Main idea

The charge of paperwork bears on the existence of the document: one more page, one more approval, one more box to fill in. The real reckoning happens elsewhere, by comparing what the document costs to write and maintain with what it removes — the explanation repeated fifty times, the arbitration reopened with every big prospect, the dependence on one person's memory, the constant navigation between implicit rules, unspoken exceptions and contradictory decisions.

Nothing forces that reckoning to come out positive. A document that is too long, or too numerous, turned into meeting minutes, approved by too many people, or written after the fact to justify a decision already imposed, does indeed add load without removing anything. It is those uses that produce paperwork, not the act of writing.

The distinction is therefore between documentation and bad documentation, and it is settled by a subtraction, not by a principle. What makes the calculation hard is that the cost is visible and immediate while the gain is invisible and diffuse: the discussions that did not take place leave no trace.

Why it matters

This gives the only criterion that allows a documentary system to be arbitrated without pitting two reflexes against each other — "we must keep records" against "we are drowning in process" — neither of which is verifiable.

It also says what to observe in order to evaluate an existing system: not the number of documents produced, but whether or not the re-discussions it was meant to avoid have disappeared.

Nuances and limits

The subtraction is asymmetrical in time: the writing cost is paid once, now, while the gain is spread over months, which spontaneously tips the judgement the wrong way.

And some documents do not aim to remove discussions: a record required by an auditor or a regulator is judged on its compliance, not on its yield.

Open questions

• How do you establish that a re-discussion did not take place, when its absence produces no observable signal?