US WARN Layoff Notice Aggregator & Delta Monitor (OR/TX API)
Pricing
from $4.40 / 1,000 results
Aggregate US state WARN Act layoff & closure notices into one normalized schema and track new / updated / removed deltas. Oregon (HECC export) and Texas (TWC workbooks) live and keyless; CA/NY/WA/WI/IL/NJ scaffolded. For RevOps, recruiting, econ research and CRE. No API key.
State
state
Optional
Two-letter code of the state whose WARN registry issued the notice (the jurisdiction, not the company HQ).
Company / job site
company
Optional
Employer or job-site name named in the WARN notice.
City
city
Optional
City where the layoff or closure occurs.
County
county
Optional
County where the layoff occurs, where the state publishes it (null otherwise).
Notice date
notice_date
Optional
Date the WARN notice was filed / received by the state (YYYY-MM-DD).
Effective date
effective_date
Optional
Date the layoff or closure takes effect (YYYY-MM-DD).
Employees affected
employees_affected
Optional
Number of workers affected by the notice.
Layoff type
layoff_type
Optional
Normalized bucket: layoff, closure, or relocation (null where the state does not distinguish).
Industry
industry
Optional
Industry / sector where the state publishes it (null otherwise).
Region / workforce area
region
Optional
Workforce development area or region the site falls in, where published.
Notice ID
notice_id
Optional
State-assigned WARN number where one exists (null for states without an id; the delta uses a composite key then). NOT unique: Oregon files one WARN number per event and one row per site, so 330 live rows carry only 196 distinct numbers.
Source URL
source_url
Optional
The official state source the record was pulled from. Texas: the TWC yearly workbook (e.g. https://www.twc.texas.gov/sites/default/files/oei/docs/warn-act-listings-2026-twc.xlsx), or the Socrata resource URL on rows served by the fallback / pre-2020 archive (see source_backend). Oregon: the HECC export form.
Retrieved at
retrieved_at
Optional
ISO 8601 timestamp of this run.
Change type
change_type
Optional
Delta mode only: new, updated, or removed.
Old employees affected
old_employees_affected
Optional
Delta mode: employee count in the prior snapshot.
New employees affected
new_employees_affected
Optional
Delta mode: employee count in the current snapshot.
Old effective date
old_effective_date
Optional
Delta mode: effective date in the prior snapshot (YYYY-MM-DD).
New effective date
new_effective_date
Optional
Delta mode: effective date in the current snapshot (YYYY-MM-DD).
Old layoff type
old_layoff_type
Optional
Delta mode: layoff type in the prior snapshot.
New layoff type
new_layoff_type
Optional
Delta mode: layoff type in the current snapshot.
State name
state_name
Optional
Full name of the state whose WARN registry issued the notice.
Job-site state
site_state
Optional
Two-letter state of the actual job site WHERE THE SOURCE PUBLISHES IT. Texas is always TX. Oregon does NOT publish a job-site state column: it is read from a ", XX" suffix on the published location (21 of 330 rows on 2026-09-05, e.g. "Los Angeles, CA") and is null otherwise -- never a guessed "OR". See site_state_basis.
Site state matches jurisdiction
site_state_matches_jurisdiction
Optional
true when site_state equals the issuing state, false when the notice covers an out-of-state job site, null when the source does not publish a site state.
Layoff type (raw)
layoff_type_raw
Optional
The exact reason string the state published, before bucketing. null means the state does not publish a reason at all (Texas). A non-null raw value with a null layoff_type means the state published a value outside the three normalized buckets (Oregon "Other").
State received date
state_received_date
Optional
The date the state agency received the notice (YYYY-MM-DD). Texas publishes this separately from notice_date and the two differ on 1,249 of 2,368 live rows (52.7%). For Oregon the source publishes one date, so this equals notice_date.
Notice key
notice_key
Optional
Stable cross-run key for this exact notice row (state + notice id + company + city + effective date). Use it to join deltas. Oregon WARN numbers are NOT unique -- one WARN number can cover 19 store closures -- so notice_id alone is not a key.
Identity key
identity_key
Optional
Key for the notice-site independent of every changeable field (state + notice id + company + city). Two rows sharing an identity_key are the same site across runs even if the effective date moved.
Source dataset ID
source_dataset_id
Optional
Identifier of the dataset or export this row came from: "twc-warn-act-listings" for Texas rows read from the Texas Workforce Commission yearly workbooks (the primary since 2026-10-10), the Socrata id "8w53-c4f6" for Texas rows served by the Socrata fallback or the pre-2020 archive, or "hecc-warn-download" for Oregon's HECC export, which replaced the retired Socrata dataset ijbz-jpx8 in September 2026.
Source status
source_status
Optional
Per-source outcome for the state that produced this row: "ok" (the source answered) or "failed". A row only exists when its own source answered; use run_states_failed to see which OTHER requested states did not.
Source row count
source_row_count
Optional
Exact total row count the source reported for this dataset on this run -- for Texas from TWC, the distinct rows across every TWC year workbook (the workbooks overlap: the 2024 file repeats 2023, so duplicates across files are removed first; 2,274 on 2026-10-10; pre-2020 archive rows are not counted here); Socrata count(*) when Texas was served by the Socrata fallback; the parsed data-row count of the export for Oregon. The free truncation guard.
Source last notice date
source_last_notice_date
Optional
The newest notice date present in the source dataset on this run. If this is far in the past the registry itself has stopped publishing -- not an absence of layoffs.
Source stale days
source_stale_days
Optional
Whole days between source_last_notice_date and this run. Texas measured 4 days stale on 2026-10-10 from the TWC workbooks (the former Socrata source had frozen at 2026-06-23, 109 days). Oregon measured 4 days stale on 2026-09-05.
Source truncated
source_truncated
Optional
true when maxResults cut this state short of its matching rows, so the state is only partially represented. null when not determined.
Run: states requested
run_states_requested
Optional
Comma-separated list of the states this run attempted.
Run: states OK
run_states_ok
Optional
Comma-separated list of states that answered successfully.
Run: states failed
run_states_failed
Optional
Comma-separated list of requested states that did NOT answer. null when every state answered. A state listed here has no rows in this dataset for reasons of availability, not because it had no layoffs.
Run: failure reason
run_states_failed_reason
Optional
Per-state error text for every state in run_states_failed.
Run complete
run_complete
Optional
true only when every requested state answered. false means this dataset is a partial view of the requested registries.
Site state basis
site_state_basis
Optional
How site_state was determined. "published_location_suffix" = the source printed a ", XX" state suffix on the location. "not_published" = the source printed a bare city, so the job-site state is genuinely unknown and site_state is null rather than a guessed in-state code. "jurisdiction_is_single_state" = the registry only ever covers its own state (Texas). Oregon measured 21 of 330 rows with a published suffix on 2026-09-05.
Source format
source_format
Optional
How this state was retrieved: "twc_xlsx" for the Texas Workforce Commission yearly workbooks, "socrata" for a JSON open-data API (Texas fallback / pre-2020 archive), "hecc_xlsx" for Oregon's two-step HTML form + xlsx export.
Source window (verbatim)
source_window
Optional
The reporting window the source itself printed on the export, verbatim. Oregon prints "WARNs Received: 9/5/2016 to 9/5/2026" on the sheet -- the site chooses this window, it is not requested, so it is republished rather than assumed. null for sources that publish no window.
Source window start
source_window_start
Optional
Parsed start of source_window (YYYY-MM-DD). Notices older than this are outside what the export covers -- their absence is not evidence they do not exist.
Source window end
source_window_end
Optional
Parsed end of source_window (YYYY-MM-DD).
Source gate status
source_gate_status
Optional
Three-valued outcome of the live drift gate for this source: "ok" (every probe measured what it expected), "drift" (the source answered and the answer no longer matches its measured contract), "unavailable" (a probe could not complete, so nothing was learned about drift). A row only exists when its own source was "ok"; the other two appear in run_states_failed_reason.
Source probes passed
source_probes_passed
Optional
Comma-separated ids of the live drift probes that passed on this run (Oregon: form_page, report_link, xlsx_payload, sheet_present, header_contract, row_floor, source_window, positive_canary; Texas TWC: index_page, year_links, xlsx_payload, header_contract, year_consistency, date_parse, row_floor, positive_canary; null when Texas was served by the Socrata fallback). Published so a future tightening of the gate is evidence-based rather than guessed.
Source note
source_note
Optional
Plain-language statement of what this source does and does not publish, so a null column cannot be misread as an error. Oregon's says that HECC publishes no separate job-site state, county or industry column.
Source backend
source_backend
Optional
Texas only (null on Oregon): which backend served this row. "twc_xlsx" = the Texas Workforce Commission yearly workbooks (primary, current to within days). "socrata" = the frozen data.texas.gov copy 8w53-c4f6, used ONLY when TWC could not be reached (source_failover true). "socrata_archive" = a notice dated before 2020 -- TWC publishes no pre-2020 workbook, so those come from the Socrata copy on full-history runs.
Source failover
source_failover
Optional
Texas only (null on Oregon): true when the primary TWC workbooks were UNAVAILABLE on this run and the row was served by the frozen Socrata copy instead -- such rows can be months stale. false when the primary served the run (including planned pre-2020 archive rows). The fallback is never used when TWC answered but failed its contract checks; Texas fails instead.
Source failover reason
source_failover_reason
Optional
Why the Texas primary was not used, when source_failover is true (the TWC gate's unavailable probes). null otherwise.
