Standards · How we work
Changelog
A dated record of data corrections, database refreshes, and feature changes to PlainSpending. Every entry here is corroborated by our source history, see our corrections process for how a reported issue becomes an entry.
- Correction
Per-capita "leader" claim fixed for single-recipient concentration
A state per-capita ranking could be driven by one large, concentrated recipient rather than broad-based spending. The affected claim was rewritten to disclose the concentration instead of stating a plain "leader."
- Correction
Per-capita rankings caveated for USASpending "multiple recipients" placeholder
USASpending.gov uses a "MULTIPLE RECIPIENTS" placeholder for some award records. Per-capita rankings that could be distorted by this placeholder now carry an explicit caveat.
- Data refresh
Database rebuild, DOGE state enrichment + OMB outlay categories
Full data rebuild (etl_version 1, stage:base v2) adding DOGE savings broken out by state and OMB outlay-category classification.
- Data refresh
OMB outlay categories seeded
Added Office of Management and Budget outlay-category classifications to the spending dataset.
- Correction
Fixed: USASpending placeholder wrongly ranked #1 recipient in all 56 states
A data-processing bug let the "MULTIPLE RECIPIENTS" placeholder rank as the #1 recipient across every state, and produced a broken $0 narrative on affected pages. Traced to the source join, fixed at the ETL, and every affected page was re-rendered.
- Data refresh
County-state and agency-award-category backfill
Backfilled state assignment for counties and agency award-category classifications across the existing dataset.
- Fix
DOGE tracker routing fix
Fixed database adapter parameter numbering and orphaned-parent redirect handling introduced by the DOGE Impact Tracker launch.
- Launch
DOGE Impact Tracker added
Added agency- and state-level tracking of DOGE-reported contract and grant terminations (initial load: ~29,000 terminations).
Notice something that looks wrong? Report it - see our corrections process.