Glossary
What our detectors look for, what the legislative terms mean, and why the patterns matter. Plain language, no jargon.
The Detectors
Each detector scans official congressional data for a specific statistical pattern. When a bill exceeds the threshold, a signal enters the editorial queue for human review.
The Ledger — Fiscal Concentration
Compares who a bill names as beneficiaries (families, workers, veterans) against where its money actually goes (agencies, contractors, corporations). Flags bills where institutional dollar allocation exceeds individual beneficiary language by 2:1 or more, with at least $1M in institutional funding.
The Fast Track — Stealth Velocity
Identifies bills that went from introduction to law in under 60 days with fewer than 5 cosponsors. The median bill takes roughly 340 days. Speed plus low visibility puts a bill in a statistical minority worth examining.
The Kitchen Sink — Omnibus Bloat
Flags bills that span an unusually high number of unrelated policy domains. When a single vote covers dozens of subjects, constituents cannot track what their representative actually supported.
The Divide — Party-Line Votes
Detects roll call votes where both parties voted with near-perfect unity on opposite sides — Democrats above 90% unified one way, Republicans above 90% unified the other. The vote became a partisan marker, not a policy debate.
The Stampede — Rush Job
Flags bills that reached a floor vote within 3 days of introduction. Same-day votes leave minimal time for public review, committee examination, or amendment.
The Shortcut — Committee Bypass
Identifies enacted legislation that received fewer committee reviews than normal. Committees are where bills get vetted in public — when that step is skipped, transparency is reduced.
The Rewrite — Most Amended
Tracks bills with an unusually high number of text revisions. When the final version differs substantially from the introduced text, original endorsements were based on a different bill.
The Shell Game — Amendment Contradiction
Measures text divergence between a bill’s introduced and final versions using computational comparison. High divergence, extreme text growth (gut-and-replace), or formally amended titles generate a signal.
The Name Game — Title Mismatch
Compares populist keywords in bill titles against the actual funding structure. When the title says “families” but the dollars go to institutions, the gap between branding and substance is measurable.
The Eraser — Quiet Repeal
Flags bills that seek to repeal, abolish, or sunset existing law with low cosponsorship. Undoing a law deserves the same scrutiny as creating one.
The Fog — Opacity
Identifies bills with significant cosponsorship or legislative activity that lack a Congressional Research Service summary — the standard plain-language explanation that helps the public understand what legislation does.
The Press Box — Media Pressure
Measures media coverage volume, outlet diversity, tone distribution, and polarization for enacted legislation. Flags bills whose coverage is 3× the median or whose outlet tone is highly polarized.
The Dead Letter — Ghost Bill
Flags legislation stuck at its initial referral for over 180 days with few cosponsors. These bills were introduced but show no sign of advancing.
The Rerun — Zombie Bill
Tracks legislation reintroduced across three or more consecutive congresses without ever being enacted. The pattern may reflect persistence or positioning.
The Tombstone — Bill Dead
Bills stuck at committee referral for 60+ days with no hearing, markup, or vote. In the 118th Congress, zero bills in this position became law.
The Revival — Lazarus Bill
Bills that showed new legislative progress after 90+ days of dormancy. Revivals are statistically rare and indicate something changed — a deal, a political shift, or attachment to a must-pass vehicle.
The Blindfold — Deliberation Gap
Flags bills where text changed too close to a floor vote for anyone to meaningfully read what was in the bill. Detects same-day text revisions near passage, compressed version timelines (multiple versions in 3 days or fewer), and last-minute text growth. When the text changes faster than legislators can read it, the vote is not informed by the text.
Legislative Terms
Committee Referral
After a bill is introduced, it is assigned to one or more committees with jurisdiction over its subject matter. Most bills die at this stage without a hearing.
Cosponsor
A member of Congress who formally signs onto a bill in support, but is not the original sponsor. Cosponsorship is a public signal of support and can indicate how much backing a bill has.
Enrolled Bill
The final version of a bill that has passed both chambers of Congress in identical form. This is the text sent to the President for signature.
Floor Vote
A vote taken by the full House or Senate, as opposed to a committee vote. Roll call votes record how each member voted individually.
Markup
The committee process of debating, amending, and rewriting a bill before it goes to the full chamber. This is where most substantive changes happen.
Omnibus Bill
A single bill that packages together many separate measures, often spanning unrelated policy areas. Legislators must vote yes or no on the entire package.
Roll Call Vote
A recorded vote where each member’s position (yea, nay, present, not voting) is individually logged. This is the public record our party-line detector analyzes.
CRS Summary
A plain-language description of a bill written by the nonpartisan Congressional Research Service. When available, it is the most accessible way for the public to understand what legislation does.
How Scores Work
Every signal carries an anomaly score between 0 and 1. This is not a judgment — it measures how far the bill deviates from the statistical norm for its detector category.
- 0.70 – 1.00 — Strong statistical outlier
- 0.40 – 0.69 — Moderate outlier
- 0.00 – 0.39 — Mild outlier
A high score means the pattern is pronounced. It does not mean the bill is bad, corrupt, or unusual for the wrong reasons. Every score requires human editorial review before publication.