No. XII · MMXXVI · Systems

Institutions After Scale

Most governance quietly assumes a crowd of people. When capacity stops tracking size, its core instruments come undone.

For two centuries the size of an institution was a usable proxy for its capacity and the governance it needed. This paper takes seriously a class of institution where that proxy has broken.

What it finds

Governance is a technology for arranging people. Headcount was its substrate, not a feature.
When capacity stops tracking size, some instruments do not weaken, they vanish.
Accountability, legitimacy, oversight, succession all quietly assumed people would be there.

Read the full paper →

Summary

For two centuries the size of an institution was a usable proxy for its capacity, its risk, and the governance it needed. Almost the entire apparatus of institutional trust, the board, the separation of decision management from decision control, the segregation of duties, the audit, the bench, the regulatory threshold, was built on top of that proxy and silently requires people to staff it. This paper takes seriously a class of institution in which capacity no longer tracks size: small in headcount, dense in capability, machine assisted, acting at a scale its size does not predict. Working from real governance and institutional theory, it shows that several classical instruments do not merely weaken at low headcount but become structurally undefined, present as names and absent as functions. Accountability needs a forum that is not the actor. Legitimacy of the durable kind is conferred, not asserted. Every thread points off the institution and onto the layer beneath it.

What this paper holds

Governance is a technology for people

Almost the entire governance canon, the board, separated decision control, segregated duties, the audit, succession, is a set of techniques for arranging people relative to one another. Headcount was never a feature it tolerated. It was the substrate it ran on.

Some instruments do not weaken, they vanish

Below a certain number of parties, several classical instruments do not run badly. They do not run at all, while remaining nameable in a policy document. An instrument present as a name and absent as a function is more dangerous than its admitted absence.

What the canon needed was not people

The separations in the canon were never about the count of bodies. They were about genuine independence of judgement, interest, and standpoint. Headcount was merely the cheapest way history ever found to manufacture independence.

The inverse of many hands

The classical accountability pathology is many hands and diffused blame. The institution after scale has the exact inverse: one hand and no forum. An actor reporting to itself is not being held accountable. It is keeping a diary.

Throughput legitimacy does not trade

The form is strongest at the legitimacy that delivery buys and structurally weakest at the legitimacy of the black box, which is the one dimension that cannot be bought back by good results. It fails at the first test where results stop speaking for themselves.

Trust became external, not optional

Removing the people did not remove the trust functions. It removed the internal raw material out of which trust used to be constructed, which means trust did not become unnecessary. It became unconstructible from inside.

The value is in not controlling it

A standard the institution can swap when it stops being flattering is not a standard. The worth of the trust layer is exactly proportional to the institution not being able to author or replace it.