The Foundation Series · Part 5 of 5 · July 2026

ERISA and Non-ERISA: One Harm, Different Remedies

The same bought expert, the same rigged denial — and, depending only on where your coverage came from, a heavy price or almost none.

Parts 1 through 4 were about state common-law insurance, where the special relationship turns the covenant of good faith into a tort-backed duty and a policyholder denied in bad faith can recover far more than the benefit itself. But a very large share of Americans get their health and disability coverage through their jobs — and employer-provided benefits are not governed by state insurance law at all. They are governed by a federal statute, the Employee Retirement Income Security Act of 1974 (ERISA), built on a different foundation and offering a different, and far thinner, set of remedies. Part 5 is about the identical underlying problem in both systems: a financially interested decisionmaker, leaning on financially interested experts, to deny a claim.

Two roads lead to the same distrust. State law reaches the insurer’s duties through the special relationship — Wilson v. 21st Century is the modern statement that a “genuine dispute” cannot be manufactured out of a conclusion the insurer has no basis to hold. ERISA reaches a parallel place through trust law: the entity that decides your claim is a fiduciary, and when it both decides claims and pays them from its own pocket, that structural conflict is a factor the reviewing court must weigh (Metropolitan Life Ins. Co. v. Glenn, 2008), while a plan “may not arbitrarily refuse to credit a claimant’s reliable evidence” (Nord). Look past the vocabulary — “special relationship” and “bias” on one side, “fiduciary” and “conflict of interest” on the other — and both regimes answer the same question the same way. The convergence is most complete on the expert retained to evaluate the claim: the framework the Ninth Circuit named in Demer v. IBM (2016) — compensation, volume of assignments, a pattern of insurer-favorable outcomes, ripening into a rebuttable presumption of bias — is regime-neutral because bias is. The federal regulator went further than the states and wrote the inquiry into law: hiring, compensating, or retaining an expert “must not be made based upon the likelihood that the individual will support the denial of benefits.”

If the wrong is identical, you would expect the consequences to match. They do not — and the reason is an accident of history, not a judgment about insurance. ERISA is a pension-protection statute, born of the 1963 Studebaker collapse, whose remedies run to the plan, not the person; its definitions swept in the employer health and disability coverage it was never designed to govern. Massachusetts Mutual Life Ins. Co. v. Russell read the statute’s remedies as a closed, “comprehensive and reticulated” list; Pilot Life Ins. Co. v. Dedeaux held the state bad-faith tort preempted and unsaved; Aetna Health Inc. v. Davila barred any state remedy that “duplicates, supplements, or supplants” the federal scheme — leaving what Justice Ginsburg called a “regulatory vacuum.” The result is stark: the same misconduct carries a deterrent price under a non-ERISA policy and almost none under an ERISA plan. Which is exactly why the bias inquiry — the one lever that survives the crossing into federal territory — matters most where the claimant has the least.

Part 5, in two editions

Part 5 publishes in two companion editions on Expert Bias Report. The free edition states the convergence and the asymmetry; the paid edition works through the ERISA case law, the Demer framework, the federal regulation, and the statutory architecture that explains the remedial divide.

Free edition · The convergence and the asymmetry

ERISA and Non-ERISA: One Harm, Different Remedies

Two regimes — one built on trust law, one on contract — arrive at the same suspicion of the expert retained by the payer, and two regulators converge on one duty: the expert must be chosen for the reliability of the judgment, not the reliability of the result. Then the asymmetry, traced to its source: a pension statute stretched over insurance claims, its sweeping preemption clause clearing away the state-law protections with nothing built in their place. The industry priced the vacuum in its own words — a 1995 memorandum tallying twelve claims settled for $7.8 million that ERISA would have reduced to “between zero and $0.5 million.”

Read the free edition →

Paid edition · The case law & the architecture

The Same Harm in Two Regimes: ERISA, Non-ERISA, and the Hired Gun

The structural parallel in full — Firestone deference, Glenn conflict-weighting, Abatie procedural scrutiny, Nord — set beside the state-law duties, then the Demer framework read closely and carried across the ERISA line in both directions. The Department of Labor’s impartiality rule, quoted whole. And the root: a pension-protection statute, an exclusive-remedy design (Russell), and the most sweeping preemption clause Congress has written (Pilot Life, Davila) — the architecture that explains why the two systems agree almost entirely on what is wrong and disagree almost entirely on what to do about it.

Read the paid edition →

Where this fits. Part 5 closes the Foundation Series by carrying its conclusion into federal territory: the biased-expert problem does not change when the statutory label on the policy changes, because a medical examiner’s reliability is a fact about the examiner — the money, the volume, the pattern — not about the regime. What changes is the price the law puts on the wrong, and that divergence is the residue of a statute built for a different purpose. The companion Recap gathers the whole argument into a single reference. As throughout this site, the ideas and the seminal authority are free; the implementing corpus is the paid product.

The series

  • Part 1 — The covenant in every contract. What the law means by an implied promise of good faith — what it protects, and against what.
  • Part 2 — What “good faith” means. The Summers–Burton debate and the practical answer California worked out that the academy never reached.
  • Part 3 — Why insurance is different. How the same covenant becomes, in the insurance relationship, the source of a tort-backed duty.
  • Part 4 — The attendant duties. Good faith in claims handling resolves into concrete obligations, each of which condemns the biased expert.
  • Part 5 — ERISA and non-ERISA. The same bias problem under federal benefits law; two regimes, the same suspicion of the interested decisionmaker, sharply different remedies. (This page.)
  • Recap — What your insurer owes you. The whole argument as one checklist — every duty, the case that anchors it, and whether the protection survives into ERISA.

Subscribe to follow the series →

Related

Part 5 develops the argument first made in Demer’s Paradigm for Assessing Biased Insurance Experts (Advocate Magazine, 2024) — that the four-factor bias framework applies with equal force whether a claim arises under ERISA or ordinary state-law insurance — and operationalized across the Practical Tips.

This page summarizes Part 5 of The Foundation Series as published on Expert Bias Report. Holdings and quotations derive from the project’s primary reading of the opinions — Metropolitan Life Ins. Co. v. Glenn (2008), Black & Decker Disability Plan v. Nord (2003), Demer v. IBM Corp. LTD Plan (9th Cir. 2016), Massachusetts Mutual Life Ins. Co. v. Russell (1985), Pilot Life Ins. Co. v. Dedeaux (1987), and Aetna Health Inc. v. Davila (2004) — and from the project’s analysis in treatise/12-regulatory/12-04-erisa-regulations.md (29 C.F.R. § 2560.503-1(b)(7)). The Studebaker history and the “worse off” assessment follow Stumpff (2011); the 1995 UnumProvident memorandum, Langbein (2007). Educational and informational only; not legal advice.