The Standard

The standard makes strangers' work fit together. A bolt enters a thread, a message reaches another machine, a shipment passes between vehicles, and an account can be compared with an account made elsewhere. Each successful connection carries a prior agreement about dimensions, meanings, or procedures. The earlier Protocol brought those quiet agreements into view. Their economic importance begins with the effort they spare: fewer translations, fewer incompatible purchases, fewer occasions on which a buyer must discover from the beginning what a seller means. Their political importance begins at the same place. Whoever chooses the common specification helps decide which investments retain their usefulness, which producers can enter, and which differences everyone else must pay to overcome. Coordination produces a benefit and distributes a power.

Compatibility makes the value of a choice depend on choices already made by others. Imagine two communication systems. A gives a user four units of value before connections are counted; B gives six. Add one tenth of a unit for each existing user on the same system. With eighty users on A and twenty on B, a newcomer values A at twelve and B at eight. The newcomer chooses A despite B's higher stand-alone quality. If the distribution were reversed, B would offer fourteen against A's six. The example isolates a coordination mechanism: the installed population becomes part of the product. It does not establish that the historically dominant system is inferior, or that every market tips to one supplier. Those conclusions require evidence about alternatives, compatibility, costs, and what users actually value.¹

The difficulty of moving alone can make a jointly beneficial change privately unattractive. Suppose ten firms use the same old reporting format. Each would gain eight units from a complete move to a better common format and pay three to convert. A coordinated move produces a gain of five per firm, fifty altogether. An isolated pioneer, unable to exchange files with the other nine, instead loses twelve units of compatibility while paying the three-unit conversion cost. Waiting is sensible for the individual even when everyone would welcome the completed change. A published transition date, dual-format software, or a converter can alter the choice. The obstacle is the path from one compatible arrangement to another; describing the destination's superiority leaves that path unbuilt.²

A common standard can be governed through several different arrangements, each with its own discipline. A public authority may prescribe it, a company may control it, or a community may develop it through an open process. The IETF's 2004 mission statement offers one explicit account of the last arrangement: participation, public records, technical judgment, and experience with working implementations. It defines a standard as a specification for interoperating, while expressly separating publication from a power to compel adoption. That distinction matters. A protocol may become commercially indispensable because others use it, even where its author cannot command anyone to use it. The practical strength of the convention and the legal authority of its maker are separate questions.³

The distributional contest often concerns the conditions of implementation as much as the specification itself. If conformity requires a patented component, access to a private interface, or a certification process controlled by an incumbent, the common language can carry a toll. A technically sound requirement may still impose costs that a large firm bears more easily than a small entrant. Conversely, an open document alone does not supply the engineering time, equipment, or security needed to implement it reliably. The useful inquiry follows the whole route into the market: who can read the specification, who may implement it, what permissions and fees are necessary, and who can challenge a refusal. Each answer can change the range of producers able to compete within the shared convention.⁴

Maintaining a standard requires decisions about change as well as decisions about entry. Compatibility protects investments already made; revision can repair defects, improve performance, or meet a new need. Freezing every interface would make yesterday's limitations permanent. Replacing them without warning can strand users who built in reliance on the old arrangement. A credible process therefore needs reasons for revision, opportunities to test alternatives, a usable transition, and a record of objections and their resolution. Openness helps make that process inspectable, but effective participation still takes time and expertise. Funding, representation, and the ability to test a proposed design affect whose difficulties become visible before the common rule is settled.⁵

The standard succeeds when it lets independent people do more together while leaving its own governance open to examination. Its value lies in the connections it makes possible; its danger lies in turning those connections into an unanswerable claim on everyone who relies on them. No single institutional form resolves the tension in every setting. The right design depends on the costs of incompatibility, the pace of change, the importance of safety, and the power attached to implementation. These are questions about governing a common dependency. The next plate brings that dependency under one roof, where the owner of a marketplace can alter the conditions of exchange for everyone inside it. The standard coordinates the participants; the forum decides the terms on which they may meet.