The art of the two-tier exhibit index

Think about what the adjudicator is actually doing. They have a petition asserting that the beneficiary satisfies a set of criteria, and a binder of several hundred pages. Their task is to verify each assertion. The index is the only tool they have for navigating between the two.

A flat list of thirty-one exhibits in the order they happened to be gathered does not support that task. A two-tier structure does, because it mirrors the shape of the argument.

Tier one is the statutory or argumentative grouping: Exhibit 4 — Original contributions of major significance, 8 CFR 204.5(h)(3)(v). Tier two is the individual document: Exhibit 4.1 — Declaration of Dr. Elena Moreau, Pages 84 to 97. The adjudicator reads a criterion in the cover letter, goes to tier one, and finds every document supporting it grouped together with page ranges.

The alternative — one flat sequence with the criterion mapping buried in the cover letter — forces the adjudicator to hold the mapping in their head while paging through a binder. That is where 'evidence not established' findings come from in records that contain the evidence.

• Tier one: criterion or argument grouping with its regulatory citation.
• Tier two: individual document with a full page range.
• A one-line statement of what the exhibit establishes.
• Bates range in the same row, matching the stamps on the page.

Compare two index entries. 'Exhibit 4.2 — Peer review records.' Versus 'Exhibit 4.2 — Peer review records for Nature Materials and IEEE TPAMI, 2022 to 2025, Pages 84 to 112, substantiating judging of the work of others under 8 CFR 204.5(h)(3)(iv).' The second entry does the adjudicator's work for them, and it takes no more space in the row.

This is also a self-check for the drafting team. If you cannot write a one-line statement of what an exhibit establishes, that exhibit may not belong in the packet.

The instinct is to hide a weak criterion. The better practice is to make coverage visible during drafting so the firm can see it. An index built from the criteria outward — rather than from the documents inward — shows immediately when a criterion has one thin exhibit under it, while there is still time to gather more.

This is the difference between an index generated at the end of assembly as a formality and an index used as a working instrument throughout. The second version prevents RFEs; the first only describes what happened.