Common carrier

Common carrier is a legal status imposed on certain enterprises that provide infrastructural services to the public — transportation, communication, energy, and increasingly digital platforms — requiring them to serve all customers without discrimination and at published rates. The common carrier obligation is not merely a consumer protection rule. It is a governance architecture that embeds nondiscrimination into the technical layer of infrastructure, preventing the owner of a bottleneck from leveraging control over the physical network to dominate the markets that depend on it.

The concept predates the regulatory state. At English common law, carriers — innkeepers, ferrymen, wagoners — were bound by the custom of the realm to serve all comers. The obligation arose not from statute but from the nature of the service: a provider who held themselves out as serving the public could not arbitrarily refuse a customer. This customary obligation was codified and extended in the nineteenth century as the state confronted the power of railroads, telegraphs, and telephone networks.

From Custom to Statute

The modern common carrier framework was constructed in response to the railroad monopoly. The Interstate Commerce Act of 1887 established the first federal regulatory commission and imposed common carrier obligations on interstate railroads: nondiscrimination in rates, reasonable service, and published tariffs. The logic was straightforward. A railroad that both owned the tracks and competed in the markets that used the tracks possessed an irreducible conflict of interest. The common carrier obligation was the state's attempt to manage this conflict without breaking up the railroads.

The pattern repeated in telecommunications. The Kingsbury Commitment of 1913 and the subsequent regulatory framework treated AT&T as a common carrier, requiring it to connect any caller to any other caller on nondiscriminatory terms. The Carterfone decision of 1968 extended this principle from service to equipment: the network operator could not control what devices attached to its infrastructure. The common carrier framework thus evolved from a rule about service provision into a rule about network architecture — the principle that infrastructure must be open to compatible attachments.

The Common Carrier and Digital Platforms

The contemporary debate over whether digital platforms — search engines, social media networks, app stores, cloud infrastructure — should be treated as common carriers replicates the historical pattern. Proponents argue that platforms have become essential infrastructure: a business that cannot appear in Google search results, an app that cannot be distributed through the Apple App Store, a service that cannot be hosted on Amazon Web Services — these are effectively excluded from the digital economy. The platform's control over the bottleneck is structurally analogous to the railroad's control over the tracks or the telephone company's control over the wires.

Opponents argue that platforms are not common carriers because they exercise editorial judgment — curating content, ranking search results, moderating speech — that is inseparable from their service. A search engine that could not rank results would not be a search engine; a social network that could not moderate content would not be a social network. The common carrier obligation, on this view, is incompatible with the functions that make platforms valuable.

This objection confuses the layer at which the common carrier obligation applies. The telephone company was a common carrier at the transport layer: it could not refuse to carry a call based on its content. But it was not a common carrier at the application layer: it did not dictate what callers said to each other. The platform as common carrier would be obligated to provide nondiscriminatory access to its infrastructure — data, APIs, distribution — without being obligated to abandon curation, ranking, or moderation at the application layer. The distinction is the same one that the End-to-end argument makes in network design: the infrastructure layer should be dumb and neutral; the application layer should be intelligent and competitive.

The Limits of Common Carrier

The common carrier framework has limits. It assumes that the service being provided is sufficiently standardized that nondiscrimination is meaningful. It does not work well when the service is inherently differentiated — when the quality of the service depends on the provider's judgment about what to include, exclude, or prioritize. And it depends on regulatory capacity: a common carrier obligation that is not enforced is merely a promise, and the history of common carrier regulation is a history of regulatory capture, gradual erosion, and eventual abandonment.

The deeper limitation is that common carrier obligations treat the symptoms of concentration without addressing its causes. A firm that is required to act as a common carrier remains a monopolist; it merely loses some of the instruments through which it can exploit its monopoly. The Essential facilities doctrine and Structural separation represent more aggressive interventions, but even these are temporary without ongoing governance. The common carrier is a compromise between the efficiency of integration and the justice of access — and like all compromises, it is unstable.

The common carrier obligation is the original network neutrality: not a rule imposed on infrastructure but a recognition that infrastructure cannot be both a monopoly and a marketplace. The owner of the road cannot charge the toll and compete with those who travel on it. This principle is not a regulatory innovation. It is a rediscovery of something English common law knew centuries ago: that the public nature of a service creates public obligations, and that those obligations cannot be waived by contract, consolidated by merger, or outrun by technological change. The failure to enforce common carrier obligations in digital markets is not a failure of law. It is a failure of memory — the forgetting of what infrastructure has always demanded.