.

None of this settles the underlying normative question, which is distinct from the descriptive one. That systemic digital platforms function economically the way historical infrastructure once did is a structural observation with genuine support. That they should therefore be regulated the same way is a normative argument, not an empirical finding — and the Interstate Commerce Commission's own arc, from necessary corrective to agency Congress judged worth abolishing, is a reminder that the infrastructure toolkit carries its own costs, to be weighed rather than imported by analogy alone.

The tenth law: systemic coordination infrastructure is public infrastructure in economic function regardless of its ownership structure. Governing it as a competitive market is governing it incorrectly — and the harms that result are the predictable consequence of the mismatch between governance framework and economic reality. Whether the infrastructure toolkit is applied wisely, or eventually accumulates the same pathologies it was built to correct, is the harder question the historical analogy alone cannot answer.