Explainer
How Federal Bills
Actually Work
The version taught in civics class — introduced, debated in committee, voted on, signed — is real, but it describes the exception. Congress’s own researchers open their guide to the process by saying it is “rarely predictable” and varies “significantly from bill to bill.” Here is the version the paperwork actually shows.
The Short Version
Almost all bills die quietly in committee — of the 17,000-plus introduced in a two-year Congress, roughly nine in ten never receive any vote at all. Of the bills that do pass, most are uncontroversial measures moved under fast-track procedures. The contested lawmaking that affects your life increasingly happens inside a few giant, must-pass packages negotiated by leadership and voted on quickly — which is exactly where provisions that could not pass on their own catch a ride.
Two skills protect a reader: knowing that a bill's title is chosen marketing with no legal force, and knowing where to look instead — the operative text, the money language, and the disclosure filings that show who is pushing. This page covers both, and is honest about what the public record cannot show.
The real pipeline, with base rates
The textbook diagram isn't wrong about the stages — it's wrong about the traffic. Here is how the volume actually flows, with the measured rates at each choke point.
Most bills die quietly in committee
The vast majority of bills are never voted on at all: they are referred to a committee and no further action is taken.
Of the 17,000–19,000 measures introduced in a recent two-year Congress, roughly nine in ten had no action beyond committee referral, and 1–3% were enacted into law (GovTrack historical statistics; verified against the live table 2026-07-13).
What our data shows: Bill status and action history for federal legislation, via the Congress.gov data our bill pages are built on. GovTrack, Historical Statistics about Legislation · CRS R42843, Introduction to the Legislative Process
Suspension of the rules: where most floor votes live
A fast-track House procedure for bills leadership considers uncontroversial: debate is capped at 40 minutes, no amendments are allowed, and passage needs a two-thirds vote.
In the 117th Congress, 783 of 1,106 measures receiving House floor action (71%) moved under suspension, with average debate under 11 minutes (CRS R48591).
What our data shows: Which recorded votes were taken under suspension is present in the vote metadata we ingest; not yet surfaced as a filter. CRS R48591, Suspension of the Rules: House Practice in the 117th Congress
Conference committees are mostly gone
The textbook step where a House–Senate conference committee reconciles two versions of a bill has largely been replaced by leadership-negotiated text shuttled between the chambers ("ping-pong").
CRS reports that conference-committee use "has decreased" in recent Congresses, and more than three-quarters of legislation that became law passed the second-acting chamber without amendment (CRS R41003).
What our data shows: Bill action histories show amendment-exchange sequences; we do not currently classify resolution method. CRS R41003, Amendments Between the Houses
The 60-vote wall and the reconciliation bypass
Most major Senate legislation effectively needs 60 votes to end debate (cloture). Budget reconciliation is the standing exception: a once-a-budget-cycle vehicle that passes with a simple majority but is restricted to budgetary provisions by the Byrd Rule.
Cloture motions filed per Congress grew from 137 (111th) to 336 (117th), and most recent party-line landmark laws — the 2017 tax law, the American Rescue Plan, the Inflation Reduction Act — moved through reconciliation (senate.gov; CRS R48640).
What our data shows: Vote records for cloture and passage votes; we do not currently tag which laws moved via reconciliation. U.S. Senate, Cloture Motions by Congress · CRS R48640, The Senate’s Byrd Rule: FAQ
Reading a bill without being misled
Bills are layered documents with a grammar of their own. Four rules cover most of what trips up a first-time reader. (The Congressional Research Service's guide, “Understanding Federal Legislation”, is the best free deep version of this section.)
Findings and purposes are not the law
The opening "findings" and "purposes" sections of a bill explain why Congress says it is acting. They generally create no rights or duties — only the operative "shall / may not" provisions do.
CRS’s bill-reading guide: statements of purpose and findings "generally do not create legal rights or duties," and a bill’s captions and title "may not [carry] much weight" with a court (CRS R46484).
What our data shows: Full bill text with links to the official version on Congress.gov, so readers can jump past findings to operative text. CRS R46484, Understanding Federal Legislation
Most bills edit other laws
Major bills rarely state rules from scratch; they amend existing statutes ("strike subsection (b) and insert…"). What a bill does is often invisible without reading the law it edits.
Committee reports must include a comparative print showing exactly what a bill changes in existing law — the Ramseyer rule in the House (Rule XIII cl. 3) and the Cordon rule in the Senate (Rule XXVI ¶12). That "Changes in Existing Law" section is the cleanest public redline.
What our data shows: Links to committee reports and to the current U.S. Code text a bill amends. CRS 98-305, Senate Committee Reports: Required Contents · Office of the Law Revision Counsel, U.S. Code guide
Effective dates and sunsets shape what a law really is
Provisions can start years later or expire on a schedule. Expiration dates are also a budget-scoring device: a provision "ending" inside the ten-year window scores cheaper even when everyone expects an extension.
The Committee for a Responsible Federal Budget catalogs 20 recurring budget-gimmick types, including arbitrary sunsets and shifting costs outside the scoring window (CRFB, "Playing By the (Budget) Rules," 2018).
What our data shows: Not currently extracted; sunset tracking is a candidate for future bill-page metadata. CRFB, Playing By the (Budget) Rules (2018)
Why titles mislead
A short title is the one part of a bill written purely for the public. Courts give it almost no weight; readers give it almost all of theirs. The register we use on this site is descriptive on purpose: a title describes X while the operative provisions chiefly do Y — a checkable statement, not an accusation of intent.
Short titles are marketing
A bill’s short title is chosen by its drafters and has no legal force over the operative text. Acronym and slogan titles have grown sharply as a persuasion device.
Reverse-engineered acronym titles ("backronyms") appeared roughly three times before 1988 and nearly seventy times in the following two decades (Sagers, Georgetown Law Journal, 2015); by 2021–22 about 10% of introduced bills carried one (The Atlantic, 2022, data by Noah Veltman). Experimentally, acronym titles are about twice as memorable (Sheppard, Moshirnia & Sullivan, NYU J. Legis. & Pub. Pol’y, 2022).
What our data shows: Our bill classification records a bill’s stated subject; comparing stated subject with who actually lobbies in support is computed today (see #lobbying-vs-stated-subject) but not yet surfaced in the UI. Sagers, 103 Geo. L.J. 1307 (2015) · Wong, The Atlantic (Nov. 12, 2022) · Sheppard, Moshirnia & Sullivan, 24 NYU J. Legis. & Pub. Pol’y (2022)
No federal single-subject rule
Forty-three state constitutions require a bill to stick to one subject, clearly expressed in its title. The U.S. Constitution has no such rule, which is why federal packages can carry unrelated provisions under one name.
The single-subject requirement exists in 43 state constitutions; the federal government has none (State Court Report / Council of State Governments).
What our data shows: Nothing to compute — this is structural context for why the mechanisms on this page are legal. State Court Report, single-subject rules analysis
When the backers don’t match the label
A measurable divergence signal: industries lobbying in support of a bill that do not appear anywhere in the bill’s stated subject. It does not prove intent — it flags where the label and the interest map disagree.
In our own audited computation across federal bills with both subject tags and lobbying positions, roughly one in six showed unstated supporting industries (PTP subterfuge audit, 2026-05-07; methodology on file).
What our data shows: Computed today per bill (unstated supporters and a divergence score, from Lobbying Disclosure Act positions vs. stated subject tags); not yet surfaced in the UI. When it ships, its tooltip will link here. Lobbying Disclosure Act filings (Senate LDA database)
Must-pass vehicles and their passengers
When something must pass — government funding, defense authorization, a debt-limit increase — it becomes a vehicle, and vehicles attract passengers. This is the modern home of most contested policy change.
Omnibus packages and continuing resolutions
When the twelve annual spending bills don’t pass on time, Congress runs the government on stopgap continuing resolutions and then bundles spending into giant omnibus packages negotiated by leadership — often released days or hours before the vote.
Congress has enacted at least one continuing resolution in all but three fiscal years since FY1977 (207 CRs through FY2025); FY1997 was the last year every regular appropriations bill passed on time (CRS R46595).
What our data shows: Bill pages for appropriations vehicles, including the full text and action timeline. CRS R46595, Continuing Resolutions: Overview
Policy riders
Provisions attached to a bill that must pass — often ones that bar an agency from spending money on a specific activity. The urgency of the vehicle carries the rider past opposition it could not survive alone.
Appropriators themselves count them: Sen. Barbara Mikulski described "about 40 or 42" policy riders in the FY2016 omnibus negotiations (Roll Call, Dec. 2015); advocacy tallies of a single draft FY2024 bill counted at least 51 (Clean Budget Coalition).
What our data shows: Full text of appropriations bills; rider-level extraction is not something we compute today. Roll Call (Dec. 8, 2015), Mikulski on omnibus riders · Clean Budget Coalition, FY2024 FSGG rider count
Earmarks (now "Community Project Funding")
Provisions directing funds to a specific named recipient or project. Banned by party rule in 2011, they returned in 2021 under new names with each request publicly attributed to its requesting member.
The FY2008 omnibus drew nearly 9,000 earmark requests; the modern replacement regime (since Feb. 2021) publishes every funded project with its requesting member (Roll Call, 2007; CRS R46722).
What our data shows: Earmarks are the most joinable influence artifact — member-attributed by design. Not yet ingested; a candidate dataset for future member pages. CRS R46722, Community Project Funding · Roll Call (Dec. 17, 2007), FY2008 omnibus earmarks
Who actually writes the text
No single author writes a federal bill. Member offices, committee staff, and two nonpartisan drafting offices assemble it — and outside interests participate earlier and more concretely than most people assume, in ways the research has actually measured.
Outside interests propose; staff and counsel dispose
Lobbyists and interest groups routinely supply draft bill language. Committee staff treat those drafts as starting points and pass them to the nonpartisan Offices of Legislative Counsel, who rewrite them into statutory text — the research finds staff do not adopt lobbyist language verbatim.
In the canonical interview study of Senate committee counsels, staffers reported lobbyists are involved in drafting but insisted they "never used lobbyists’ language verbatim": "No one simply drops in the language of the lobbyist" (Nourse & Schacter, NYU Law Review, 2002).
What our data shows: Which organizations lobbied on which bills (from LDA filings naming bill numbers), and which lobbyists previously held government positions — the observable trace of the propose side. Nourse & Schacter, 77 N.Y.U. L. Rev. 575 (2002) · CRS RS20735, The House Office of Legislative Counsel
Influence concentrates on must-pass bills
Lobbying activity clusters on the handful of bills guaranteed to become law. Well-organized interests arrive with language pre-positioned; short text-to-vote windows favor them over everyone else.
The 2014 "cromnibus" spending package drew lobbying from 852 distinct clients — nearly double the next most-lobbied bill of that year (475) — and passed the House less than three days after its text was released (OpenSecrets, 2015).
What our data shows: Per-bill lobbying-client counts and registrant lists from quarterly LDA filings. OpenSecrets (Feb. 2015), Crowding in on cromnibus
What disclosure shows — and hides
Much of this page is possible because lobbying is disclosed under oath. But the window has edges, and honesty requires drawing them.
The Lobbying Disclosure Act window
Registered lobbyists file quarterly reports naming their clients, the issues (often specific bill numbers) they lobbied, which chambers and agencies they contacted, and their prior government positions — but not which side they argued, or which provision.
The LD-2 form requires "specific lobbying issues," and registration is triggered only above a threshold: more than one lobbying contact plus at least 20% of a person’s time for that client over three months (House LDA guidance).
What our data shows: Everything in the filings: registrant, client, bill numbers, lobbyist names, covered-position history — the substrate of our Revolving Door data. House Clerk, Lobbying Disclosure Act guidance · Senate LDA filing database
The disclosure gap ("shadow lobbying")
Advocacy that stays under the registration thresholds — strategic advisers, sub-threshold contacts, grassroots campaigns — appears nowhere in the filings. Any lobbying dataset, including ours, undercounts influence for this reason.
Survey-based research estimates roughly one unregistered policy-advocacy professional for every registered lobbyist (LaPira & Thomas); a 2023 working paper tracking ex-congressional staff at lobbying firms measured that specific shadow population at 3.8–10.7% of the registered workforce (d’Este, Draca & Fons-Rosen, CAGE wp652) — different populations, both real.
What our data shows: Only registered activity. This gap is why our surfaces describe lobbying data as a floor, not a census. LaPira & Thomas, Revolving Door Lobbying (Univ. Press of Kansas) · d’Este, Draca & Fons-Rosen, Shadow Lobbyists, CAGE wp652 (2023)
Does money buy votes? What the evidence says
The most common assumption about Congress — donations purchase floor votes — is also the one the evidence supports least. What the research finds instead is more specific, and more useful to a reader of this site.
Direct tests of contributions-for-votes mostly come up empty or mixed. The most-cited review found little robust effect of PAC money on roll-call voting once party and ideology are accounted for (Ansolabehere, de Figueiredo & Snyder, Journal of Economic Perspectives, 2003); later re-analyses found significant effects in a minority of tests (Stratmann, Public Choice, 2005) — a genuinely mixed record, not a settled one. We ran a pre-registered version of this test ourselves on recent congressional voting and industry money and found the same null: after conditioning on party and correcting for multiple comparisons, almost no member-industry pairs showed a defensible link between contributions and votes.
Where the evidence isstrong is upstream of the vote. Donors get meetings: in a randomized field experiment, congressional offices were three to four times more likely to grant senior-staff meetings when attendees were identified as donors (Kalla & Broockman, American Journal of Political Science, 2016). Money follows power: industries shift contributions toward members who gain agenda-setting positions (Fouirnaies, AJPS, 2018) and pull them from members exiled from relevant committees (Powell & Grimmer, Journal of Politics, 2016). And most money supports allies rather than converting opponents (Bonica, AJPS, 2014).
That is why this site's surfaces juxtapose money, committee seats, lobbying, and votes without asserting causation: the documented influence channels are access and agenda, and the honest way to show them is side by side, with the inference left to you.
What Our Data Shows — and Doesn't
Shows: bill text, status, and vote records from the official Congress.gov record; which organizations lobbied on which bills, from sworn Lobbying Disclosure Act filings; which lobbyists previously held government positions; and campaign-finance totals from FEC data — each linked to its primary source.
Doesn't show:which side a lobbyist argued (filings don't say), advocacy below the registration thresholds, who asked for a specific provision, or anyone's motive. Where a pattern is computable — like a mismatch between a bill's stated subject and the industries supporting it — we present it as a flag to investigate, never as a conclusion.
Glossary
- Bill vs. resolution
- Bills (H.R., S.) and joint resolutions can become law. Concurrent and simple resolutions cannot — they express positions or set internal rules.
- Markup
- The committee meeting where members amend and vote on a bill before it can reach the floor.
- Filibuster / cloture
- In the Senate, debate on most legislation continues until 60 senators vote to end it ("cloture") — the practical source of the 60-vote threshold.
- Reconciliation
- A budget procedure that lets certain tax-and-spending legislation pass the Senate with a simple majority, limited by the Byrd Rule to budgetary matter.
- Suspension of the rules
- Fast-track House procedure: 40 minutes of debate, no amendments, two-thirds to pass. Where most uncontroversial bills move.
- Rider
- A provision attached to a must-pass bill, often unrelated to it, that would struggle to pass on its own.
- Omnibus / continuing resolution (CR)
- An omnibus bundles many spending bills into one package; a CR extends current funding temporarily when the bills aren’t done.
- Authorization vs. appropriation
- An authorization permits a program and caps its funding; an appropriation actually provides the money. Only appropriations spend.
- Engrossed / enrolled
- Engrossed = the official text one chamber passed. Enrolled = the identical final text both chambers passed, sent to the president.
- Conference committee
- A temporary House–Senate panel that reconciles differing versions of a bill. Now rare; leadership negotiation usually replaces it.
- Byrd Rule
- The Senate rule that strips non-budgetary ("extraneous") provisions out of reconciliation bills.
- Findings
- A bill’s opening statements about why Congress is acting. Context, not law — they create no rights or duties.
Sources & Further Reading
- Congressional Research Service, “Introduction to the Legislative Process in the U.S. Congress” (R42843) — everycrsreport.com
- Congressional Research Service, “Understanding Federal Legislation: A Section-by-Section Guide” (R46484, Killion) — everycrsreport.com
- Congressional Research Service, “Authorizations and the Appropriations Process” (R46497, Saturno) — everycrsreport.com
- Barbara Sinclair, Unorthodox Lawmaking: New Legislative Processes in the U.S. Congress, 5th ed. (CQ Press, 2016) — sagepub.com
- Nourse, V. & Schacter, J., “The Politics of Legislative Drafting: A Congressional Case Study,” 77 N.Y.U. L. Rev. 575 (2002) — nyulawreview.org
- Kalla, J. & Broockman, D., “Campaign Contributions Facilitate Access to Congressional Officials,” AJPS 60(3), 2016 — onlinelibrary.wiley.com
- Ansolabehere, S., de Figueiredo, J. & Snyder, J., “Why Is There So Little Money in U.S. Politics?” J. Econ. Perspectives 17(1), 2003 — aeaweb.org
- U.S. Senate, Cloture Motions by Congress — senate.gov
- Per-mechanism sources are cited inline in each entry above; the full research file behind this page, including the verification pass, is available on request via our methodology page.
Last reviewed 2026-07-12. Corrections to this page are logged and dated; if you spot an error, use the contact link in the footer.