The spine is settled: Iskanian bars wholesale PAGA waivers because the claim is the state's; Viking River lets the employer compel the individual claim; Adolph holds the representative claim survives that arbitration, stayed under section 1281.4. Compelling pauses the representative claim — it does not end it.
Three joints are contested, and each is dispositive. Whether a plaintiff may plead a headless action with no individual claim is before the Supreme Court in Leeper. Whether an arbitral loss precludes the representative claim divides the districts. And whether the employer can keep the arbitration it compels turns on the fee-default statutes after Hohenshelt. The strategic question — whether to move to compel at all — is answered only by running all three.
Why the arbitration question comes first
For a restaurant operator with an arbitration program, the first question in a meal-and-rest matter is rarely whether a break was missed. It is whether the representative claim can be directed away from court — and, increasingly, whether moving to do so helps or harms. The reason the question leads is structural. A large share of food-service employees sign arbitration agreements with class and representative-action waivers at hire; the individual value of a single missed-break claim is small; and the economics that make the litigation worth bringing depend on aggregation. The representative vehicle is the case. The contest over whether that vehicle reaches a courtroom is therefore typically dispositive of the matter's scale, and it is fought before the merits of any individual break.
What has changed is that the contest is no longer one-directional. A decade ago the employer's interest was simply to compel; today, after Adolph severed the link between individual arbitration and representative dismissal, and after the fee-default statutes created a way to lose the forum one has won, the motion to compel is a genuine strategic decision with downside as well as upside. This page develops the settled spine, the three contested joints that determine the outcome, and the decision — treated last, because it is the synthesis of everything before it — of whether to compel at all.
The distinction the whole area runs on
Every move below depends on a single distinction that Viking River drew and named, and imprecision about it is the most common source of error in the analysis. A PAGA action contains two analytically separate claims. Understanding the rest of the page requires holding them apart.
The claim for civil penalties for the Labor Code violations the plaintiff personally suffered. After Viking River this component is arbitrable — the FAA permits compelling it — and it is the claim a motion to compel actually moves.
The claim for penalties for violations suffered by other aggrieved employees. It is the volume of the case, it cannot be waived (Iskanian), and under Adolph it is stayed rather than dismissed when the individual claim goes to arbitration.
Fig. 1. The Viking River taxonomy. The individual claim is the lever; the representative claim is the load. The headless device is an attempt to plead only the right-hand box; the preclusion question asks whether losing the left-hand box in arbitration destroys standing for the right; and the fee-default rules govern whether the employer can hold the left-hand box in arbitration at all.
The settled spine
Three decisions built the framework that governs an action with an ordinary individual component, and each is best understood through the qui-tam premise that anchors all of them. Iskanian held the representative waiver unenforceable because a PAGA claim is not the employee's private claim at all: the statute deputizes the employee to sue as a proxy for the state, which is the real party in interest and which has not agreed to anyone's arbitration clause. An employee cannot contract away an action that belongs to the sovereign. Viking River accepted that premise but held the FAA preempts Iskanian on one point only — California could not forbid dividing the action so as to send the individual claim, the employee's own, to the arbitration the employee did agree to. The Court compelled the individual claim and predicted the representative claim would collapse for lack of a plaintiff with standing. Adolph then held that prediction wrong as a matter of California law: PAGA standing requires employment by the alleged violator plus one violation suffered, and arbitrating the individual claim removes neither, so the representative claim persists — stayed, not dismissed.
Fig. 2. The settled spine. The result is a stay-and-sequence procedure: arbitration of the individual claim does not by itself end the representative action. Whether anything ends it depends on the three contested joints that follow.
The sequence in operation
The choreography is established. The employer moves to compel the individual claim; the court stays the representative claim under Code of Civil Procedure section 1281.4; the two proceed in parallel, with the arbitration ordinarily resolving first. What the stay does — and does not do — is the point practitioners most often misstate. The stay preserves the representative claim in suspension; it does not resolve it. The claim's fate turns entirely on what the arbitration produces, and there are three possibilities, only one of which ends the case.
If the plaintiff prevails individually, the representative claim revives with its standing intact and proceeds to certification-style proof in court — the employer has gained a forum on the individual claim but no containment of the representative one. If the matter settles individually, Kim v. Reins teaches that settlement of the claim for individual redress does not strip aggrieved-employee status, so the representative claim again survives. Only the third outcome — a merits determination that the plaintiff suffered no violation, reduced to a confirmed award — offers the prospect of ending the representative claim, and it does so not through the stay but through preclusion. That is why the preclusion question, not the stay, is where the real leverage of compelling resides. Two gateway refinements precede all of this: whether a clear and unmistakable delegation clause sends arbitrability itself to the arbitrator, and whether the agreement will survive the enforceability attack the plaintiff will mount to keep the individual claim in court in the first place.
The preclusion question: the only path that ends the case
Because Adolph foreclosed the argument that arbitration itself defeats representative standing, the employer's remaining route to ending — not merely pausing — the representative claim runs through issue preclusion. The theory is exact: PAGA standing requires that the plaintiff be an aggrieved employee, meaning one against whom a Labor Code violation was committed; if an arbitrator finally determines, on the merits, that no violation was committed against the plaintiff, that determination collaterally estops the plaintiff from relitigating the identical fact in court, and the representative claim fails for want of a qualified plaintiff. Whether that theory holds is itself split across the districts, and a defendant must know which way the assigned court leans before betting the strategy on it.
The Second District holds that once an arbitrator finally determines no Labor Code violation occurred, issue preclusion bars the plaintiff from relitigating that fact to establish aggrieved-employee status — and it applies regardless of the capacity in which the plaintiff appears, because the identical factual issue was actually litigated and decided. The PAGA claim, predicated on a violation the plaintiff failed to prove, fails for want of standing.
Fig. 3. The preclusion sub-split. The mechanism is the employer's single most powerful tool — it converts an individual arbitration win into a representative dismissal — but it is unsettled (the capacity objection) and fragile (it dies with the award). A defendant who compels in reliance on preclusion should confirm the controlling district view and protect the award against vacatur.
The interaction with the headless device is the strategic crux. The preclusion route requires an individual claim to litigate and lose; the headless pleading is the plaintiff's attempt to deny the employer that target by removing the individual claim from the case. The two contested joints are therefore the same fight approached from opposite ends — the employer pressing to establish an arbitrable individual claim it can win and give preclusive effect, the plaintiff pressing to plead a claim with no individual component to compel. Which prevails is the question now before the Supreme Court.
The headless device
Adolph and the preclusion line together produced a foreseeable counter-move. If the representative claim survives arbitration but an individual loss in arbitration can preclude it, the plaintiff's optimal play is to keep the individual claim out of arbitration altogether — to deny the employer both the forum and the preclusion vehicle. The device is to plead the PAGA action with no individual claim at all: to seek civil penalties only for other aggrieved employees, expressly disclaiming any individual recovery, so that there is nothing individual for the employer to compel and nothing on which an adverse arbitral finding could operate. The action proceeds, if it proceeds, without a head. Whether the device works is the question the Courts of Appeal have divided on, and the division reduces, with deceptive simplicity, to a single conjunction in the statute.
The split: one word, two readings
PAGA authorizes an action brought “on behalf of himself or herself and other current or former employees.” Everything turns on the “and.” Select a reading for the authority, the textual argument, and the consequence for an arbitration program.
The statute authorizes an action “on behalf of himself or herself and other current or former employees.” On this reading the “and” is conjunctive: a viable PAGA action must contain both an individual and a representative component, so a plaintiff cannot disclaim the individual claim, and it may be compelled to arbitration even where the complaint omits it. Williams reinforces the point from the limitations side — the plaintiff must hold a timely individual claim as to at least one violation, because the filing period runs from the individual injury.
Fig. 4. The two readings of the statutory “and.” The Second District (Leeper, Williams) reads it conjunctively; the Fourth and Fifth (Packers-Rodriguez, CRST), building on Balderas, read it permissively. Note the doctrinal tension within the Rodriguez name: Rodriguez v. Lawrence Equipment (preclusion) and the unrelated Packers-Rodriguez (headless) point in opposite strategic directions and should not be conflated.
Before the California Supreme Court
The division prompted the California Supreme Court to grant review in Leeper v. Shipt on its own motion, limiting the issues to the two the split presents: whether every PAGA action necessarily includes both an individual and a non-individual claim regardless of the pleading, and whether a plaintiff may bring a non-individual action only. Williams and the Packers-Rodriguez line were granted review and are held behind Leeper as the lead vehicle. As of this writing the matter remains pending; its disposition should be confirmed before reliance.
One feature of the posture deserves weight in any present assessment. The Leeper complaint was filed in March 2024, before the operative date of the 2024 reform, and the panel nonetheless construed the current text — which reviewing courts have read to signal that the Supreme Court will likely analyze the pre-amendment statute when it decides the case. The significance for a restaurant defendant is that the reform may resolve much of the question prospectively regardless of how Leeper comes out: the amended standing provision requires an aggrieved employee to have “personally suffered each of the Labor Code violations alleged,” and Williams independently demands a timely individual claim tied to the limitations period. Both press toward requiring a head for any post-reform matter. A defendant should therefore distinguish two regimes — pre-reform conduct, where Leeper may control directly, and post-reform conduct, where the statutory standing requirement supplies a convergent argument for an individual component even if Leeper is read narrowly or decided for the plaintiff on the older text.
Holding the forum: the fee-default trap
An employer can win the motion to compel and still lose the arbitration — not on the merits, but by operation of a statute that practitioners on the defense side underestimate. Code of Civil Procedure sections 1281.97 and 1281.98 provide that in an employment arbitration the drafting party, the employer, must pay the arbitration provider's invoices within thirty days of the due date; a failure to do so is a material breach that allows the employee to withdraw from arbitration, return to court, and seek sanctions. The provision exists because the drafting party controls the forum it insisted upon, and a late or withheld payment can stall the very arbitration the employee was forced into. For a defendant whose entire containment strategy depends on keeping the individual claim in arbitration long enough to obtain a preclusive award, an inadvertent fee lapse can surrender the forum and hand the plaintiff a return ticket to the courtroom — after the employer has already conceded arbitrability and paid to get there.
Hohenshelt v. Superior Court (2025) 18 Cal.5th 310 softened the rule without removing the trap. The California Supreme Court held the fee-payment statutes are not preempted by the Federal Arbitration Act, rejecting the employer's preemption argument, but it also rejected the rigid construction several Courts of Appeal had adopted under which any late payment, for any reason, worked an automatic forfeiture. Properly construed, the Court held, forfeiture follows only from nonpayment that is willful, grossly negligent, or fraudulent, with relief available for a good-faith mistake, inadvertence, or excusable neglect — the ordinary contract principle against forfeiture. The post-Hohenshelt applications confirm the softer line: in Wilson v. Tap Worldwide a payment initiated on the due date and received one business day later was not strategic and did not forfeit arbitration, and Wilson v. VXI Global Solutions extended the same relief to the initiation fees governed by section 1281.97.
Hohenshelt is relief from the harshest outcomes, not a license for casual payment. The defensible practice is a calendared, redundant fee-payment protocol that treats every arbitration invoice as a thirty-day deadline with no slack, because the cost of a missed deadline — loss of the compelled forum — dwarfs the administrative burden of paying on time.
A second exposure runs alongside the statute. The arbitration provider's own rules may be stricter than California law: the AAA's employment rules treat a fee default as an automatic waiver of the right to arbitrate without regard to intent or prejudice. Where the agreement incorporates those rules, the provider's stricter standard may govern the proceeding regardless of the Hohenshelt gloss — so the choice of provider and the rule set the agreement adopts are themselves part of the fee-default risk, not merely the statute.
Drafting the agreement to the framework
The framework rewards precise drafting and punishes its absence in three concrete ways, and the agreement that does not address all three is a paper defense. The first is the severability architecture, and Viking River itself is the lesson. The agreement there contained a wholesale waiver of representative actions that was invalid under Iskanian — but it also contained a severability clause providing that any portion of the waiver that could be enforced would be. That clause is what permitted the Court to strike the invalid representative-waiver portion while preserving, and compelling, the individual PAGA claim. The drafting imperative is therefore a severability clause written to save individual-PAGA arbitration as a freestanding obligation that survives the invalidation of the representative waiver around it.
The second is the inverse hazard, the one that converts a drafting choice into a self-inflicted defeat: the poison pill. An agreement that provides the opposite — that if the representative-action waiver is unenforceable, the entire arbitration agreement is void, or that no PAGA claim in any form may be arbitrated — hands the plaintiff the result the plaintiff wants. Under such a clause the invalidity of the representative waiver takes the individual-arbitration right down with it, leaving the employer with nothing to compel and the whole action in court. Several employers learned after Viking River that their pre-existing agreements were drafted exactly this way, written for a pre-2022 world in which the representative waiver was thought to stand or the agreement was thought to fall as a unit. Reviewing the existing agreement for a latent poison pill is often more urgent than perfecting a new one.
The third is enforceability itself, because none of the above matters if the agreement is unconscionable. Armendariz v. Foundation Health Psychcare Services (2000) 24 Cal.4th 83 requires a lawful employment arbitration agreement to supply a neutral arbitrator, adequate discovery, a written and reasoned award subject to limited review, the full range of statutory remedies, and the employer's payment of the costs unique to the arbitral forum — and the agreement must not be unconscionable, which California measures as a sliding-scale combination of procedural unfairness in formation and substantive one-sidedness in terms. An agreement engineered for containment that is oppressive in how it was presented at hire, or lopsided in its terms, can be struck in whole or in part, defeating the leverage it was meant to create. The through-line is that the three drafting goals are cumulative: a severability clause that saves individual arbitration, the absence of a poison pill that would sink it, and clean Armendariz compliance that keeps the whole agreement enforceable. An agreement that achieves two of the three still fails.
The decision: whether to compel at all
The doctrine assembles into a decision the older practice treated as automatic and current practice cannot. Before Adolph, compelling the individual claim plausibly ended the representative claim, so the motion was reflexive. After Adolph it does not — compelling merely stays the representative claim, and whether anything more follows depends on preclusion, which is contested; on the pleading, which may be headless; and on fee discipline, which can forfeit the forum. A motion to compel that buys only a parallel proceeding, adds fee-default exposure, and delivers no containment is affirmatively worse than not moving. The considerations below are the ones that determine which case a defendant is in; select each to see how it cuts.
Is the individual claim strong on the merits for the employer, and is the agreement clean enough that an award will be confirmed?
Fig. 5. The decision to compel, decomposed. The reflexive motion is obsolete: compelling is strongly indicated where preclusion is realistic and the agreement is clean, and contraindicated where the individual claim is weak, fee discipline is uncertain, or the pleading is headless and the individual claim must first be established. The analysis is matter-specific and illustrative; it decides no case.
The synthesis is that compelling is no longer a default but a bet, and the bet pays only when the pieces align: an arbitrable individual claim that exists and can be compelled, merits strong enough that the award will be favorable, a controlling district that gives that award preclusive effect, an agreement clean enough to be enforced and to survive confirmation, and fee discipline reliable enough to hold the forum to the end. Where those align, compelling is the most powerful move available — it is the only path that ends rather than pauses the representative claim. Where they do not, the better course is frequently to litigate the representative claim in court and contain it through the trial-management and penalty defenses developed elsewhere in this section, rather than to finance a second forum that returns no containment.
The multi-front overlay
A final dimension governs the increasingly common situation in which more than one PAGA action targets the same employer over overlapping conduct — a frequent occurrence in food service, where multiple firms may file on the same wage-and-hour theory across overlapping aggrieved-employee groups. Turrieta v. Lyft (2024) 16 Cal.5th 664 held that a PAGA plaintiff in one action lacks standing to intervene in, object to, or move to vacate the judgment in another plaintiff's overlapping PAGA action: the plaintiff's status as the state's deputy confers authority to prosecute the plaintiff's own case, not a personal interest in a different deputy's case. For the defendant the consequence is architectural. A defendant facing parallel representative claims may be able to resolve one — typically the one with the most cooperative posture or the weakest claim — and obtain a court-approved judgment that binds the state's interest, without affording the competing plaintiffs a procedural foothold to intervene and disrupt the settlement. The first matter to reach an approved judgment can thus foreclose the others, which makes the choice of which front to settle, and the speed of getting it approved, a strategic decision in its own right. The settlement architecture and the LWDA-approval mechanics that govern that resolution are developed in Certification and the Trial Plan.
Whatever the resolution of the headless question, a representative meal-and-rest claim that survives — because the agreement could not be enforced, the pleading was held validly headless, the individual claim was won by the plaintiff, or the employer elected not to compel — returns to court as a representative proceeding without class certification. How a court structures proof of violations across the aggrieved group without a class device, within the due-process limits that constrain representative evidence, is the second front on which a large meal-and-rest matter is won or contained, and it is the subject of the next page.