Who we cite, and how often.
Most outlets publish a sources page listing the institutions they intend to rely on. This one is generated from the site itself. Every figure below is counted directly from the source lists attached to our published stories, so it cannot drift from what we actually cite.
It also means the list is unflattering where it should be. Outlets we treat as secondary still appear often, because a story frequently needs a news report to establish that an event happened before a primary document exists. Our rule is in the methodology: a news citation points at a primary document or carries a verbatim quote. It is never the sole support for a dollar figure or a criminal charge.
- 5,272source citations
- 1,332distinct domains
- 1,024stories with source lists
- 258government and court domains
Charging documents, Inspector General reports, GAO audits, court dockets, Federal Register notices, congressional committee releases, and agency newsrooms. These are tier one and are cited first wherever they exist.
Secondary reporting. Used to point at a primary document, to carry a verbatim quote, or to establish that an event was reported — never as the sole support for a dollar figure or a charge.
Posts and broadcasts where the speaker is the source. Quoted as statements by that person, not as independent verification of what they assert.
Background and biographical lookups. Never load-bearing for a factual claim on its own.
Endpoints this site queries at runtime rather than citing in a story. Everything else on the site is a static citation to a document or article.
Counts cover the source lists attached to registry entries and regenerate on every deploy. They do not include the inline citations inside a story’s own sources panel, which are frequently broader. A domain appearing here is not an endorsement of that outlet, and a high count is not a measure of trust — it reflects how often a domain happened to hold a document or quote a story needed. Corrections: contact us.