Justice in a Locked Room: How Mandatory Arbitration Erases the Paper Trail on Workplace Abuse
When a warehouse worker in the Inland Empire files a complaint about unpaid overtime, or a meatpacking employee in the Midwest reports that supervisors ignored repeated safety violations, the instinct is to imagine that complaint entering some form of official record—a court filing, a regulatory database, a public docket. For the majority of American workers, that assumption is wrong. Their grievance will instead travel into a private arbitration system designed, structurally and legally, to ensure that what happens there stays there.
Mandatory pre-dispute arbitration clauses—standard language buried in employment agreements, offer letters, and onboarding paperwork—now govern the workplace disputes of an estimated 60 million American workers, according to research from the Economic Policy Institute. These clauses do not merely redirect disputes away from courts. Paired with confidentiality agreements that are themselves often mandatory, they ensure that the facts of individual cases, the names of employers, the nature of violations, and the outcomes reached are sealed from public view. The data simply ceases to exist.
A Parallel Legal Universe Without a Public Record
Civil litigation, whatever its limitations, generates records. Court filings are indexed. Judgments are published. Class action settlements require judicial approval and public notice. This transparency is not incidental—it is foundational to the ability of researchers, journalists, regulators, and advocacy organizations to identify patterns of abuse across industries and employers.
Arbitration produces none of that infrastructure. Cases are administered by private firms—the American Arbitration Association, JAMS, and a constellation of smaller providers—under procedural rules that typically prohibit parties from disclosing the existence of proceedings, let alone their substance. When a decision is reached, it is delivered privately to the parties. No appellate court reviews it for consistency. No database catalogs it alongside similar cases. No regulator is automatically notified.
For human rights data purposes, this is not a gap in the record. It is the deliberate elimination of the record.
The Aggregation Problem: Why Individual Silence Becomes Systemic Blindness
The human rights implications of this architecture extend well beyond individual workers who lose access to courts. The deeper problem is one of aggregation—or rather, the systematic prevention of it.
When workplace violations are adjudicated publicly, patterns emerge. A researcher examining federal court filings might identify that a particular staffing agency generates a disproportionate number of wage theft claims. An investigative journalist reviewing OSHA citations alongside civil suits might notice that a specific food processing company's injury rates cluster around particular facilities. A policy advocate might aggregate discrimination complaints to demonstrate that a sector-wide problem demands regulatory intervention rather than case-by-case resolution.
Mandatory arbitration forecloses all of this. Each case is atomized, sealed, and effectively deleted from the evidentiary universe. The worker who prevails in arbitration—a relatively rare outcome, given that research consistently shows employees win at substantially lower rates in arbitration than in court—receives a confidential settlement and moves on. The employer faces no public accountability, no reputational consequence, and no contribution to any data set that might one day illuminate the scope of the problem.
This is not a side effect of the system. It is, for the employers who lobbied aggressively for the legal architecture that enables it, the primary feature.
The Regulatory Vacuum That Confidentiality Creates
Federal and state labor regulators operate, in theory, independently of the civil litigation system. The Department of Labor's Wage and Hour Division, the Equal Employment Opportunity Commission, and state-level equivalents can investigate and enforce regardless of whether a worker has filed a private lawsuit. In practice, however, these agencies depend heavily on complaint data and litigation outcomes to identify enforcement priorities, allocate investigative resources, and build systemic cases.
When arbitration absorbs the bulk of workplace disputes, regulators lose a critical intelligence stream. They cannot see what they cannot count. An employer who settles dozens of discrimination claims in confidential arbitration proceedings appears, to a regulator scanning public data, to have a clean record. An industry whose wage theft practices are addressed exclusively through private arbitration generates no public pattern that might trigger a sector-wide investigation.
The EEOC has explicitly acknowledged this problem, noting in policy statements that confidential settlements impede its ability to identify systemic discrimination. Yet the agency lacks authority to pierce arbitration confidentiality agreements in most circumstances, and legislative efforts to address the issue have stalled repeatedly in Congress—with the notable exception of the 2022 Ending Forced Arbitration of Sexual Assault and Sexual Harassment Act, which carved out those specific claim types from mandatory arbitration enforcement.
That carve-out, while meaningful, illustrates the broader problem by contrast. Wage theft, racial discrimination, unsafe working conditions, retaliation against whistleblowers—all remain fully subject to mandatory arbitration and its accompanying confidentiality regime.
What the Data We Do Have Reveals
The irony of researching mandatory arbitration's impact on data availability is that it requires working with the limited data that has managed to surface despite the system's opacity. Several sources offer partial windows.
The American Arbitration Association publishes aggregate consumer and employment arbitration statistics, though these figures are stripped of identifying information and provide no granular breakdown by employer, industry, or violation type. Academic researchers who have obtained case-level data through litigation discovery or regulatory access have consistently found that employees fare significantly worse in arbitration than in court—winning less often, receiving smaller awards, and facing procedural disadvantages that compound over time.
Research by Cornell University's Alexander Colvin, among others, has documented that the "repeat player" effect—whereby employers who frequently use arbitration develop structural advantages over first-time employee claimants—systematically tilts outcomes. Arbitrators who depend on employer referrals for their caseload have measurable financial incentives to rule in employers' favor, a conflict of interest that the confidential, unreviewed nature of proceedings helps conceal.
Worker centers, legal aid organizations, and community-based advocates who work directly with low-wage workers report that mandatory arbitration functions in practice as a near-complete barrier for workers in sectors like domestic work, agriculture, food service, and retail—industries that disproportionately employ immigrants, women, and workers of color.
The Advocacy and Research Imperative
For organizations working at the intersection of labor rights and human rights data, mandatory arbitration represents a foundational methodological challenge. Any effort to measure the prevalence of workplace abuse in the United States that relies on litigation records, regulatory complaints, or court-published outcomes is working with a dataset that has been systematically truncated by design.
This demands a multi-pronged research response. Community-based surveys and worker testimony projects, while resource-intensive, can capture violations that the formal record erases. Investigative partnerships between advocacy organizations and journalists have produced some of the most significant exposés of arbitration abuse. Legislative advocacy for transparency requirements—mandating that arbitration providers publish case-level data disaggregated by employer and claim type—represents perhaps the most direct structural remedy available.
The 2022 arbitration reform legislation demonstrated that targeted congressional action is possible when political conditions align. Expanding its scope to cover wage, discrimination, and safety claims would represent a significant step toward restoring the evidentiary infrastructure that accountability requires.
Until then, the locked room remains locked—and the violations conducted inside it remain, by design, uncounted.
