What a good vote rationale actually looks like

This is a statement of the drafting standard the Institute teaches, not a study. We have not surveyed or coded a sample of published rationales, and we make no claim about how common any of these traits are in practice.

The traits below come from the fields the CIP-136 vote-rationale metadata standard asks a document to carry, and from what a reader needs in order to check a decision without asking the author a follow-up question.

First, context: what the action actually asks for, stated before any judgement. Second, evidence: what the author checked, where it came from, and what it said as distinct from what they inferred. Third, tradeoffs: what is given up under the author's own conclusion, stated as a cost they accept. Fourth, decision: the position, together with concrete conditions that would change it.

Length is not one of the traits. A rationale under 300 words can carry all four, and a very long one can carry none of them.

Structure is not correctness. A rationale carrying all four traits can reach a conclusion you find wrong, and a scattered one can be right. This describes how reasoning is communicated, not the quality of the judgement behind it, and never which way anyone should vote.