106 Companies, Rewritten Clauses, and a Preemption Argument Nobody Expected

Most commentary on mass arbitration is anecdotal by necessity — a decision here, a settlement there, a provider rule change that everyone reads the same way. Richard Frankel's study in the Vanderbilt Law Review does something different. It looks at what companies actually put in their contracts.

The dataset

Frankel analysed arbitration provisions from 106 large consumer-facing companies, examining how those provisions had been modified in response to the mass arbitration campaigns of the preceding years.

The findings describe a coordinated redrafting wave rather than scattered experimentation. Three patterns dominate.

Pre-arbitration procedures are now the norm. Most companies in the sample require some process — notice of dispute, an informal resolution period, a settlement conference, documentary preconditions — before a claim may be filed at all.

Sequential batch processing is widespread. Many provisions require that claims proceed in tranches rather than simultaneously, in the family of structures that produced Achey and Rios.

Provider selection became a cost lever. Some companies moved to providers offering reduced fee schedules — a decision that, after the Ninth Circuit's treatment of a bespoke provider's mass rules, now carries its own risk.

The claim-suppression argument

Frankel's normative argument is that these modifications produce claim-suppressing consequences, and he identifies two mechanisms.

The first is dismissal on procedural grounds. Every added precondition is a new way for a demand to fail without any assessment of its merits. Applied to a single claimant, a notice requirement is a modest formality. Applied across ten thousand claimants, strict enforcement of individualised notice requirements becomes a filter that removes a meaningful share of the inventory on grounds having nothing to do with whether the underlying claims are good.

The second is delay through staggering. Sequential processing spreads the respondent's cost and correspondingly extends every claimant's wait. For claims of modest individual value, delay of sufficient magnitude functions as denial — a point on which the case law has since converged independently.

An important feature of the argument is that it does not depend on attributing bad faith to any individual drafter. Each provision, considered alone, has a legitimate justification. Suppression is the aggregate property of the stack.

The preemption move

The study's most consequential contribution is a doctrinal one, and it inverts the usual argument.

For thirty years, FAA preemption has run in one direction: state law that obstructs enforcement of arbitration agreements is displaced. That is why unconscionability — a generally applicable contract defence preserved by the FAA's saving clause — has been almost the only surviving route of attack.

Frankel argues that heavily engineered modern clauses may have travelled outside the statute's protection entirely. The FAA was enacted to enforce agreements to submit disputes to arbitration: a relatively streamlined, arbitrator-decided alternative to litigation. A regime that layers mandatory notice, informal resolution windows, evidentiary preconditions, tranche-based sequencing, bespoke bellwether protocols, and provider-specific mass rules on top of that core is not obviously the same thing.

If it is not, the consequence is significant. State legislatures could regulate such procedural apparatus without confronting preemption, because there would be nothing the FAA protects to preempt with.

How to read it

The argument has not been adopted by any court, and it faces obvious difficulties — not least drawing a principled line between permissible procedural specification and disqualifying complexity. Parties have always been free to design their own arbitral procedures, and that freedom is itself a core FAA value.

But the framing matters for anyone advising on either side. It supplies claimants' counsel with an argument that operates upstream of unconscionability, at the level of statutory coverage rather than contract defence. And it supplies defence counsel with a reason to be cautious about the accumulation of procedural terms that seems, clause by clause, entirely rational: the more elaborate the machine, the more available the argument that it is no longer arbitration in the sense the statute contemplates.

That is a different kind of drafting risk from unconscionability, and it does not respond to the same fixes.