A staggered layoff is the most common way employers try to stay under the WARN Act's notice obligation. The pattern: at a site with several hundred active employees, cut 40 people in October, 40 in November, 40 in December. Each round looks small. Each round individually falls under the 50-worker mass-layoff floor at 29 U.S.C. § 2102(a)(2), and at a site that size each round also falls under the alternative 33-percent test that applies for mass layoffs of 50 to 499 employees. No notice gets sent. No state dislocated-worker unit gets called.

Congress addressed this exact pattern when it wrote WARN. The rule lives in 29 U.S.C. § 2102(d), the 90-day aggregation provision (Cornell LII, verified May 2026). The rule operates retroactively: the third-round cut can sweep in the first and second rounds and turn three sub-threshold events into a single covered mass layoff with back-pay liability for everyone in all three rounds.

What the statute actually says

29 U.S.C. § 2102(d) reads: "For purposes of this section, in determining whether a plant closing or mass layoff has occurred or will occur, employment losses for 2 or more groups at a single site of employment, each of which is less than the minimum number of employees specified in subparagraphs (B) or (C) of subsection (a)(2) of this section but which in the aggregate exceed that minimum number, and which occur within any 90-day period shall be considered to be a plant closing or mass layoff unless the employer demonstrates that the employment losses are the result of separate and distinct actions and causes and are not an attempt by the employer to evade the requirements of this chapter" (Cornell LII, verified May 2026).

Three points to pull out of that paragraph. The aggregation window is 90 days. The trigger threshold is the same 50-worker (or 33-percent) plant-closing or mass-layoff floor that applies to a single round. The default rule is aggregation; the employer carries the burden of proving the rounds were unrelated.

How the 90-day window actually runs

The window is a rolling lookback. It is not a fixed calendar quarter. A November 1 termination triggers an aggregation review that pulls in every employment loss at the same site within any 90-day period ending on November 1, reaching back to August 3. A December 15 termination pulls in every loss since September 16. Each individual termination has its own rolling lookback under 20 CFR § 639.5 (verified May 2026).

The math changes depending on which date the analysis anchors on. A first-round worker laid off on August 1 will not retroactively gain a WARN claim from a round terminated on November 15, which sits outside the 90-day frame. A first-round worker laid off on August 15 will be inside the lookback when the third round lands on November 13, exactly 90 days later.

The Department of Labor regulation at 20 CFR § 639.5(a)(1) confirms the rolling-window construction: aggregation applies "within any 90-day period," and the employer cannot select which 90-day frame to use (verified May 2026).

The "separate and distinct actions and causes" defense

Aggregation is the default outcome. The employer can defeat it only by proving the rounds were unrelated. The regulation at 20 CFR § 639.5(a)(1)(ii) puts the burden on the employer: "The employer has the burden of proof to show the cause of the employment losses was separate and distinct."

What helps the employer: documentation that round 1 was caused by a discrete event (a customer contract cancellation, a regulatory shutdown, a natural disaster damaging a specific facility) and round 2 was caused by a different discrete event (a software-business pivot, an acquisition, a hiring freeze imposed by a new owner). The cleaner the causal separation, the stronger the defense.

What hurts the employer: documentation that all rounds came from a single board approval, a single program label, a single budget line item, or a single executive sponsor. When the company calls all three rounds "Project Renewal" in internal memos, the case for separate causes weakens. When the same executive signs off on each round under the same restructuring plan, the case weakens further. When the severance agreement template is identical across rounds, that is more evidence of a unified program.

Courts apply this test as a multi-factor analysis. The Sixth Circuit's decision in Morton v. Vanderbilt University, No. 15-5417 (6th Cir. Jan. 5, 2016) (opinion), is on-point appellate treatment of the 90-day timing question. Morton held that for aggregation purposes employment terminates when wages and benefits cease, not when notice issues, so a second group whose pay continued past the 90-day mark fell outside the aggregation window. The decision sits inside the regulatory framework at 20 CFR § 639.5(a)(1)(ii), which puts the burden of demonstrating separate-and-distinct causes squarely on the employer.

How retroactive back-pay works

When round 3 pushes the aggregated total over the threshold, the rounds that came before round 3 also become part of a covered mass layoff. Workers terminated in rounds 1 and 2 within the 90-day window also have a WARN claim, even though their individual round looked sub-threshold at the time.

The remedy under 29 U.S.C. § 2104 is back pay for each day of the violation period, up to 60 days, plus the value of benefits the employer would have provided (verified May 2026). For a worker who left in round 1 thinking WARN did not apply, the discovery that round 3 swept them into coverage can mean a meaningful claim.

The statute of limitations question matters for these retroactive claims. WARN does not contain its own limitations period. Federal courts borrow the most analogous state limitations period, which commonly ranges from two to six years depending on the state and the analogous statutory analog the circuit selects (wage-claim, contract, personal-injury). Circuits also split on the accrual question, with some running the clock from the date of the qualifying employment loss and others from the date the aggregated total crossed the threshold. The practical implication for a multi-round program is that the clock for an aggregation claim usually runs no earlier than the round that crossed the threshold, but the exact deadline turns on the state and circuit involved and should be calendared with employment counsel.

Recognizing the pattern from the inside

The signal pattern that points to a continuing program rather than separate events:

Vague employer responses to direct questions about quarterly headcount changes. Answers like "we don't disclose specific numbers" or "the cuts were across multiple business units" are consistent with an employer trying to obscure the running total.
Repeated use of a single program label across rounds. "Phase 1 of Project Renewal," "Phase 2 of Project Renewal," and "next wave of the Project Renewal initiative" all point to a single program, which cuts against the separate-causes defense.
A single executive sponsoring all rounds. Restructurings sponsored by one VP or one board committee tend to be a single causal action. Genuinely unrelated rounds usually come from different decision-makers responding to different events.
A single severance schedule template across rounds. Variations across rounds (different severance multipliers, different release-of-claims terms) suggest different programs. Identical severance terms suggest a single program.
Cuts spaced just outside notable thresholds. At a large site, three rounds of 49 workers each in 88 days is consistent with engineering, not coincidence; the count sits one short of the 50-worker floor in each round and the spacing sits two days outside the 90-day window.

What discovery uncovers in WARN litigation

When a WARN aggregation claim goes to litigation, the discovery requests track the separate-causes test directly. Plaintiffs' counsel typically requests:

Board minutes for the relevant period, showing whether one approval covered all rounds.
HR personnel files for everyone terminated, showing the program label assigned to each separation.
Severance-payment ledgers, showing whether the company paid out on a single budget line or separate ones.
Internal communications labeling the rounds, including project-named emails, restructuring slide decks, and Slack channels.
Compensation analyst spreadsheets used to score employees for selection.

A defense based on separate causes that survives this discovery is rare. A defense that does not survive discovery turns into back-pay liability for everyone in all rounds within the 90-day window.

State mini-WARN aggregation: a layered floor

Federal § 2102(d) is the floor. Several states impose stricter aggregation rules, and the state rules apply in addition to the federal test, not instead of it.

Jurisdiction	Aggregation window	Causation defense	Notes
Federal WARN	90 days	Employer proves separate and distinct causes	50-worker floor; 33-percent test for sites under 500
New York (NY-WARN)	90 days	Stricter showing required for the defense	Covers employers with 50+ full-time employees; mass-layoff trigger is 25 workers if they comprise at least one-third of the active workforce, or 250+ workers regardless of percentage; 90-day notice
New Jersey (Millville Dallas Airmotive Plant Job Loss Notification Act, N.J.S.A. § 34:21-1 et seq.)	90 days	Defense applies but must be documented contemporaneously	Severance of one week's pay per year of service attaches when aggregation triggers and the employer did not give the full 90-day notice; full and timely notice can substitute for the severance obligation
California (Cal-WARN, Cal. Lab. Code § 1400 et seq.)	30-day window built into the mass-layoff definition itself	Anti-evasion principle applies but no codified 90-day federal-style hook	Lower employer-coverage threshold (75 employees) and the 30-day mass-layoff window mean closely-spaced rounds reach the trigger at sites federal WARN would not capture
Illinois (IL-WARN)	90 days	Defense available; mirrors federal test	75-employee-employer threshold with 25-worker mass-layoff trigger

A worker laid off in a state with a mini-WARN should run the math under both the federal § 2102(d) test and the state aggregation rule. A round that fails the federal aggregation test may still trigger a state claim, especially in California, New York, and New Jersey where the floors are lower. The layoff calculator carries each state's aggregation window and causation standard in its data layer.

The interaction with comparable-role offers

A common employer move in a multi-round restructuring is to offer some affected workers a comparable role instead of a severance package. Under 20 CFR § 639.3(f)(1), an employee who accepts a comparable role and continues working for the same employer without a break in employment is generally not considered to have suffered an "employment loss" under WARN (verified May 2026).

A comparable-role acceptance reduces the employment-loss count for that worker. It does not affect the count for workers in earlier rounds who were terminated outright. A worker who accepts a comparable role in round 3 may still be covered by a class WARN claim that captures the rounds 1 and 2 employees who left without a comparable offer. The reverse is also true: a round 1 worker who was laid off outright is not affected by a round 3 colleague who accepted a transfer.

The companion piece on comparable-role offers in restructurings walks through the comparability test in detail.

When to call an employment lawyer

A WARN aggregation claim is fact-intensive. The signals that the math may favor the worker:

Multiple rounds of cuts at the same site within a 90-day window, where each round individually was below 50 workers.
An employer that refuses to confirm the running headcount or program label.
A single executive sponsor or single board approval covering all rounds.
A standardized severance template applied across rounds.
Cuts spaced at intervals that look engineered to fall just below the threshold.

When several of these signals are present, the math may produce a coverable claim and the employer's separate-causes defense may not survive discovery. An employment lawyer can pull the public filings (state dislocated-worker unit notices, SEC restructuring disclosures for public companies) and the discovery record to test the aggregation theory.

For the underlying WARN basics, see the 60-day WARN notice piece. For the parallel question of how multi-round severance packages interact, the multi-round layoff hub walks through the negotiation dynamics, OWBPA decisional units, and ERISA plan rights. For the OWBPA disclosure list that often accompanies a covered round and the older-worker selection-rate math it enables, see how to read your OWBPA disclosure list. For the accept-vs-decline framing if your employer offers you a voluntary exit ahead of the involuntary round, see voluntary separation package vs. involuntary RIF. The layoff calculator checks your specific facts against federal and state WARN thresholds. Methodology lives at /methodology and the editorial standards are documented at /about.

FAQ

How does the 90-day window apply to my termination date?

Each individual termination date triggers its own 90-day lookback. A worker terminated on November 1 has a window running from August 3 to November 1. Every employment loss at the same site of employment within that window counts toward the aggregated total. The DOL regulation at 20 CFR § 639.5(a)(1) treats the window as rolling rather than fixed (verified May 2026), so a worker terminated later in a staggered layoff has a longer rear-view that captures earlier rounds.

What evidence does the employer need to win the separate-causes defense?

The employer carries the burden under 20 CFR § 639.5(a)(1)(ii) and must show that each round resulted from a discrete, unrelated business event (verified May 2026). Documentation that helps the employer: distinct board approvals for each round, different program labels in internal memos, different budget allocations, and different executive sponsors. Documentation that hurts the employer: a single program name, one budget line, a single executive sponsor, and a uniform severance template. The Sixth Circuit in Morton v. Vanderbilt University, No. 15-5417 (6th Cir. Jan. 5, 2016) (opinion), applied the timing question as a fact-intensive analysis where the precise dates of wage cessation, not notice dates, control whether two rounds fall inside the 90-day window.

Can a round 1 employee bring a WARN claim months after the round 1 termination?

Yes, when round 3 pulls round 1 inside an aggregated window. The triggering event for the aggregation claim is round 3, not round 1. WARN borrows analogous state limitations periods (commonly two to six years), with circuit splits on whether the clock runs from the qualifying employment loss or from the date the aggregated total crossed the threshold; under either approach, a round 1 worker terminated in August can typically still file when round 3 in November pushes the aggregated total over 50 workers, but confirm the applicable period with employment counsel for your circuit. The 29 U.S.C. § 2104 remedy of up to 60 days back pay plus benefits applies to every worker in every round inside the 90-day window, not just the round that crossed the threshold.

Does the aggregation rule apply across different sites of employment?

No. 29 U.S.C. § 2102(d) aggregates only employment losses "at a single site of employment." Layoffs at different sites are treated separately under the WARN coverage analysis (verified May 2026). The "single site of employment" definition under 20 CFR § 639.3(i) covers a single building, a campus, or a group of buildings under common operational control. A truly mobile workforce (drivers, traveling sales staff, remote engineers) is treated as employed at the site they are assigned to or report to. Different states may apply their own site-of-employment rules under state mini-WARN statutes.

What about state mini-WARN laws when federal aggregation does not trigger?

State mini-WARN statutes apply in addition to federal WARN, not instead of it. A round of layoffs that falls below the federal 50-worker floor may still trigger a state aggregation rule with a lower threshold. New York, New Jersey, California, and Illinois all impose lower employee-count thresholds and may aggregate rounds the federal statute would not capture. The state-specific overlays are described on the calculator's per-state pages and tie into the same 90-day rolling-window logic where applicable.

WARN's 90-Day Aggregation Rule: When Round 3 Triggers Back-Pay for Rounds 1 and 2