Skip to main content

Audits and open gaps

Routine, not a concession

An audit that happens only when somebody complains is a political event: whoever demands it is making an accusation, whoever grants it is admitting something, and the result is read as a verdict on the complaint rather than on the election.

An audit that always happens is a control. Nobody had to accuse anyone, nothing is conceded by running it, and the result carries information instead of implication.

The difference is entirely in the schedule. Same procedure, opposite meaning.

Gaps are recorded, not argued away

Where a jurisdiction cannot produce what reproduction requires, that is written down as an open gap — and it stays written down until it is closed.

  • The gap names what is missing, who holds it, and what was said about it.
  • It is not softened into "data limitations" and it is not inflated into an accusation.
  • It closes when the material appears, and the closing is published too.

This is the same discipline used for every other kind of record on this site. A published gap is a real finding; a confident guess is not. See Naming the gaps and The public ledger of what is missing.

Why the ledger matters more than any single audit

One audit tells you about one election. A standing, public list of which jurisdictions can and cannot produce reproducible records tells you about the system — and it accumulates, which no individual audit does. Ten years of that ledger is the most useful election-integrity artefact anybody could build, and nobody has built it because it is boring and slow.

It also does something an accusation never does: it gives an administrator a specific, achievable thing to fix, and credit for fixing it.

Where it goes

Into the same catalog as everything else — Problems — and the same public repository at Open Data, cloneable in full.

Also in this area