Datasetpublic

abigailhaddad/legislative-issue-tracker

Legislative Issue Tracker Bills, legislative actions, floor speeches, hearings, and committee reports that touch a specific federal statute — currently the Paperwork Reduction Act (44 U.S.C. ch. 35, subch. I) — with every mention classified as amends, exempts, references, or related. The distinction is the point: Congress amends the PRA rarely (254 bills) but exempts individual programs from it constantly (845 bills). Built by abigail-64/legislative-issue-tracker. Everything… See the full description on the dataset page: https://huggingface.co/datasets/abigailhaddad/legislative-issue-tracker.

sourceHugging Facecc0-1.0updated 15h agoView on Hugging Face

0likes1.7kdownloads

Dataset Card

Legislative Issue Tracker

Bills, legislative actions, floor speeches, hearings, and committee reports that touch a specific federal statute — currently the Paperwork Reduction Act (44 U.S.C. ch. 35, subch. I) — with every mention classified as amends, exempts, references, or related.

The distinction is the point: Congress amends the PRA rarely (254 bills) but exempts individual programs from it constantly (845 bills).

Built by abigail-64/legislative-issue-tracker. Everything here is derived from public government APIs (Congress.gov, GovInfo, EveryCRSReport.com). Only short passages around each statutory citation are stored — never full document text.

Files

data/*.parquet — pipeline state:

file	one row per
`bills`	bill (Congress.gov metadata: sponsor, dates, latest action, law)
`actions`	recorded legislative action
`bill_issues`	bill x issue link: classification, passages, version list
`candidates`	bill text version found by search or bulk-data scan
`text_matches`	(issue, package) analysis cache: classification + passages
`text_diffs`	consecutive text-version pair: change stats + issue-relevant hunks
`speeches`	Congressional Record section: speakers, speech vs inserted bill text
`documents`	hearing or committee report mentioning the issue
`cosponsors`	bill x cosponsor
`crs`	CRS report matched by title
`scan_state`	bulk-data scan watermark, per issue
`versions_done`	bill whose text-version list has been read from Congress.gov
`unavailable`	upstream record that keeps failing, so the pipeline stops retrying
`usc_credits`	enacted bill: the U.S. Code sections and package the Code's own amendment credits name it under
`summary_checks`	machine-generated: an LLM's comparison of a curated timeline summary against the CRS summary of the same law

summary_checks is the only table here that is not derived mechanically from a government source. It holds proposals — a model's opinion about whether a curated sentence is supported by the Congressional Research Service's own summary — recorded with the model that produced it. It exists to decide what a person reads first. Nothing in this dataset is marked verified on its say-so, and it should not be treated as ground truth about any Act of Congress.

site/*.json — the same data shaped for the web UI.

Quick start

python

import duckdb

duckdb.sql("""
  SELECT match_type, count(*) AS bills
  FROM read_parquet('hf://datasets/abigailhaddad/legislative-issue-tracker/data/bill_issues.parquet')
  GROUP BY 1 ORDER BY 2 DESC
""").show()

Caveats

—Classification is rule-based — regex over the clause containing each citation — not human-reviewed. confirmed = true marks the small number of hand-tagged rows.
—Coverage is 1993–present via GovInfo full-text search, plus bulk-data scans of the 113th Congress onward. Earlier bills are hand-tagged only.
—44 U.S.C. ch. 35 also contains FISMA (subchapter II) and CIPSEA (subchapter III). Those are classified related, not PRA.
—In the Congressional Record, most mentions are bill or amendment text reprinted into the Record rather than a member speaking; speeches.mention_kind distinguishes them.