Two weeks per year of service. It’s the number that gets read aloud in a conference room with a box of tissues on the table, and it’s the number most people nod at because it sounds like a lot. Ten years in, that’s twenty weeks of pay — five months. Plenty of runway, in theory, to land something new.
Then the COBRA paperwork shows up.
For a family that was paying a modest payroll deduction for health coverage while employed, the same plan without an employer subsidizing most of the premium can run well over a thousand dollars a month, sometimes close to double that depending on the plan tier and the state. The severance check was built around replacing a paycheck. It was never built around replacing a paycheck *and* the chunk of the health premium the employer had been quietly absorbing all along.
A Formula Built for a Different Economy
The “weeks per year of service” formula has been the default in American layoffs for decades, and it has barely changed even as the thing it’s supposed to cushion — the cost of staying insured between jobs — has changed enormously. The formula was largely set when health coverage was cheaper, job tenure was longer, and the gap between employer-subsidized premiums and unsubsidized ones was narrower. None of those conditions reliably hold anymore.
Severance policies tend to get written once, by HR and legal teams thinking about payroll liability and legal risk, and then left alone for years. They’re rarely revisited with health insurance costs in mind, because health insurance costs aren’t really what severance was designed to solve. It was designed to buy goodwill, reduce the odds of a wrongful termination claim, and give someone a cushion while they look for work. Covering the actual cost of staying insured was never the explicit design goal — it just gets treated as if it is, because that’s the biggest gap most people feel.
What COBRA Actually Costs
COBRA lets a laid-off employee keep the exact same health plan they had, for up to 18 months in most cases, but at the full premium — the share the employer paid plus the share the employee paid, plus often a small administrative fee on top. For an employer-sponsored family plan, the full premium is frequently the biggest line item an employee never sees directly, because it mostly shows up on a pay stub as a modest deduction while the company quietly covers the rest.
Once that subsidy disappears, the sticker shock tends to be severe. A plan that cost a few hundred dollars a month out of a paycheck can suddenly cost well over a thousand for the same coverage, with no change in the plan itself — same deductible, same network, same card in the wallet. The severance formula didn’t anticipate that jump, because it was never calculated against it in the first place.
The Gap, in Practice
Twenty weeks of severance pay sounds like five months of security. Subtract a COBRA premium that’s suddenly consuming a meaningful share of that money every month, plus whatever portion goes to rent, a mortgage, groceries, and the ordinary cost of living, and the runway shortens considerably faster than the headline number suggests. Add a job search that takes longer than hoped — which, in many fields, has become common rather than exceptional — and the math gets uncomfortable well before a new offer arrives.
This is part of why unemployment can feel financially worse than the severance number implied it would be, even for people who received what looked like a generous package. The formula measured income replacement. It didn’t measure the cost of staying insured while that income is being replaced.
Where There’s Room to Push
Severance packages are often presented as fixed, especially in group layoffs where everyone gets the same formula applied to their tenure. But the package isn’t always as non-negotiable as the paperwork implies, particularly for individual departures, senior roles, or situations involving any ambiguity about the reason for termination.
A few things tend to be worth asking about, even when the base formula seems set:
Whether the company will cover COBRA premiums directly for a set period, rather than just increasing the severance check and leaving the employee to pay COBRA out of it. Some employers will do this because it’s tax-advantaged for both sides compared to equivalent cash.
Whether the severance calculation can be adjusted upward specifically to account for health insurance costs, framed not as asking for more money generally but as closing a documented gap.
Whether outplacement services, which some packages include, can be converted into cash or extended benefits instead, if the job search support isn’t something that’s actually going to be used.
None of this is guaranteed to work, and companies vary enormously in how much flexibility they’re willing to show, especially in packages tied to a mass layoff rather than an individual negotiation. But it’s rarely explored, because severance paperwork is usually presented all at once, during an emotionally difficult meeting, with a short window to sign.
The Marketplace Alternative
COBRA isn’t the only option, and for many people it isn’t the cheapest one. Losing employer coverage counts as a qualifying life event for marketplace health insurance, which opens a special enrollment window outside the usual annual period. Depending on income during the months of unemployment, marketplace subsidies can make a plan meaningfully cheaper than COBRA, even though the coverage itself may look different — different network, different deductible structure, sometimes a different insurer entirely.
The tradeoff is continuity. COBRA keeps the same doctors, same prescriptions, same plan rules, which matters for anyone mid-treatment for something ongoing. A marketplace plan might save money but require switching providers or rethinking a treatment plan, which isn’t always a fair trade depending on the circumstances.
What the Formula Doesn’t Say
None of this is a reason to assume severance is calculated in bad faith. Most HR teams apply the same formula to everyone precisely to avoid the appearance of favoritism, and the formula itself is usually a known quantity set well in advance of any particular layoff. The problem isn’t malice. It’s that the formula answers a question — how many weeks of pay does this person get — that isn’t the same question as how much it will actually cost this person to stay insured while they look for the next job.
That distinction is worth sitting with before signing anything, and worth raising out loud, even in a conversation that everyone in the room would rather have end quickly.












