Why WARN Numbers and News Headlines Don't Match

A company announces 1,000 job cuts in the news. The official layoff tracker shows a different number. Neither figure is necessarily wrong. They are usually measuring different things, at different times, under different rules.

Understanding that gap matters, because the WARN data and the headlines answer different questions. Headlines tell you what a company said. A WARN filing tells you what an employer formally notified a state agency about, for a specific site, on a specific date.

The numbers right now

Layoff Atlas currently shows 679 WARN notices covering 65,527 reported workers in its 90-day notice window, with 33 counts unavailable. Washington leads with 17,073 reported workers, followed by California with 14,001, Illinois with 4,626, Texas with 3,652, and Ohio with 3,180. The upcoming effective weeks are heavy: 5,503 workers in the week of October 26, 5,061 in the week of November 2, and 6,568 in the week of November 9.

Those totals are a snapshot of filed notices, not a count of everyone who lost a job in America. That distinction explains most of the mismatch with the news.

Reason 1: Three different dates

Every layoff story has at least three dates, and each source tends to pick a different one:

• The announcement date, when a company tells the press or its staff.
• The notice date, when the employer files the WARN notice with the state.
• The effective date, when the job losses actually take effect.

Take a recent filing: Starbucks filed a notice on October 23, 2026, for 135 workers in Seatac, Washington, effective December 31, 2026. A news story in October, a WARN count in October, and a jobs-lost count in December would all describe the same event differently. Foster's 169-worker Winston-Salem filing, noticed October 21, is not effective until March 5, 2027.

Reason 2: Not every layoff requires a WARN notice

The federal WARN Act only covers larger employers and larger layoffs. In general, it applies to employers with 100 or more full-time workers, and to plant closings affecting 50 or more workers or mass layoffs of 50 to 499 workers where at least 33 percent of the workforce is affected. Smaller cuts, and cuts at smaller employers, simply never produce a WARN filing.

That means a headline total that adds up every announced cut — including a 20-person office closure or a 40-person team reduction — will almost always run higher than the WARN total for the same company. The law never required a filing for those jobs.

Reason 3: State rules and coverage gaps differ

WARN coverage is not uniform across the country. Some states have their own mini-WARN laws with lower thresholds or broader coverage, so their data is richer. Other states rely on the federal floor alone. And Layoff Atlas itself notes that its registry covers 51 jurisdictions, but only 48 have imported records — a coverage gap is not the same as zero layoffs in the missing places.

A national headline total, by contrast, does not respect state lines or filing rules. It counts what companies announce, wherever they announce it.

Reason 4: Headlines count announcements; filings count sites

A company announcing "1,000 cuts nationwide" may file a dozen separate WARN notices, one per site, spread across several states and several effective dates. Some of those sites may fall below the WARN threshold and never appear. Others may be filed weeks after the announcement, or in phases. A tracker that organizes data by site and effective date will therefore show a different shape — and often a different total — than a single headline number.

Headlines can also sweep in categories a WARN filing never touches: overseas roles, contractors, voluntary departures, or restructuring phases not yet filed. A WARN filing is tied to one location and one worker count for that site.

Reason 5: Some counts are simply unavailable

Even within the WARN system, not every record carries a worker count. In the current 90-day window, 33 counts are unavailable, and those workers are excluded from the totals rather than estimated. News reports rarely leave a blank — they round, estimate, or repeat a company figure.

Undated and recently imported records add another wrinkle. Some filings arrive without a published notice date and sit in a separate "recently imported" group. They are real records, but they do not enter the 90-day notice window in the same way.

How to read the two numbers together

The practical approach is to use each source for what it is good at:

• Use headlines to learn that something was announced, and roughly how big a company says it is.
• Use WARN filings to confirm the details: which sites, how many workers per site, when the notice was filed, and when the cuts take effect.
• When the two disagree, check the dates first. Most gaps close once you line up announcement, notice, and effective dates.
• Treat any total as a floor, not a census. WARN totals exclude small layoffs, uncovered employers, and unavailable counts — and these are reported plans, not confirmed job losses.

The next time a headline number and a filing number disagree, the disagreement itself is information. It usually tells you something about timing, thresholds, or coverage — not that one side invented its figure.

Data source: Layoff Atlas