Average Severance Package in 2026: What the Data Says, and How Yours Compares

“Was my package normal?” is usually the second question after a layoff (right after “what do I do now”), and it deserves a real answer, not a shrug about how everything varies. The data exists, the formulas behind most employer plans are knowable, and several of the biggest ones are filed publicly. Here’s where the market actually sits in 2026, what moves a package off the average, and what “typical” means for an offer that’s already on your table.

What does the cross-industry data say?

The broadest current benchmark comes from Challenger, Gray & Christmas’ 2025 severance analysis of more than 8,000 packages, which put the average at 19.3 weeks of pay, up sharply from 15.6 weeks a year earlier, as employers competed for clean exits through the 2024–2025 cutting cycles. Averages blend a retiring executive’s 18 months with a warehouse worker’s four weeks, so the number that does the real work in practice is the formula behind most plans:

About two weeks of pay per year of service, often with a base floor, sometimes with a cap. Adjusted up in big tech and parts of finance, down in retail, hourly, and most healthcare roles.

A package near that line is market-typical regardless of how it feels in the moment, and packages can sit above it for structural reasons (group cuts with WARN exposure, long tenure, senior level) or below it for equally structural ones (small employer, no plan on file, sub-two-year tenure).

Apply the norm to a concrete case before reading the sector table, because the formula does more work than the average. A hypothetical eight-year employee at a $130,000 salary ($2,500/week): the two-weeks-per-year norm implies 16 weeks, about $40,000 gross. That sits below the 19.3-week average, yet is squarely typical for that tenure, because the average is pulled up by long-tenure and executive packages. The same person at a 16-weeks-base-plus-two tech employer sees 32 weeks ($80,000) and at a one-week-per-year retail employer sees 8 weeks ($20,000). All three are “normal” for their employer class. That’s the sense in which a benchmark is useful: it locates a package, it doesn’t grade the person holding it.

How do industries actually compare?

The spread by sector is wider than most coverage admits. These figures summarize the employer-specific guides in the layoff recovery hub, each of which carries the sources and the per-company mechanics:

Sector	Typical cash structure	What moves the real value
Big tech (top tier)	16 weeks base + 2 weeks/yr (Google, Meta); ~2-month minimum + ~2 weeks/yr (Microsoft)	Equity: continued vesting vs forfeiture; 6-month COBRA subsidies
Tech (harsher end)	~8–12 weeks at mid-levels (Amazon); 4 weeks + 1/yr capped at 26 (Oracle)	Amazon’s 5/15/40/40 RSU forfeiture; 1-month COBRA at Oracle
Banking	2 weeks/yr with caps (JPMorgan: lesser of 52 weeks or $400K); discretionary by level (Goldman)	Bonus exclusion by default; garden leave; deferred-comp vesting
Healthcare/retail	1–2 weeks/yr with tier caps (CVS files its plan publicly: senior tiers cap at 44 weeks)	License/recertification support; thinner health bridges
Federal	Statutory: 1 week/yr (first 10 years) + 2 weeks/yr after, paid via payroll	VSIP buyouts capped at $25,000; FEHB rules differ entirely

Two patterns are worth naming, and both argue for reading your own sector’s row rather than the blended average, because cross-sector comparisons mislead in predictable directions. First, the cash weeks are the most visible and least decisive number for equity-compensated workers: the gap between continued vesting and cliff forfeiture on a mid-tenure grant routinely exceeds $200,000, several times any plausible difference in severance weeks. Second, health-coverage bridges have hard dollar value: a 6-month subsidized COBRA window is worth roughly $4,800 at average single-coverage rates under the 2026 math, and it’s part of the package whether or not it appeared in the headline number.

What’s legally guaranteed, and what’s plan generosity?

Less is guaranteed than most people assume. No federal law requires severance. The floor is: final wages (state law governs timing and PTO payout), the WARN Act’s 60 days of notice or pay-in-lieu for covered mass layoffs at 100+ employee firms (plus state mini-WARN variants), and, for workers 40 and over, the OWBPA consideration windows under the federal age-discrimination statute: 21 days to consider an individual release, 45 days in a group cut (with required disclosure of affected ages and roles), and 7 days to revoke after signing.

Those windows matter at the recovery stage for one simple and frequently missed reason: they’re a legal minimum clock, and while one is still open, the agreement sitting on your table is still reviewable. If yours is, or if you’re inside the 7-day revocation period, that’s the moment to check the package against the market rather than after. Detailed negotiation strategy is its own discipline (and squarely lawyer territory when real money or unusual claims are involved); the recovery-stage job is narrower: know what the document promises, and verify it arrives.

Why did packages get richer in 2024–2026, and will it hold?

The 15.6 → 19.3-week jump in one year wasn’t generosity; it was mechanics. Three forces pushed packages up through the cycle. Litigation hygiene: bigger group layoffs mean more OWBPA releases, and a release is only worth signing if the consideration attached to it clears the employee’s alternatives; thin packages produce unsigned releases. WARN exposure: pay-in-lieu of the 60-day notice gets baked into headline numbers at covered employers, and several states extended mini-WARN coverage downward. Optics in a documented era: the 2023–2026 layoff cycles were the most publicly tracked in history, and package quality became visible employer branding, the same dynamic that produced the named programs (Project Bora Bora, Project Voyage) covered in the company guides.

None of those forces guarantee the trend continues, which is why the dating discipline matters: a benchmark from 2024 understates today’s market by roughly four weeks of pay, and a 2026 figure will misstate 2027’s. The numbers in this guide carry their dates on purpose. BLS’s duration data is the companion figure worth tracking, because a 19-week package against a 23-week average search is the real arithmetic a household plans around. Both numbers, with their sources, live in the layoff recovery statistics for 2026.

What does “below average” actually mean for a signed package?

Honest answer: for the cash terms, it usually means the employer’s plan was simply thin, not that something went procedurally wrong, and (post-signature, post-revocation) not that there’s a renegotiation to run. The work that is still live after signing:

• Verification. Final paycheck math, PTO payout per your state’s rule, the severance payment schedule, each promised COBRA-subsidy month, and the equity plan’s separation treatment per the grant agreement. (Larger employers’ severance plans are often ERISA documents filed publicly: CVS’s grade-tier plan sits on SEC EDGAR, which is how its caps are checkable at all.) Shortfalls against the written agreement are enforcement questions: document everything first, then consider consulting an employment attorney if the dollars are material; many take wage-claim matters on contingency or flat fee.
• The offsets. A below-average package makes the recoverable money elsewhere matter more: unemployment filed correctly in week one, the right health-insurance call, and a 401(k) handled without the withholding trap.
• The plan. Thin package or rich one, the sequence is the same: the 90-day money plan is built to run on either.

What are the non-cash components actually worth?

Pricing the whole package, not just the weeks, changes plenty of comparisons. Rough valuations for the common components, so nothing gets left as a vague “benefit”:

• Subsidized COBRA months: about $790/month single, $2,290/month family at average group rates. A 6-month family subsidy is worth roughly $13,700 of after-tax money, which is why it appears in the 2026 health-coverage math as a first-class term.
• Equity treatment: continued vesting or acceleration is valued grant-by-grant (unvested value × the fraction covered); for mid-tenure tech workers it routinely exceeds the entire cash line, per the company guides above.
• Outplacement services: typically a $1,500–$10,000 retail value depending on tier: worth using for the resume rewrite and interview practice, worth approximately nothing if it goes unused. It is a service, not money; treat it accordingly in comparisons.
• The paid-notice period: 60 days of WARN pay-in-lieu is ~8.6 weeks of salary that some employers fold into the headline number and others state separately. Normalize before comparing offers across employers.
• A neutral reference / non-disparagement commitment: economically unpriceable but operationally real in a reference-checked search; verify what the agreement actually promises versus what HR said verbally.

Benchmarks are context, not a verdict on you. The 2024–2026 cycles cut deep enough that package quality tracked employer policy and timing far more than individual performance. Worth remembering when the comparison stings, and worth saying plainly in interviews, where “part of a documented restructuring” is a complete answer.