Skip to content

bill-publishing.md

Latest commit

 

History

History
115 lines (93 loc) · 6.96 KB

bill-publishing.md

File metadata and controls

115 lines (93 loc) · 6.96 KB
Raw

How appropriations bills are published: formats, venues, and timing

The stock answer to a question that comes up constantly: for a given version of an appropriations bill, what formats exist, where, published by whom, and when? This drives why we have both an XML and a PDF input path (see ADR 0010) and when to steer a user to one over the other.

Scope: the twelve regular appropriations bills in each chamber, drafted by the twelve House and twelve Senate Appropriations subcommittees. The same pattern holds for ordinary legislation, but appropriations is where the timing gap bites, because the high-value diffing happens before a bill is formally published.

The one fact that drives everything

A bill is authored in XML, but the XML is not public until GPO publishes it — which happens after the committee stages where staffers most need to compare versions. At the chair's-mark and markup stages, the only public artifact is a PDF. The authoritative XML appears one to two days after the bill is formally introduced or reported, as a downstream product of the Government Publishing Office (GPO), not the committee.

You can see the XML existing privately: the committee-print PDFs posted on docs.house.gov carry their internal source path in the document title (e.g. U:\2026OMNI\MIN41203.xml). The bill is XML upstream; what the public gets at that moment is the PDF rendered from it.

Who publishes what, and where

Actor Publishes Where Format(s)
House/Senate Appropriations committees Chair's mark, committee print, markup amendments, committee report docs.house.gov (Committee Repository), appropriations.house.gov, appropriations.senate.gov PDF (bill text); the only .xml on a Committee Repository event page is the meeting manifest, not the bill
House Rules Committee The version made in order for the floor rules.house.gov PDF, sometimes XML
GPO (Government Publishing Office) Authenticated bill text + status govinfo (BILLS collection), Bulk Data Repository XML, PDF, TXT, HTML
Library of Congress Mirrors GPO, adds bill metadata congress.gov XML (the "Download XML" link), PDF, TXT

The Senate publishes XML for schedules, votes, nominations, and floor proceedings on senate.gov, but no bill-text XML. Senate bill XML exists only downstream via GPO/govinfo. Senate Appropriations marks are posted as PDF on appropriations.senate.gov, same as the House.

The lifecycle of a version, by format

Reading top to bottom is the chronology a single bill travels.

Stage Public artifact Where Public XML?
Subcommittee mark (chair's mark) PDF committee print docs.house.gov / committee site No
Full-committee mark PDF committee print docs.house.gov / committee site No
Markup amendments PDF docs.house.gov No
Introduced / reported (gets an H.R./S. number) XML, PDF, TXT, HTML govinfo + congress.gov Yes — ~1–2 days after
Floor version (Rules print) PDF, sometimes XML rules.house.gov Sometimes
Engrossed / enrolled / public law XML, PDF, TXT, HTML govinfo + congress.gov Yes

The "No" rows are the pre-publication window. They are exactly the stages an appropriations staffer cares about most: comparing a chair's mark against last year's enacted bill or the President's request, and tracking amendments through markup. That is the access gap the PDF pipeline exists to close.

When does the XML show up, and how do I know?

  • Trigger. XML is produced when GPO processes a formally introduced or reported bill. A pre-introduction committee print does not get bill-text XML; it may get a committee-print (CPRT) entry, typically PDF.
  • Lag. Text is usually on govinfo and congress.gov one to two days after introduction. congress.gov refreshes from govinfo early each morning (~6 a.m. ET); a bill introduced today generally appears tomorrow. Delays grow when many bills land at once or a bill is very large.
  • How to check. Look up the bill on congress.gov; the Text tab lists each version with a Download XML link once GPO has it. The Bill Texts Received Today page shows the day's arrivals. On govinfo, browse the BILLS collection.

Coverage by Congress (history)

  • Bill-text XML: 113th Congress (2013) forward on govinfo. Earlier bills are PDF/TXT only.
  • Bill-status XML (metadata, not text): 108th Congress forward, via the GPO Bulk Data Repository.

Today tools/fetch_bills.py fetches bill text from govinfo bulk data by default (--source api selects the Congress.gov API v3 instead), and tools/fetch_bill_archives.py loads bill-status metadata from the govinfo bulk repository. Per-bill version enumeration also uses govinfo BILLSTATUS; only download-all year-range discovery still uses the Congress.gov committee API. See ADR 0004 for the rationale (issue #10).

Practical guidance: prefer XML when it exists

When a published XML version exists for the bill in hand, use it, and point users to it. XML is the better source — it is the authenticated structure the bill was authored in, so amount and hierarchy extraction is exact rather than reconstructed from a rendered page (the PDF path is inherently lossier; see ADR 0002 and the validation in ADR 0009). The PDF pipeline is for when no published XML exists yet — not a parallel choice for bills that have one.

A genuine discussion draft (released and never formally introduced) may never get public XML at all; for those, PDF is the only option there will ever be.

Sources