One hundred dollars, per employee, per pay period
For a Labor Code provision that does not carry its own civil penalty, section 2699(f) supplies the default: one hundred dollars for each aggrieved employee for each pay period in which the violation occurred. Two structural features of that sentence do all the damage. The unit of accrual is the pay period, not the violation and not the employee — so a single recurring practice generates a fresh penalty on every pay date it persisted. And the penalty runs per aggrieved employee, so the same practice multiplies across the entire affected workforce. A restaurant that pays biweekly produces twenty-six pay periods per employee per year; a practice baked into the payroll repeats on every one of them, for everyone, until it is fixed.
The reform did not lower the default. It left one hundred dollars in place and instead worked at the edges — narrowing the old elevated tier, adding two reduced tiers, halving the figure for weekly payrolls, and capping the result for compliant employers. The order of operations matters and is worth stating plainly: first determine the per-period rate from the tier (§ III), apply the weekly halving if it applies (§ IV), multiply by the employee and period counts to reach the gross (§ V), and only then apply the reasonable-steps cap (developed on the caps page) and the distribution. This page works the first three steps; the cap is the next page.
The first multiplier is a contested number, not a given
The penalty multiplies across "aggrieved employees," and the size of that population is one of the most consequential and most litigated inputs. The reform sharpened the definition in the employer's favor at the threshold: under section 2699(c)(1), the named plaintiff must have personally suffered each Labor Code violation alleged, within the one-year limitations period — which constrains the theories a plaintiff may pursue representatively, though not directly the size of the represented group once a theory is properly anchored. That standing question is the subject of the standing page; what matters here is that the aggrieved-employee count is not conceded by being pleaded.
For the penalty arithmetic, the count is the number of employees who experienced the particular violation during the penalty window — and it is rarely the entire roster. A practice that affected only servers does not sweep in the kitchen; a regular-rate error that touched only employees who earned a nondiscretionary bonus does not reach those who did not. Pressing the gap between the pleaded class and the actually-affected population is a front-end defense that shrinks the first multiplier before any cap is reached, and it is often where the real reduction in exposure begins.
What sets the rate — and when malice removes the cap
The per-period rate is not always one hundred dollars. The reform created a graduated structure: three tiers move the rate down, and one moves it up. The heightened tier carries a second consequence that is easy to miss and easy to overstate. Its two triggers are not equivalent — a finding that the conduct was malicious, fraudulent, or oppressive both doubles the rate and removes the reasonable-steps caps entirely (§ 2699(g), (h)), while a prior unlawful finding within five years doubles the rate but leaves the caps available, though it makes the reasonable-steps showing far harder to sustain. The caps otherwise reduce even a doubled penalty. Classify the violation:
The baseline. One hundred dollars for each aggrieved employee for each pay period in which the violation occurred. This is the rate a plaintiff assumes across the entire class and limitations period to build the gross figure — and the rate the reasonable-steps caps act upon.
§ 2699(f)Fig. 1. The § 2699(f) penalty tiers. All of them — the reduced $25/$50, the $100 baseline, and the heightened $200 under (f)(2) — are reduced by the 15% and 30% caps; what removes the caps is not the rate but a finding of malicious, fraudulent, or oppressive conduct (§ 2699(g), (h)). The reduced and heightened amounts are mutually exclusive characterizations of the same per-period unit, not additive.
The reform did not make the penalty smaller; it made the penalty depend on the employer's conduct — and reserved one tier that no amount of later compliance can reduce.
Section 2699(o) halves the penalty for weekly payrolls
Because the penalty accrues per pay period, an employer that pays weekly was, before the reform, exposed to roughly twice the penalty of an identical employer paying biweekly — fifty-two periods a year against twenty-six, for the same underlying conduct. Section 2699(o) corrects that by halving the otherwise-applicable penalty where the regular pay period is weekly rather than biweekly or semimonthly. The halving applies to whichever tier governs — it takes the hundred-dollar default to fifty, the fifty-dollar isolated-event amount to twenty-five — and it is applied before any reasonable-steps cap, so a weekly-pay employer that also earns the fifteen-percent cap stacks the two reductions. For a restaurant industry in which weekly pay is common, this is a real and frequently overlooked adjustment to the gross figure.
The three-number product
The gross penalty is a product of three numbers — the aggrieved-employee count, the number of pay periods in which the violation occurred, and the per-period rate from the tier, halved for weekly pay. The calculator below builds that figure and then shows the distribution, so the same inputs display both the number a plaintiff asserts and how it would be split. It deliberately stops at the gross-and-split stage: the reasonable-steps cap, which is where the figure actually contracts, is the next page, and the anti-stacking limits that prevent the derivative penalties from compounding are developed on the anti-stacking page.
Gross = aggrieved employees × pay periods × per-period rate (halved if weekly). Pick the tier and the pay frequency.
- Fees are separate. A prevailing plaintiff's attorney's fees and costs are awarded on top of the penalty, not carved out of it.
- The cap is not in this number. This is gross; the 15%/30% reduction is applied next, on the caps page.
- Malice is the trap, not the rate. A malice finding both doubles the rate and removes the caps; a prior finding only doubles the rate. The caps reduce even a doubled penalty otherwise.
Fig. 2. Illustrative. Lab. Code § 2699(f), (o). Gross penalty only; the reasonable-steps cap, the § 2699(e)(2) discretion to reduce, and any underlying wage liability are separate, and attorney's fees are additional. The split is of the penalty after any cap, shown here on the gross for illustration.
Most of the penalty is the state's — and the fees are the engine
The reform raised the employees' share of any recovered penalty from twenty-five to thirty-five percent, leaving sixty-five percent to the Labor and Workforce Development Agency. That split has two consequences worth holding. First, the penalty is fundamentally a public recovery — the represented employees receive a minority share, which is part of why PAGA is framed as law enforcement rather than a class device and informs the standing and manageability analyses. Second, and decisive in practice, the attorney's fees and costs awarded to a prevailing plaintiff are separate from and additional to the penalty, not a percentage of the employees' share. A modest penalty can fund substantial fee exposure, which is why the realistic settlement value of a PAGA matter is rarely just a fraction of the gross penalty — the fee award is its own line, and it travels with the merits.