One Claimant's Harassment Claim Can Blow Up Your Entire Arbitration. Your Completion Model Should Account for That.

Here is something most plaintiff firms running mass arbitration campaigns have not modeled: a single claimant's sexual harassment allegation can pull your entire matter out of arbitration.

Not that claimant's case. The entire matter.

The Second Circuit said so last month in Diaz-Roa v. Hermes Law, and the implications for completion economics are significant enough that every firm running a live campaign should be rethinking how they build their claimant roster.

What the Court Actually Said

The Ending Forced Arbitration of Sexual Assault and Sexual Harassment Act (EFASASHA) lets a claimant elect to void an arbitration agreement if their dispute relates to sexual harassment or assault. That much was clear from the statute. What was not clear, until Diaz-Roa, was the scope of that election.

The Second Circuit held that when a claimant invokes EFASASHA, the entire arbitration agreement becomes voidable, not just the harassment-related claims. In Diaz-Roa, that meant wage and contract claims that had nothing to do with harassment also escaped arbitration. The court read the statute's text ("at the election of the person alleging conduct constituting a sexual harassment dispute... no predispute arbitration agreement... shall be valid or enforceable") as applying to the agreement itself, not to individual claims within it.

For a single-plaintiff employment case, that is a meaningful expansion of litigation options. For a mass arbitration campaign with thousands of claimants, it is a denominator problem.

Why This Is a Completion Variable, Not Just a Legal Issue

Think about how most firms model a mass arb matter. You start with a claimant pool. You model your release threshold (typically 75% to 85%). You build an outreach campaign designed to push enough claimants through signing to hit that number. Your entire completion budget, your per-claimant outreach cost, your admin fee structure, all of it sits on top of a denominator: the total claimants in the arbitration.

Now imagine you are running a 5,000-claimant consumer arbitration against a company with a mixed workforce. Some of those claimants are employees or former employees. Some of those employees experienced harassment. If even a subset of your claimant pool includes individuals who can invoke EFASASHA, and if courts follow Diaz-Roa's logic to hold that the election voids the arbitration agreement broadly, your denominator is not stable.

Claimants who exit arbitration do not just reduce your count. They change the math on every remaining claimant. Your per-claimant completion cost goes up. Your threshold percentage applies to a smaller pool. Your outreach budget, which you sized for 5,000, is now spread across 4,700 or 4,200 or whatever number survives the triage.

This is not hypothetical risk. It is a modeling input you should be running right now.

The Triage Problem Nobody Is Solving at Intake

Most mass arb intake flows are built to answer one question: does this person have a claim that fits the arbitration clause? Eligibility screening, documentation, consent, signing. The workflow is designed to move claimants through a funnel as fast as possible.

What almost no intake flow asks is: does this claimant have a secondary claim that could invoke a statute like EFASASHA? Because if the answer is yes, that claimant is not just a regular member of your pool. They are a variable that could change the procedural posture of the entire matter.

I am not suggesting firms should screen out harassment victims. That would be wrong and probably unethical. What I am suggesting is that firms need to know, before they build their completion model, how many claimants in their pool might have the ability to elect out of arbitration entirely. Because that number changes your budget, your timeline, and your threshold math.

Three Things to Model Differently Starting Now

First, build an EFASASHA flag into your intake. If your claimant pool intersects with employment, gig work, or any context where harassment claims arise, you need a screening question that identifies potential EFASASHA-eligible claimants early. Not to exclude them. To count them accurately in your denominator.

Second, stress-test your completion budget against denominator shrinkage. Run your completion model at 100% of your expected pool, then at 95%, then at 90%. If your per-claimant outreach cost becomes uneconomic at 90% of pool, you have a fragile model. The firms that hit thresholds consistently are the ones that build slack into their denominator assumptions.

Third, talk to your admin about what happens when claimants exit mid-campaign. If your claims administrator charges per touch and you lose 300 claimants from a 5,000-person pool after you have already started outreach, those touches are sunk cost. An admin that charges per fully signed claimant (release and closing statement both executed) absorbs that risk differently than one billing per contact attempt. That is not a minor distinction when your denominator is moving.

The Illinois Uber Ruling Compounds This

The Diaz-Roa ruling does not exist in isolation. The Illinois Supreme Court recently held that a widow's Uber arbitration agreement, including its delegation clause, could not force arbitration of claims arising from her husband's ride. The court found that the scope of the arbitration clause did not extend to a non-signatory's wrongful death claims.

Put these two rulings together and you see a pattern: courts are increasingly willing to narrow the reach of arbitration agreements based on claim type, party identity, and statutory override. For mass arb campaigns, every one of these rulings is a denominator variable. Every claimant who can exit the arbitration framework, whether through EFASASHA, non-signatory arguments, or unconscionability findings, shrinks the pool your completion model is built on.

Completion Is a Denominator Game

Most of the completion conversation in mass arb focuses on the numerator: how many claimants can you get to sign? That is important. But the denominator, the total pool of claimants who will actually proceed through arbitration, is just as critical and far less stable than most firms assume.

The firms that will hit their thresholds in 2027 are the ones that are modeling denominator risk right now. Not just EFASASHA. Not just non-signatory arguments. Every variable that could shrink the pool between filing and resolution.

Build your completion budget for the pool you will actually have, not the pool you started with.