Practical Tips · The Demer Factors · Factor Two · June 2026
An expert’s bias shows up as a pattern, and the pattern lives in other insureds’ claim files. California has let claimants reach those files for forty years — and courts grant the discovery far more often than the industry lets on.
A bias case turns on four factors — the expert’s compensation and assignment volume, the pattern of insurer-favorable outcomes across claims, the methodology and procedural irregularities in the work, and the insurer’s own measures (or absence of measures) to secure a neutral expert. It is tempting to think most of that can be built from the file in front of you. It cannot. The single report you hold may hint that the methodology is thin or result-driven, but the proof that the thinness is a practice — together with the compensation figures, the assignment counts, the denial rates, and the repetition of the same flawed reasoning — lives in records you do not possess: the other insureds’ claim files and the insurer’s and expert’s own cabinets, reached only through discovery and subpoenas. The good news is that you are allowed to open them.
Taken one at a time, every denial backed by a hired opinion has an innocent explanation. The disability claim really was a pre-existing condition, says the insurer’s on-staff physician. The surgery really was experimental, says its reviewing doctor. The foundation cracks really came from old soil settlement, not the earthquake, says its engineer. The kitchen leak really had run for more than two weeks, says its field plumber. Each is plausible in isolation. What dissolves the innocence is repetition — and repetition of two kinds. The same reviewer, retained by the same insurer, reaches the same insurer-favorable answer across a run of other claims (the outcome pattern); and reaches it the same way each time, leaning on the same experiential say-so, ignoring the same kind of contrary evidence, over-weighting the same flawed method (the reasoning pattern). The files prove not only that the denials recur but why — and the why is often the more damning. One opinion is a judgment call; a pattern is a practice. And the pattern is structurally out of reach without discovery, because the claimant sees one file and the insurer sees all of them. Discovery of other insureds’ claim files — OICFs — is how the law corrects that asymmetry.
Asking for strangers’ claim files is not a reach. In California the right runs back more than forty years, to Colonial Life & Accident Insurance Co. v. Superior Court (1982), where the Supreme Court held that evidence of an insurer’s pattern of unfair practices is relevant to a bad-faith claim and to the recidivist conduct that supports punitive damages, and that a claimant may reach other insureds’ files to develop it. It remains the seminal authority and remains good law: although California later held there is no private right of action under the unfair-practices statute, the courts have since confirmed that statutory violations may still be used to prove common-law bad faith. The cabinet the Court opened in 1982 has never been closed.
OICF discovery is not a single-purpose tool. A segmentation of the cross-claimant records identifies eleven distinct evidentiary uses they serve — and the more of them a request can name, the harder it is to dismiss as a fishing trip:
| Selection bias | That the carrier chose this reviewer because the reviewer’s record made the outcome predictable. |
|---|---|
| Compensation & financial dependence | The volume of claims referred and the dollars paid by the carrier or its vendors — the burden-shift predicate. |
| Pattern and practice | Systemic conduct across many claimants — the “general business practice” element. |
| Outcome & denial-rate | The reviewer’s denial rate as a direct measure of one-sided output. |
| Methodology & reliability | That the reviewer’s experiential judgment is not applied consistently across like claims — selective use of the reviewer’s own criteria, the procedural irregularities behind the opinions. |
| Policy interpretation | That the carrier construes the same policy language one way for some insureds and another way for others. |
| Genuine-dispute rebuttal | Defeating the “fairly debatable” defense by showing a recurring litigation posture, not a one-off honest reading. |
| Insurer intent | Conscious disregard — the bad-faith mental state shown by repeated deployment of the same reviewer with knowledge of the record. |
| Punitive damages | The recidivist, reprehensible pattern that supports a punitive award. |
| Statutory & UIPA violations | The unfair-claims-practices showing under section 790.03(h) and the parallel showing under the Unfair Competition Law. |
| Class actions | Commonality and the substantiation of class allegations. |
Relevance is the real discipline, and it is satisfiable: the requirement is substantial similarity — a meaningful nexus such as the same adjuster, expert, exclusion, denial rationale, or kind of loss. A request tied to the biased participant and bounded in time, geography, and scope meets it; “every file from this disaster” does not. Burden collapses more often than insurers admit, because it is frequently self-inflicted — courts have refused to let a carrier defeat discovery by pointing to its own opaque recordkeeping, since a company that cannot search its own claims “is entitled to little sympathy” for that choice. Privacy is the objection most overstated: the Insurance Information and Privacy Protection Act contains express exceptions for disclosures under a judicial order and disclosures otherwise permitted by law, and the privacy interest is ordinarily answered by redaction — you usually need not contact a single other insured to use what their file teaches.
By my own survey of the California orders, courts grant OICF discovery far more often than they deny it — roughly seven in ten times — and most denials turn on a fixable framing mistake, not on any principle that the files cannot be had. And the granting does more than open a cabinet. Courts do not order discovery to prove something the law refuses to recognize; every order letting a claimant develop the pattern of insurer-favorable outcomes is a court affirming, by its conduct, that an inference of bias drawn from a pattern is real and cognizable. The discovery and the standard it serves are the same argument, made twice.
Read the deep-dive → See the case study → The bias-evaluation service →
Turn the principle into a request with Framing an OICF Request That Survives, see the most-cited denial dissected in The Case Insurers Cite Most to Deny Discovery, and the standard the pattern serves in The Standard Comes Before the Factors.
Distilled from the project’s own survey of California and national OICF-discovery orders. The seminal authority is Colonial Life & Accident Ins. Co. v. Superior Court, 31 Cal.3d 785 (1982). The full body of granting and denying orders — and the failing behind each denial — is reserved for the subscriber deep-dive. Educational and informational only; not legal advice.