Correcting a CARF filing: amendments, DocRefId, and MessageTypeIndic

New, corrected, or deleted

The message header declares what kind of submission this is through a message-type indicator. A file is either fresh data or a set of corrections; you don't mix them casually. Each record inside then says whether it's a new record, a correction of an earlier one, or a deletion of one sent before.

The identifiers that make it work

Every document you send carries a unique DocRefId. To correct or delete a record later, you reference the exact DocRefId of the original. That back-reference is the whole game: it's how the authority knows which of the thousands of records you already filed this one replaces. Reuse or muddle those identifiers, and corrections attach to the wrong records or fail outright, which is one of the rejection patterns in the CARF XML errors that get filings rejected.

Why sloppy first filings hurt twice

This is the practical reason identifier hygiene matters from day one. If your original submission had duplicate or unstable DocRefIds, you can't cleanly amend it, because there's no reliable handle to point back at. A tidy first filing is what makes every future correction a small operation instead of a reconciliation project.

Keeping the audit trail

Corrections also have to be auditable: you should be able to show what you filed, what you changed, and when. Our CARF and DAC8 reporting platform assigns stable identifiers, builds correction and deletion messages that reference the right originals, and keeps the full history, so an amendment is a click rather than a hand-built XML edit. For how this fits the wider filing flow, see the CARF XML schema, section by section, and the OECD exchange-of-information hub for the correction rules.