Recoverable vs. Non-Recoverable Penalties Under PAGA: What the Statute Actually Authorizes

PAGA is a procedural vehicle, not a source of substantive rights. It authorizes an aggrieved employee to recover, on the state’s behalf, civil penalties that the Labor Commissioner could have recovered, and it supplies a default penalty under section 2699(f) for Labor Code provisions that carry no civil penalty of their own. It does not convert every monetary remedy in the Labor Code into a PAGA recovery.

That limitation is the whole game. ZB, N.A. v. Superior Court (2019) 8 Cal.5th 175 is the clearest statement of it: the Court held that the unpaid-wages component available under section 558 is not a civil penalty a PAGA plaintiff may recover. Wages are recovered by employees in their own right, through their own claims, on their own limitations periods — not through a representative penalty action.

Kirby v. Immoos Fire Protection (2012) 53 Cal.4th 1244 supplies the other half. The meal and rest period remedy under section 226.7 is compensation to the employee, not a penalty imposed on the employer. It follows that the premium itself is not PAGA-recoverable, even though the underlying failure to provide the break is a Labor Code violation for which the default penalty is available.

Stone v. Alameda Health System (2024) 16 Cal.5th 1040 added a question that precedes all of this. Section 2699(b) gives “person” the meaning it carries in Labor Code section 18, and it does so “for purposes of this part” — the whole of PAGA, not merely the default penalty. A public entity is not a person on that definition, so it is outside PAGA entirely: neither the section 2699(f) default penalty nor a penalty specifically provided elsewhere in the Labor Code reaches it. The Court of Appeal had split those two categories and allowed the second; the Supreme Court rejected the distinction and disapproved Sargent v. Board of Trustees of California State University (2021) 61 Cal.App.5th 658 as to nondefault penalties.

The affirmative side of the ledger belongs in the same analysis, because it is what remains after the subtractions. Section 2699(k)(1) authorizes an aggrieved employee to recover the section 2699(f) default penalty and to be awarded injunctive relief, and provides that an employee who prevails “shall be entitled to an award of reasonable attorney’s fees and costs,” including the section 2699.3 filing fee. It also preserves other remedies expressly: nothing in PAGA limits an employee’s right to pursue remedies available under other state or federal law, separately or concurrently. A category analysis that strips the penalty column without accounting for fee exposure and for the claims that survive outside PAGA has not valued the case.

The practical order of operations therefore runs: is this defendant within PAGA at all, is this requirement one PAGA reaches, has the agency already acted on the same facts, then is there a specific civil penalty, then is it one an aggrieved employee may recover, then does the default penalty apply. The first question is categorical rather than incremental — it zeroes the entire penalty column rather than trimming it — and it reaches past counties and cities to hospital authorities, special districts, and joint powers authorities. It also has a mirror image worth pricing: where a public entity is the client rather than the employer, both the wage order exclusion and the section 220(b) exemption turn on employees “directly employed by” the entity, so a staffing agency’s workers are not covered by either and the exposure consolidates on the agency rather than dividing across the two.

Three provisions remove categories from the case wholesale, and each operates before any question about wages versus penalties is reached.

The first is the one most often overlooked. Section 2699(k)(2) provides that no action shall be brought under PAGA “for any violation of a posting, notice, agency reporting, or filing requirement of this code, except if the filing or reporting requirement involves mandatory payroll or workplace injury reporting.” Posting and notice items are routine filler in a PAGA notice — the required wage order posting, the various notices that must be given or displayed — and they are attractive to plead because they are easy to prove and apply to the entire workforce. The statute does not reach them. The exception is narrow and worth reading precisely: it restores filing and reporting requirements only where the requirement involves mandatory payroll or workplace injury reporting, and it does not restore posting or notice requirements at all.

The second is section 2699(l). No action lies where the agency, on the same facts and theories, has cited the employer within the section 2699.3 timeframes for a violation of the same section, or has initiated a proceeding under section 98.3. Where a Labor Commissioner citation already covers the conduct, the corresponding category is not reduced — it is unavailable.

The third is arithmetic rather than categorical, and it is routinely missed in modeling. Section 2699(o) provides that the penalty recovered under PAGA “shall be reduced by one-half if the employees’ regular pay period is weekly rather than biweekly or semimonthly.” A weekly payroll produces roughly twice as many pay periods as a biweekly one, and the per-pay-period penalties scale with that count. The halving therefore functions closer to a neutralizer than a discount — but a model that counts weekly pay periods without applying it overstates the penalty column by a factor of two, and that error compounds through every derivative that scales the same way.

Meal and rest periods. The section 226.7 premium is a wage and is not recoverable through PAGA. The violation itself supports the default penalty under section 2699(f)(2). Demands that list the premium as a PAGA penalty are double-counting, and the amount involved is frequently the largest single line in the demand.

Overtime and minimum wage. Unpaid overtime is a wage. The default penalty attaches to the violation; section 1197.1 supplies a specific penalty structure for minimum wage violations. The wage differential itself belongs to the employees’ own claims.

Wage statements. Section 226(e) provides its own statutory penalty, subject to a per-employee cap. This is one of the categories where the statute genuinely supplies what the plaintiff is asking for — but the scienter element and the cap both matter, and both are routinely ignored in demands.

Waiting time. Section 203 is a penalty rather than a wage, but its recoverability through PAGA depends on how the claim is characterized, and it applies only to employees who separated within the relevant period. A demand that applies section 203 to the entire workforce rather than to separated employees is not a close question.

Expense reimbursement. Section 2802 obligations are reimbursement of the employee. The default penalty attaches to the violation. The reimbursement column is larger than the receipts, though, in a direction the penalty analysis does not capture: section 2802(b) carries interest at the judgment rate accruing from the date each expenditure was incurred rather than from a later payment date, and section 2802(c) writes the employee’s attorney’s fees into the definition of the recoverable amount itself rather than leaving them to a separate fee statute.

The discipline is to run all three questions for every category rather than reasoning by analogy from the ones that are familiar: is there a specific civil penalty statute; if so, is it one PAGA authorizes an aggrieved employee to recover; and if not, does the default penalty apply.

The reduction is large because the non-recoverable categories are usually the ones carrying the biggest numbers. Premium pay and unpaid wages scale with hours and rates; penalties scale with headcount and pay periods. A demand that adds both together, and then applies penalties to the wage components as well, can easily double the figure it should have produced.

The practical output is a category-level table showing, for each alleged violation: the amount demanded, the amount actually authorized, the statutory basis for the difference, and the resulting reduction. Presented that way the analysis is difficult to argue with, because each line is a legal proposition rather than a negotiating position.

It also has a second use. The same table tells you which categories are worth litigating. There is little value in spending discovery on a category whose maximum authorized penalty is immaterial once the wage component is stripped out.

One qualification belongs with any figure derived from the reform’s caps. Section 2699(e)(2) gives the court discretion in both directions: it may award less than the maximum specified by the statute, “including the penalty amounts in subdivisions (g) and (h),” and it may also, notwithstanding those limitations, exceed them — in either case where confining the award otherwise “would result in an award that is unjust, arbitrary and oppressive, or confiscatory.” The fifteen and thirty percent figures are therefore the statutory measure rather than an absolute ceiling, and a model that treats a qualified cap as a guaranteed outcome is making the same category of error as a demand that treats a maximum as an entitlement.

The first published decision to apply that subdivision after the reform put the point sharply. In Taduran v. James R. Glidewell, Dental Ceramics (Cal. Ct.App., June 17, 2026, G064718, as mod. July 1, 2026), the court quoted section 2699(e)(2) in its amended form — cross-references to subdivisions (g) and (h) included — and held that it prescribes no method at all: “[a]fter calculating the maximum civil penalty on a per pay period basis, the trial court is not precluded from using any reasonable method to reduce that amount, including applying a reduction on a percentage, per pay period or per employee basis.” The scope of that authority is reviewed de novo and its exercise for abuse of discretion. On the facts the court affirmed a reduction leaving the plaintiff less than one percent of what he sought. The decision does not construe the cap standards themselves, and the conduct predated the reform, so it is not authority on what “all reasonable steps” requires. What it settles is narrower and still useful: the discretion that qualifies the caps has no formula, which cuts against a demand insisting the court must discount per pay period and equally against a defense model that assumes any particular method will be chosen.

The most common response is to recharacterize: to plead the same conduct under a different section that does carry a recoverable penalty, or to argue that the wage component is recoverable as an element of the penalty rather than as wages. ZB forecloses the second move for section 558, and the first is answered by asking what the recharacterized section actually requires — the elements rarely fit conduct that was pleaded as something else.

The second response is to concede the point and pivot to volume: accepting the category analysis while asserting violation rates that restore the original number. That is a different argument, and it is met with the data rather than with the statute.

Both responses are improvements over the original demand, which is the objective. The category analysis is rarely the end of the negotiation, but it reliably moves the starting point.