California Layoffs: What the 2025 WARN Data Shows

California recorded 1,691 layoff events affecting 97,751 workers across 673 companies in 2025, according to layoff records compiled from state WARN Act filings. That makes it the single largest source of WARN-documented layoff activity in the country for the year — more reported events than any other state, and close to a quarter of every affected worker captured in the national dataset. This piece breaks that activity down by place and by month, and is equally clear about what the WARN record can and can't show.

The figures here come from LayoffTracker.us, which assembles layoff records exclusively from notices that employers file under the federal and state Worker Adjustment and Retraining Notification (WARN) Act. They cover the 2025 calendar year as of the dataset's snapshot dated August 5, 2026. Two terms below do a lot of work and are not interchangeable: an event is a single WARN filing, and the affected-worker count is the headcount an employer attached to that filing. A month heavy with events is not necessarily a month heavy with affected workers — and California's 2025 record is a clean example of the two pulling apart.

California in national context

Across the country, the dataset records 5,105 WARN events affecting 410,707 workers at 2,520 companies in 2025. California alone accounts for about 33.1% of the national events, 23.8% of the affected workers, and 26.7% of the companies — a larger share of national layoff activity than any other single state, on all three measures. By affected-worker count, the next-largest states trail well behind: Washington (30,707 affected across 120 events), Texas (24,367 across 300), New York (21,461 across 623), and Florida (20,423 across 322), with Georgia, Illinois, and Pennsylvania following in the 12,000–16,000 range. California's lead is a matter of scale, not a close race — its 97,751 affected workers exceed the next three states combined. Worth noting in passing how differently states' totals are composed: New York reaches its total across 623 events while Washington reaches a comparable one across just 120, a reminder that an event count and a headcount describe different things even at the state level.

One caveat belongs up front, because it shapes every national comparison in this article. National totals reflect only the states that publish WARN data — 47 states appear in the 2025 breakdown. Arkansas, New Hampshire, and Wyoming do not publish WARN notices publicly at all, so they contribute nothing to the national figure. Oklahoma files events but does not include affected-employee counts in its filings, so its headcount reads as unreported rather than zero. The practical effect is that the national denominator is a floor, not a full census of U.S. layoff activity. California's share is best read as its share of what was reported — directionally large, but not a precise slice of a complete national picture.

Where the layoffs landed

Within California, the activity is geographically lopsided. Los Angeles is the clear leader, with 144 events and 24,040 affected workers — on its own, roughly a quarter of the state's entire affected-worker total. After Los Angeles, the largest concentrations recorded are San Francisco (61 events, 5,257 affected), San Diego (72 events, 3,160), Sacramento (34 events, 2,692), Santa Clara (54 events, 2,213), and San Jose (48 events, 2,147), followed by a cluster of Bay Area and Central Valley cities — Sunnyvale, Pleasanton, Bakersfield, West Sacramento — each in the four-figure range for affected workers. The Los Angeles basin and the greater Bay Area together carry most of the state's WARN-documented headcount.

Below that top tier, the activity fans out across the state's other population centers: the Central Valley (Bakersfield, West Sacramento, and Folsom each register four-figure affected-worker counts alongside Sacramento), the Inland Empire (San Bernardino and Ontario), and Orange County (Irvine, Anaheim, and Long Beach). None of these individually rivals Los Angeles, but together they show that WARN activity in 2025 was distributed across the state's inland regions as well as its coastal metros, not confined to any one part of California.

Treat that city ordering as directional rather than exact. City names in this data are raw entries lifted from individual WARN filings, and they carry the noise that implies: typos, inconsistent casing, address fragments, and placeholder values. The dataset spreads across 325 distinct city labels for California in 2025, and the same physical city can appear under more than one spelling, which slightly undercounts wherever variants split. The clearest sign to keep in mind: a placeholder label of "Unknown" itself ranks among the larger city entries, with 34 events and about 1,100 affected workers — filings whose location simply wasn't pinned to a city. Individual rows can be lumpy in the same way: La Canada Flintridge appears with roughly 871 affected workers from just two events at two companies — a real figure, but one resting on so few filings that a single large notice swings the city's whole standing. The broad shape (Los Angeles first, a heavy Bay Area presence) is reliable; the exact rung-by-rung ranking past the top few, and any small city's placement, is not something to lean on.

How the year unfolded

Spread across the twelve months, California's event count stayed in a relatively tight band — from a low of 80 events in August to highs of 215 in June and 210 in March. Affected-worker counts swung far more widely — from a low of 3,154 in February to July's 22,438, a more than sevenfold range across the year even as the event count never varied by more than about a factor of three. That divergence is where the year's most distinctive feature sits.

Month	Events	Affected workers
Jan	124	9,853
Feb	106	3,154
Mar	210	6,548
Apr	159	6,748
May	127	6,377
Jun	215	8,998
Jul	160	22,438
Aug	80	4,966
Sep	154	7,701
Oct	135	7,160
Nov	100	6,721
Dec	121	7,087

July is the standout. It recorded 22,438 affected workers — more than double any other month — yet did so on 160 events, fewer than June's 215. The arithmetic of that gap is the point: July's filings averaged about 140 affected workers each, against a statewide average near 58 per event and June's roughly 42 per event. In other words, July's spike was driven by a smaller number of comparatively large filings, not by a surge in the number of layoffs. That is a description of the data, not an explanation of it — the WARN record shows the size of the filings, not the business reasons behind them.

Stepping back, the two halves of the year are close on events (941 in the first half, 750 in the second) but the affected-worker total is heavier in the back half (41,678 versus 56,073), almost entirely because of July. One technical note worth carrying into any month-by-month reading: each event is assigned to a month based on the date recorded in its WARN filing, and some filings record dates that are approximate rather than exact. The monthly contours are dependable; reading too much into a single month's precise figure is not.

What these numbers do and don't tell you

A few boundaries keep this snapshot honest:

• Events and headcount are separate facts. The count of filings and the count of affected workers move independently, as July makes plain. Cite whichever one the claim is actually about, and don't convert between them.
• This is WARN activity, not all layoffs. The WARN Act only requires notice for layoffs above certain size and duration thresholds. Smaller reductions, attrition, contractor and gig cutbacks, and any layoff an employer failed to file simply don't appear here. The dataset is a consistent measure of WARN-reported layoffs, which is a floor on real-world job loss, not a ceiling.
• This is a geographic and temporal view, not an industry one. Industry classification across the dataset is still being built out and is not yet complete, so this article deliberately makes no industry-level claims. Where the layoffs happened and when is well supported; what sectors they hit is not something the current data can segment reliably.
• City detail is directional. As above, lean on the broad geographic pattern, not the exact ordering.
• The data shows what was filed, not why. Nothing here supports a causal story about the California economy. These are counts of notices and the headcounts attached to them.

Methodology

Layoff data in this article is drawn from LayoffTracker.us, which compiles records from publicly filed state WARN Act notices. The dataset is limited to events that meet WARN thresholds and excludes states without public WARN disclosure (Arkansas, New Hampshire, and Wyoming); Oklahoma reports events without affected-employee counts, so its headcount is treated as unreported rather than zero. National and cross-state figures therefore reflect only the states that publish, and should be read as a floor on total U.S. activity. Figures reflect the 2025 calendar year as of the data snapshot dated August 5, 2026. For full methodology, see https://www.layofftracker.us/methodology.