The PAGA penalty is a sovereign claim, prosecuted by the employee as the state's proxy and owned three-to-one by the state. It is assessed per aggrieved employee per pay period under the default provision, section 2699(f), because the meal-and-rest statute fixes no penalty of its own — so its magnitude tracks headcount and the length of the period, not the count of missed breaks.
The 2024 reform left that penalty intact and built a graduated ladder of mitigation around it: the fifteen- and thirty-percent caps for documented reasonable steps, the statutory cure that eliminates the penalty, and — beneath both — the court's discretion under section 2699(e)(2) to cut even a capped award that would be oppressive. Each rung has its own trigger, its own burden, and a decreasing degree of employer control. None of them reaches the premium, which is a wage and survives every one.
Why the penalty is one hundred dollars
The figure is not arbitrary, and understanding its source explains both its size and where it is vulnerable. PAGA does not invent a penalty for each Labor Code section; it supplies a default penalty for those sections that carry none. Many Labor Code provisions specify their own civil penalty — section 226.3, for instance, sets $250 and $1,000 for wage-statement recordkeeping failures, and section 210 sets graduated penalties for late wage payment. Section 226.7, the meal-and-rest provision, specifies no civil penalty at all. Its only built-in remedy is the one-hour premium, which is a wage paid to the employee, not a penalty paid to the state. That gap is what PAGA fills: because section 226.7 fixes no penalty, a meal-and-rest violation draws the section 2699(f) default — one hundred dollars per aggrieved employee per pay period.
Two consequences follow immediately, and both organize the rest of this analysis. First, the penalty is statutory and per-period, so it does not vary with the gravity of any single missed break; one short lunch in a pay period generates the same hundred-dollar exposure as five. The multiplier that matters is therefore not how badly the employer failed but how many employees were affected in how many periods — which is why a modest practice, aggregated across a roster and a year, produces a figure out of all proportion to the wages actually at stake. Second, because the penalty is a creature of the default provision rather than of section 226.7 itself, every mitigation device the reform created operates on that default exposure, and the premium — section 226.7's own remedy — is left entirely outside the penalty calculus. The two move on separate tracks, and the reform's exits run only along one of them.
Two strata, one violation, separate clocks
A single non-compliant meal or rest period sets two obligations running at once, and conflating them is the most common error in an early exposure estimate. The first is the premium under section 226.7 — one hour of pay at the regular rate for each workday a compliant period was not provided. It is a wage: owed to the employee, recoverable in an individual or class action, and governed by a three-year limitations period (longer still where pleaded through the unfair-competition law). The second is the PAGA civil penalty — owed to the state, with the aggrieved employees taking thirty-five percent and the state sixty-five, and governed by PAGA's one-year period. The two have different owners, different measures, different limitations periods, and different mitigation rules, and a complete assessment runs them in parallel rather than collapsing one into the other.
Meal and rest is the paradigmatic PAGA claim precisely because the two strata reinforce each other. The section 226.7 violation is itself the predicate the penalty attaches to; the records that prove the violation are the employer's own time records; and the per-period structure converts an hourly obligation into a six- or seven-figure aggregate across a restaurant's roster. This page develops the penalty stratum and its mitigation ladder. The premium stratum is treated in The Premium Pay Remedy; the derivative wage-statement and waiting-time penalties an unpaid premium can generate — and the scienter firewall that severs them — in Derivative Penalties & Wage Statements.
The rate: initial, subsequent, and the heightened tier
Section 2699(f) states two rates — one hundred dollars for an initial violation and two hundred for a subsequent one — and the reform settled what had been the single most consequential ambiguity in the structure. Plaintiffs had long argued that the first pay period was the "initial" violation and every period after it was a "subsequent" violation at the doubled rate, so that a year of violations ran at $100 for one period and $200 for twenty-five. That reading is now foreclosed. Under amended section 2699(f)(2), the two-hundred-dollar rate applies only where a court or the Labor and Workforce Development Agency has found, within the five years preceding the alleged violation, that the same policy or practice was unlawful — or where the violation was malicious, fraudulent, or oppressive. Absent one of those predicates, the rate is one hundred dollars for every pay period, including the periods after the first.
The two section 2699(f)(2) triggers do not operate identically on the caps below. Malicious, fraudulent, or oppressive conduct does double work for the plaintiff: it lifts the rate to $200 and forfeits the fifteen- and thirty-percent caps under sections 2699(g) and (h), removing the employer's principal mitigation in the same stroke. A prior adverse finding within five years lifts the rate as well, and while it does not by its terms erase the caps, it is close to fatal to them as a factual matter — the caps require all reasonable steps to comply, and an employer that continued a practice already adjudicated unlawful will struggle to prove it took them. Either path converges on the same defense objective: holding the matter in the ordinary $100 tier and preserving cap eligibility both turn on defeating the allegation of prior knowledge or bad faith. And a PAGA notice is not a "finding" — only an adjudication by a court or the agency lifts the rate or, through the malice prong, defeats the caps.
The practical reading for a restaurant is that the ordinary first-suit matter runs at one hundred dollars across the period, and the defense's task at the rate stage is narrow and specific: ensure there is no prior adverse finding on the same practice, and deny the factual basis for a malice characterization. Where a prior finding does exist — a closed agency citation, a stipulated judgment in an earlier matter — its scope and its currency within the five-year window become a contested threshold issue, because it governs both the rate and the availability of the caps below.
The stacking architecture: what doubles and what does not
The largest swing variable in a meal-and-rest penalty model is not the rate but the question of how many penalties a single pay period generates, and the answer requires a distinction the reform sharpened. Two kinds of multiplication must be kept apart. The first is the coexistence of a meal violation and a rest violation in the same period. These are violations of distinct substantive obligations — the duty to provide a meal period and the separate duty to authorize and permit a rest period — and the prevailing view is that each carries its own per-period penalty, so a period in which an employee was denied both a compliant meal and a compliant rest can generate two underlying penalties rather than one. The estimator below treats this as a selectable assumption, because it remains contestable and because it roughly doubles the base.
The second kind of multiplication is derivative stacking, and that the reform expressly curtailed. Before 2024, a plaintiff could argue that a single unpaid premium generated not only the section 226.7 penalty but also separate PAGA penalties for the derivative violations the unpaid premium produced — a section 226 violation because the wage statement was inaccurate, a section 203 violation because the premium went unpaid at separation, a section 204 violation because it was not timely paid. Section 2699(h)(3), echoed in section 2699(m), now bars that layering: no PAGA penalty attaches to a derivative violation of sections 201 through 203, 204, or 226 unless that derivative violation independently satisfies its own scienter element. The distinction is therefore clean even if the vocabulary is not — meal and rest may stack as two underlying penalties, but the derivative penalties built on either one no longer stack on top. The scienter mechanics that govern whether any derivative penalty survives at all are developed in Derivative Penalties & Wage Statements.
The penalty build
A defensible model is built from the variable that actually drives the figure: not the headcount alone, but the number of pay periods in which each employee has at least one violation. A plaintiff's opening demand often assumes a violation in every period for every employee; the real exposure is the rate multiplied by the violating periods, summed across the meal and rest theories, then reduced by whatever rung of the mitigation ladder the employer reaches. The model below exposes each of those levers — including the two that most often separate a plaintiff's number from a defensible one, the stacking assumption and the violating-period frequency. It is an illustration, not an assessment of any matter.
Rest $100 × 14 × 95 = $133,000
State (65%) $197,600
No reasonable-steps showing, no cure: the full default stratum is in play, subject only to the court's § 2699(e)(2) discretion and its authority to limit scope and evidence at trial.
Fig. 1. The penalty build across its real levers. § 2699(f), (g)(1), (h)(1), (j), (e)(2). The estimator is schematic: it assesses penalties on violating periods per employee, lets meal and rest stack or merge, applies the reached cap, and then applies a separate § 2699(e)(2) judicial reduction — the rung the employer least controls. Figures are illustrative and exclude the premiums owed to employees, which are a separate stratum.
The mitigation ladder: three rungs of decreasing control
The reform's mitigation provisions are best read not as a menu of alternatives but as a descending ladder, ordered by how much of the outcome the employer controls. At the top sit the percentage caps, which the employer earns by its own documented conduct and which apply as a matter of right once the showing is made. Below them sits the statutory cure, which the employer initiates but which depends on the agency's or the court's acceptance and on the feasibility of making an entire group whole. At the bottom sits the court's discretion under section 2699(e)(2) — the rung the employer does not control at all, available even where the caps and the cure are unavailable, but committed entirely to the court's judgment about whether the aggregate is oppressive. A matter descends the ladder only as far as the higher rungs fail; the defense objective is to stop it as high as possible, because certainty decreases with every step down.
Fig. 2. The mitigation ladder. The rungs are cumulative and ordered: a matter reaches a lower rung only when the higher ones are unavailable, and the employer's control over the outcome falls at each step. The caps and the cure are developed below; the discretionary rung in its own section.
Rung one: reasonable steps and the percentage caps
The caps are the highest rung because they are the one the employer earns by its own conduct and can establish as a matter of right. Section 2699(g)(1) limits the penalty to fifteen percent of the amount otherwise recoverable where the employer took all reasonable steps to comply before receiving the section 2699.3 notice; section 2699(h)(1) limits it to thirty percent where those steps were taken within sixty days after the notice. Two features govern their use. The burden is the employer's — it is an affirmative showing, proven on the totality of the circumstances, including the employer's size and resources and the nature, severity, and duration of the alleged violations, so the measures that satisfy the standard for a thirty-seat independent will not necessarily satisfy it for a multi-unit operator. And the caps are forfeited by the malicious, fraudulent, or oppressive conduct that also triggers the heightened rate, while the other heightened-rate trigger — a prior adverse finding on the same practice — makes the reasonable-steps showing nearly impossible as a factual matter, because an employer that continued an adjudicated-unlawful practice cannot readily prove it took all reasonable steps to comply. The showing that preserves the caps and the showing that holds the rate at $100 therefore largely coincide.
The statute names the qualifying steps without making the list exhaustive, and the operative feature of each is that it is proven by a contemporaneous record rather than by later assertion. The reasonableness inquiry asks what the employer did and documented when it had the chance, not what it can reconstruct for litigation.
Fig. 3. The reasonable-steps factors under § 2699(g)–(h). The documented pre-notice audit is what separates the fifteen-percent tier from the thirty-percent tier, and either from the uncapped figure — which is why it is worth running before any notice arrives and is the first act in the timing sequence below.
Rung two: the statutory cure
The cure is the second rung because it can eliminate the penalty entirely, but it is below the caps because the employer controls only its initiation, not its acceptance, and because what counts as a cure is statutorily demanding. Section 2699(d) defines the term with three cumulative requirements: the employer must abate each violation alleged, come into compliance with the underlying statute identified in the notice, and make every aggrieved employee whole. Section 2699(j) then provides that no civil penalty is assessed against an employer that proactively complies and cures after receiving notice. The mechanics differ by headcount, and the threshold is the number of employees employed during the limitations period.
The employer may request an early evaluation conference and a stay of discovery and of the responsive-pleading deadline. A neutral evaluator reviews the employer's confidential submission — the disputed allegations, the violations it intends to cure, and its proposed cure plan — together with the plaintiff's response. The conference statements and discussions are inadmissible to prove liability under Evidence Code section 1152, so the employer can engage candidly without manufacturing admissions. The process need not run beyond thirty days unless the parties agree, and if the evaluator or the plaintiff disputes the sufficiency of the cure, the employer may move the court to approve it on a showing of correction.
Two features make the cure harder for a meal-and-rest claim than for a wage-statement defect, and both bear on whether it is worth attempting. First, a missed break is a completed past event, so the cure cannot undo it; section 2699(d) is satisfied only by correcting the practice going forward and paying the premiums owed, with interest, to make each affected employee whole. In a representative posture that make-whole obligation reaches every aggrieved employee, so where the same gap recurs across shifts, locations, or a long period, the completeness and cost of making the entire group whole is what determines whether the cure is realistic at all. Second, and more subtly, a cure presupposes a violation to abate. An employer that genuinely contests whether any violation occurred — that maintains compliant periods were provided and the records reflect employee choice rather than denial — cannot be made to "cure" a violation it disputes without conceding the merits. The cure is therefore most useful where the violation is clear and the group is bounded, and least useful in precisely the contested, sprawling matter where the penalty exposure is largest. There, the analysis turns to the rung below.
Rung three: judicial discretion under § 2699(e)(2)
Beneath the caps and the cure sits a provision that predates the reform and that the reform deliberately strengthened. Section 2699(e)(2) authorizes a court to award less than the maximum civil penalty where, on the facts and circumstances of the particular case, the full amount would be "unjust, arbitrary and oppressive, or confiscatory." This is the backstop that operates when the higher rungs are unavailable — when the employer cannot establish reasonable steps and cannot or will not cure, but the aggregate penalty that results is grossly disproportionate to any harm. The 2024 reform did not merely retain this discretion; it expressly extended it to operate on the cap amounts themselves, so that a court may award below the fifteen- or thirty-percent figure, or in the unusual case exceed it, under the same standard. The discretionary rung therefore sits beneath every other: even a capped penalty remains subject to reduction if it is oppressive on the facts.
The provision's limits should be stated as plainly as its promise, because overstating it is a defense error. The discretion is real but it is the court's, not the employer's, and the leading application cautions against assuming it. In Amaral v. Cintas Corp. No. 2 (2008) 163 Cal.App.4th 1157, the Court of Appeal confirmed the power to reduce an oppressive aggregate but affirmed the trial court's refusal to exercise it, holding that the employer's good-faith arguments on unsettled law did not by themselves render the maximum unjust. The lesson is that section 2699(e)(2) is not a good-faith discount that attaches automatically to a defensible position; it is a discretionary judgment reserved for awards that genuinely shock the conscience of the court, invoked by a developed record of disproportion — the ratio of the penalty to the wages actually at issue, the absence of employee harm beyond the technical violation, the employer's resources and ability to pay, and the deterrent purpose already served.
Defendants sometimes frame the disproportion as a federal due-process or Excessive Fines argument against the aggregate. That route has generally fared poorly: courts have declined to read the Constitution as a categorical cap on aggregated statutory penalties, and have rejected proposed per-violation reductions as arbitrary where untethered to actual harm. The durable argument is therefore statutory, not constitutional — section 2699(e)(2) gives a California court the express, particularized discretion that the constitutional theory asks it to find by implication, and it is the channel through which a disproportion argument is most credibly made.
For the defense, the discretionary rung reframes the worst case. A matter in which the caps are forfeited and a cure is infeasible is not, for that reason, a matter exposed to the full aggregate as a fixed sum to be negotiated. It is a matter in which the residual judicial discretion becomes the operative lever, and the work shifts from earning a cap to building the disproportion record that section 2699(e)(2) requires — which is also the record that a court invokes when it uses the case-management authority that Estrada left it.
What the ladder does not reach
Every rung of the ladder operates on the civil penalty. None discharges the premium. Because the section 226.7 premium is a wage rather than a penalty, it sits outside the PAGA structure entirely: it is owed for each workday a compliant period was not provided, it carries a three-year limitations period rather than PAGA's one-year period, and it remains recoverable in an individual or class action even after the penalty has been capped to fifteen percent, reduced under section 2699(e)(2), or eliminated by a cure. Indeed the make-whole payment that a cure requires is, in substance, the payment of those premiums — so a cure does not avoid the wage cost; it converts a matter carrying both a wage and a penalty into one carrying the wage alone.
The strategic implication runs in one direction and should discipline how a favorable PAGA result is read. The premium is the smaller, more certain number; the penalty is the larger, more contingent one. The ladder's rungs all operate on the penalty — which is where the leverage is — but they leave the wage in place, and they do not touch the separate three-year class exposure for the premiums or the derivative wage-statement and waiting-time claims an unpaid premium can generate. A complete assessment therefore models the penalty stratum and the wage-and-derivative stratum side by side; a capped or cured PAGA penalty resolves one stratum, not the matter.
Manageability, and the discretion it points to
A court has no inherent authority to strike a PAGA claim because it is unmanageable; PAGA is not a class action, and grafting a manageability-to-dismiss requirement onto it is impermissible. But the Court was equally explicit that PAGA's public purpose is subordinate to constitutional due process, and that trial courts retain their ordinary tools — expressly including limiting the evidence at trial, narrowing the claim's scope, and limiting the penalties — to keep a sprawling representative case within bounds.
The connection to the rung above is direct and easy to miss. When Estrada lists "limiting the penalties" among the tools a trial court retains, the statutory vehicle for that limitation is section 2699(e)(2). Manageability and discretionary reduction are therefore two faces of the same residual authority: the court cannot dismiss an unwieldy meal-and-rest PAGA claim, but it can decline to award penalties for periods or employees the plaintiff cannot prove with admissible, manageable evidence, and it can reduce an aggregate that proof would otherwise produce to a figure that is not oppressive. The defensive course in a large matter is accordingly not to seek dismissal but to constrain — to hold the plaintiff to reliable proof of the violating-period multiplier, to resist an extrapolation that due process would not permit, and to build the disproportion record that supports a section 2699(e)(2) reduction of whatever survives. That trial-management dimension, and the limits on representative proof, are developed in Certification and the Trial Plan.
Standing and the one-year clock
The reform also narrowed who may bring the claim and how far back it reaches, and the two limits work in tandem to bound the exposure before the penalty is ever computed. The first is a standing requirement. Under amended section 2699, an aggrieved employee must have personally suffered each of the Labor Code violations alleged in order to prosecute them representatively — abrogating Huff v. Securitas Security Services USA, Inc. (2018) 23 Cal.App.5th 745, under which a plaintiff affected by a single violation could pursue penalties for every other violation the employer committed, including categories the plaintiff never experienced. A narrow exception preserves the broader Huff standing for actions brought by qualifying nonprofit legal-aid organizations, so the reform's narrowing binds the ordinary private plaintiff rather than the legal-aid bar. The second limit is temporal. The action carries a one-year limitations period under Code of Civil Procedure section 340, and the plaintiff must have suffered the predicate violation within that year. That forecloses the argument — which had gained traction after Johnson v. Maxim Healthcare Services (2021) 66 Cal.App.5th 924 held that a time-barred individual claim did not by itself defeat representative standing — that a plaintiff could anchor a PAGA action on a violation experienced at any time, so that the limitations period imposed no real constraint. Both limits now operate as threshold filters: they strike theories and disqualify plaintiffs before the per-period penalty is ever multiplied across the roster. The standing question becomes acute where the employer has an arbitration program, because directing the plaintiff's individual claim to arbitration can test the very status on which representative standing depends — the subject of Arbitration, Standing & the Headless-PAGA Split.
The timing architecture
The provisions assemble into a sequence governed by overlapping clocks, and the clocks are where a mitigation strategy is won or lost. The first and most valuable window opens before any notice exists: the documented audit, conducted while no litigation is pending and the cost of correction is lowest, is what later qualifies the employer for the fifteen-percent cap and drives down the violating-period multiplier the plaintiff must prove. It cannot be created after a notice arrives — section 2699(g) rewards only steps taken before the notice — so the pre-notice posture is built in advance or not at all.
Once a notice arrives, two clocks run at once and they do not align. The reasonable-steps window for the thirty-percent cap runs sixty days from the notice; the small-employer cure clock under section 2699.3 runs thirty-three days. An employer eligible for both must therefore sequence its response against the shorter deadline, deciding within thirty-three days whether to pursue a cure that eliminates the penalty or to spend the period building the post-notice reasonable-steps record for the thirty-percent cap — while preserving the section 2699(e)(2) disproportion record as the backstop if neither succeeds. The penalty a matter ultimately carries is, in the ordinary case, a function of how early and how completely that work was done. That is why the analysis treats the pre-notice audit as the first defensive act, the post-notice cure-or-cap decision as the second, and the discretionary reduction as the last — rather than treating the penalty as a fixed exposure to be addressed only at settlement.