Evidence before assertion
Significant claims should be supported by public records, court materials, official actions, credible reporting, or established institutional research.
Editorial standards · Evidence, judgment, accountability
How we work to earn your trust.
Democracy Redline makes consequential judgments about the condition of American democracy. Those judgments should never require blind trust. We show what evidence supports the warning, how it affects the score, what meaningful safeguards still held, and when a published conclusion has been corrected or clarified.
Our commitments to readers
Six standards govern every published warning.
These commitments are not a request for agreement. They are the standards readers can use to examine, question, and challenge our work.
Corroboration before confidence
When a consequential claim depends on reporting rather than a primary record, we look for independent confirmation and supporting documentation.
Facts separated from judgment
We distinguish what happened, why it may matter for democratic government, and whether it is significant enough to affect the score.
Counterevidence included
We consider court rulings, institutional resistance, election safeguards, oversight, and other evidence that may limit, delay, or reverse deterioration.
Human editorial responsibility
Research tools can identify and organize developments. They do not determine the score. Final judgments remain human decisions for which Democracy Redline accepts responsibility.
Corrections without concealment
Material errors are corrected visibly. Clarifications and significant methodological changes are labeled so readers can understand what changed and why.
From event to warning
How a development becomes part of the score.
We evaluate patterns of institutional change, not the temperature of a news cycle.
A development is identified
We monitor official records, court actions, credible reporting, watchdog findings, civil-liberties analysis, election-administration materials, and established democracy research.
The underlying evidence is examined
We look for original documentation, corroboration, institutional reach, recurrence, and direct relevance to democratic safeguards.
Its significance is evaluated
We ask whether the development is isolated, confirms an existing pattern, worsens a category, or acts as a meaningful countervailing safeguard.
It is placed in context
Events are considered as part of an accumulating institutional pattern, not scored according to outrage, popularity, or headline volume.
A human editorial judgment is made
The score reflects structured judgment applied to the nine-category model. No feed, formula, headline count, or AI system independently changes it.
The reasoning is made inspectable
Readers can examine the visual score explanation, deeper analysis, category changes, evidence record, and methodology.
Evidence that can influence the score
We look for credible evidence with democratic consequence.
- Official actions and public records
- Court rulings, filings, and compliance evidence
- Corroborated investigative or institutional reporting
- Election-administration changes
- Documented civil-liberties and press-freedom developments
- Oversight findings and credible watchdog reports
- Repeated conduct establishing a broader pattern
What does not move the score by itself
Attention is not the same as evidence.
- Partisan disagreement
- Viral claims or unverified allegations
- A single inflammatory quotation
- Opinion presented without supporting evidence
- Social-media popularity
- Predictions about events that have not occurred
- Symbolic resistance with no practical constraining effect
Uncertainty and disagreement
Structured judgment is not mathematical certainty.
Reasonable people may disagree about the significance or weight of particular developments. Our responsibility is not to pretend judgment can be eliminated. It is to make that judgment visible enough to examine, question, and challenge.
Confidence follows evidence
Claims supported by primary records and independent corroboration warrant more confidence than emerging or contested accounts.
Allegations remain allegations
We do not present an accusation as an established fact, and we state uncertainty when it could materially affect a conclusion.
Importance is not movement
A disturbing event may deserve attention without being large, durable, or verified enough to change a category score.
Safeguards count by effect
Resistance receives weight according to what it prevented, delayed, reversed, or constrained, not merely what it declared.
Patterns outrank volume
The score does not move simply because one month produces more headlines. We evaluate institutional reach, recurrence, and consequence.
Counterarguments stay visible
Evidence that complicates or limits the warning belongs in the public explanation, even when it does not change the final score.
Corrections and transparency
Meaningful changes should be visible.
When a change could materially affect how a reasonable reader understands a report, we will note what changed, when it changed, and whether the score or conclusion was affected.
Correction
A factual error, faulty link, incorrect attribution, numerical mistake, or material misstatement was repaired.
Clarification
Language or context was improved without changing the underlying evidence judgment or score.
Methodology update
The scoring framework or public explanation of the model changed in a way readers should understand.
Current public record: No material correction affecting the August 2026 score has been recorded. If that changes, the correction will be identified here and on the affected publication.
Historical integrity
Each month remains part of the public record.
When a new warning is published, earlier reports remain available as records of what was known, documented, and judged at that time. Historical pages are labeled by month and should not silently change to reflect later information. When a material correction is necessary, it is identified rather than concealed inside a rewritten archive.
Verify or challenge the work
You do not have to accept the score on authority.
Start with the visual explanation, examine the analytical judgment, inspect the underlying evidence, and review the methodology. If you find a factual error, missing source, or material inconsistency, tell us.
An independent monthly warning system tracking democratic stress in the United States.