WARN notices: separate planned layoffs and realized job losses
A WARN notice describes a planned employment event, while a later notice can change its date, scope or count. Adding each published notice as a new layoff can turn one evolving event into several imaginary ones. The focused research task is to preserve notice versions and separately establish what actually occurred.
Build a site-and-event ledger with original plan, revisions, cancellations and matched actual departures. This article concerns US federal WARN information read through October 25, 2026; it does not determine employer coverage, legal notice sufficiency, penalties or state-law obligations. State systems can differ, so a search result from one source is not a complete national count.
Locate the official version relationship
The US Department of Labor’s Extending WARN Notice page explains that a postponement notice refers to the earlier notice, new action date and reason, with different treatment for different postponement lengths. The operational research point is that a later document can continue an existing event rather than announce an additional one.
Use the source’s heading Can My Employer Extend the Notice? to locate the discussion. Do not use the invented June-to-August example below to decide whether a real employer met any statutory timing requirement. The actual dates, applicable regime and original documents would be needed for that different task.
Rebuild one planned event
Assume a fictional initial notice plans three hundred position exits at one site in June. A later document explicitly links to that notice and postpones two hundred to August while cancelling one hundred. The revised announced event contains two hundred planned exits, not five hundred.
| Version | Event treatment | Count |
|---|---|---|
| Initial notice | June plan | 300 |
| Linked revision | Two hundred postponed | 200 within the same event |
| Linked cancellation | One hundred removed | −100 from original scope |
| Latest planned total | Revised event population | 200 |
The two-hundred figure is the revised total population of the announced event, before matching implementation. It is not automatically the number still waiting to leave after actual departures begin. Preserve original and current plan fields rather than rename both remaining layoffs.
In July the company separately says eighty people have left the site. The statement alone does not establish that all eighty belong to the revised WARN population or meet the same employment-loss definition. Record eighty as company-reported departures with matching unresolved. Do not add it to the two-hundred plan or subtract it until identities and scope connect them.
Reconcile implementation only after matching
If later evidence establishes that all eighty departures belong to that same revised two-hundred population, remaining unimplemented planned exits are 200 − 80 = 120. The event now has three distinct numbers: revised planned total two hundred, matched actual departures eighty and outstanding planned exits 120.
That conditional reconciliation does not say the August plan was executed early merely because the July statement existed. The matching document is a necessary additional assumption. If twenty of the eighty instead belong to other voluntary departures, only sixty are matched, and outstanding planned exits would be 140 under the otherwise unchanged scenario.
Keep the departure date distinct from payroll payment, severance payment and announcement date. This worksheet does not treat those dates as equivalent or reconstruct restructuring expenses. An employee remaining on payroll or transferring elsewhere requires the actual record and applicable definition before a position count becomes an employment-loss count.
Test a different boundary: duplicate site names
Suppose a second publication uses the parent’s name while the first uses a subsidiary and a shortened site address. Similar names can refer to the same event. Match physical location, event identifier, dates and the document’s reference to the earlier notice before counting another two hundred.
The opposite error is merging genuinely distinct events because they share an employer. A later reduction at another site or in a separate population should remain separate where documents establish it. Use an event-family identifier and a version identifier, not company name alone.
Apply the notice verification sheet
| Field | Obtain | Check |
|---|---|---|
| Identity | Employer, site and notice reference | Parent/subsidiary and address variants |
| Plan | Original count and effective dates | Persons, positions and event population |
| Revision | Linked postponement or cancellation | Replaces or changes original, not added automatically |
| Actual | Dated company or employment records | Same site, population and definition |
| Remaining | Matched implementation bridge | Revised total less matched actual only |
| Search scope | Federal/state source and extraction date | Missing notice not treated as no job losses |
When counts conflict, inspect scope, revision date, phased windows and whether a table counts notices rather than affected people. Retain the older document with a superseded label instead of deleting it. This allows another reader to reproduce why the current plan differs from the original headline.
If actual implementation is unavailable, the deliverable should say current announced plan two hundred, realised count not established. Precision in notice counts cannot remove missing implementation evidence. If a public list supplies only employer and aggregate headcount without the notice text, the event linkage remains provisional.
Preserve the cutoff and update conditions
Maintain announced plan, current plan, matched actual and unmatched departures as separate dated series. A cancellation reduces an expectation; it does not reinstate a person already counted as having left without evidence of re-employment. A postponement changes timing, not necessarily the eventual population.
Update when a linked supplement, cancellation, revised date or matched actual record appears. Absence from a searched WARN list does not prove no layoffs, because scope, timing and publication can differ. The final conclusion should identify the latest documented plan and the realised portion actually supported, without turning a notice database into a legal finding or an automatic payroll-savings forecast.
Separate another event from another publication
Assume a genuinely separate notice at another documented site adds fifty planned exits, with no overlap in people or event scope. Current combined announced population is then 200 + 50 = 250. It is not 300 + 200 + 50 = 550, because the first two documents are versions of the same original event.
If a public download still displays the original three hundred, retain its extraction date and determine whether it is a historical notice table rather than a current-plan table. A preserved original record is not itself evidence that the cancellation failed. Document version and event state answer different questions.
FAQ
Should the original 300 and postponed 200 be added?
No. In the example the later document revises the same event.
Can the reported eighty departures be subtracted immediately?
Only after evidence matches them to the revised planned population and definition.