Methodology

The four layers, the attribution scale, and the rules this archive follows.

Standing rule: no scores. Color marks whether something is in force, reversed, contested, concluded, or permanent — not good or bad.

Unscored · 97 entries · last change 2026-08-15

New here? Read this first — 30 seconds

Color is a legal fact, not an opinion. It shows — whether something is still in force, was reversed, or is contested in court — never whether it was good or bad.
Every number links to its source and to how strongly it can be tied to the action () — click any figure to see the full evidence.
Nothing here is scored, graded, or ranked. When two credible sources disagree, this archive shows both numbers instead of picking one.
Wins are in here too. See or filter the Ledger by .

Methodology version 1.0 · unchanged since 2026-08-10

These rules have not changed since 2026-08-10. If they change, the previous version stays published here alongside the new one and the change is written into the corrections log — the same rule the entries themselves follow. A methodology that can be revised silently is not a methodology.

01 — The four layers

Nearly every argument about a president's record collapses four different kinds of claim into one. This archive keeps them physically separate, because each has a different evidentiary standard and a different failure mode.

The fourth layer is deliberately empty. A tariff that raises consumer prices and protects domestic mill jobs is not "good" or "bad" — it is a transfer, and whether you favor it depends on who you think should bear the cost. The archive states the transfer and stops.

04 — When sources disagree

Show the spread, never the midpoint

Two credible studies producing $200B and $900B do not average to $550B. Both numbers appear, each with its method and its institutional orientation, and the width of the gap is presented as information.

Separate the model from the measurement

CBO’s deficit score is a projection; Treasury’s receipts are a measurement. Conflating them is the most common way a tracker loses credibility with sophisticated readers.

Name the counterfactual

"Cost $X" is meaningless without "compared to what." Prior law, current law, a policy baseline, and a no-policy world give four different numbers for the same law — all honestly.

Log every correction

A public, dated corrections ledger with the original claim visible alongside the fix. Silently editing a claim is how a reference site becomes untrustworthy overnight. Track your own error rate and publish it.

Record what remains open

When the evidence does not yet support a conclusion, the entry says so. An archive willing to write "unresolved" is more credible than one that always has an answer.

02 — scale

Every effect claim carries one of four attribution levels. Most economic outcomes are contributory or contextual, and labeling them honestly is the single strongest credibility signal this archive can send.

The action is the proximate legal or administrative cause. A signed order changes a tariff rate; the rate change is direct. Nearly all Layer 1 and 2 claims sit here.

One identifiable input among several, where a published study attempts to isolate the effect. Most credible economic findings live here — including favorable ones.

The outcome occurred during the tenure, and causation is not established. Aggregate unemployment, GDP, crime rates, and stock indices almost always belong here, whichever direction they moved.

Credible sources reach incompatible conclusions using defensible methods. The entry shows the incompatibility rather than picking a side.

03 —

Primary documents and federal statistical agencies — the Federal Register, Congress.gov roll calls, court opinions, CBO, JCT, BLS, BEA, CDC, GAO, agency inspectors general. Cited without qualification.

Peer-reviewed research and nonpartisan institutions with published methodology. Cited by name, with the method noted where it drives the result.

Advocacy and party-affiliated sources. Admissible, always labeled with their orientation, and never cited alone — a Tier 3 claim appears only alongside an opposing-orientation source making the competing claim.

Anonymous sourcing with no documentary corroboration, single-outlet reporting on contested factual matters, and any claim that cannot be traced to a document a reader can open.

05 — Editorial & legal posture

Truth and the public-figure standard

A sitting or former president is a public figure: liability for defamation requires actual malice — knowledge of falsity or reckless disregard for truth. Rigorous sourcing and a documented corrections process are the practical defense, and they are the same discipline that makes the site credible.

Report characterizations as characterizations

"The court found X" is a fact. "He lied about X" is a characterization. Write the first, attribute the second to whoever said it, and the legal and editorial risks fall together.

Fair report privilege

Accurate accounts of official proceedings and public records carry strong protection in most states. Anchoring entries to court filings, roll calls, and agency reports is both the best sourcing and the safest posture.

Structure, funding, and disclosure

Publish who funds the project, who edits it, and how entries are approved. Consider a nonprofit or fiscally sponsored structure, media liability insurance, and a named editorial contact for disputes.

Open the data

Publish the entry database in a machine-readable format under an open license. If a critic can rebuild your analysis from your own data and reach a different conclusion, that is a feature — it proves the methodology is real.

Nothing here is legal advice. Before publishing, have a media-law attorney review the corrections policy and at least a sample of entries.