Consolidation Puts Most Traffic Through Few Hands

A design that assumed many roughly comparable networks now operates with a small number of very large ones carrying much of the load.

The arrangement was conceived with many independent networks of broadly similar significance in mind, where no single participant's decisions could reshape the whole. What has developed instead is a landscape where a modest number of very large operators, content platforms and delivery providers account for a great deal of what happens. This has not required anybody to change the rules. It is the ordinary result of scale economics in a business where serving more traffic gets cheaper per unit as you serve more of it.

The effects cut both ways, and it is worth resisting a simple story. Large operators have invested heavily in capacity and resilience, and users generally experience faster and more reliable service as a result. Concentration also concentrates competence: a handful of well resourced organisations can respond to problems faster than a thousand small ones. Against that, decisions made inside a few organisations now affect very large numbers of people who had no part in making them, and the failure of one such organisation is felt far more widely than any single failure was ever meant to be.

What consolidation does not do is give anybody ownership of the whole. Even the largest participants remain networks that connect to others by agreement, and none can dictate terms to all the rest. But the practical distance between formal independence and real bargaining power has widened, and the voluntary arrangement works rather differently when the parties are very unequal in size. Observers disagree about how much this matters and what if anything should be done, and that disagreement is one of the genuinely open questions about the system's future.