Practical Tips · Case Study · July 2026
The plaintiffs in Brzezinski v. Allstate pled a selection-bias theory, then subpoenaed the examiner’s income instead of his outcomes, argued “credibility,” and won only a sliver of the discovery. A losing record that is a blueprint for winning.
Some of the most instructive discovery losses are the ones where the law was on the claimant’s side. Brzezinski v. Allstate Insurance Co. (S.D. Cal. 2012) is one. The plaintiffs had a real bad-faith theory and a court willing to find their discovery relevant — and they still walked away with a fraction of what they asked for, because the request did not match the theory, the framing invited the wrong rule, and the sequence handed the other side its best objection.
The Brzezinskis sued Allstate for bad faith over an underinsured-motorist claim, alleging Allstate unreasonably delayed and refused to use any examiner other than one doctor because of its cozy, repeat relationship with him. They subpoenaed the doctor — a non-party — for his tax forms, income percentages, and claim volume, plus a deposition. The court found the information relevant to the bad-faith claim, then granted only deposition estimates (percentages, hourly rate, evaluations per week, the fee for treating the plaintiff) and denied the tax forms and documents. On the numbers that mattered, close to a total loss on strong bones.
Their theory was a selection theory — that Allstate steered the claim to a favored examiner. Proving a selection theory generally takes two things: the insurer’s pattern of choosing this examiner over others, and the examiner’s outcome record — how often his opinions favored insurers across other claims, the pattern-and-practice files. They asked for neither. They asked for financial dependence — the income numbers — and the examiner’s report had actually come out in the plaintiff’s favor, gutting the financial angle for this report. Even a full win on that motion would not have carried the case: financial-dependence proof alone rarely substitutes for the outcome-pattern evidence a selection theory needs. The request chased the wrong factor for the theory on file.
The plaintiffs told the court they wanted the information to attack the doctor’s credibility. That word is a doctrinal trapdoor: credibility is the impeachment frame, and it let the court reach for a line of older impeachment cases — about testing a trial witness’s bias — that allow deposition questioning but refuse the financial documents. Anchored instead in Allstate’s duty to investigate fairly under Egan v. Mutual of Omaha — or, outside first-party claims, the broader implied covenant of good faith and fair dealing — the money would have been breach evidence, and the impeachment cases would not have fit.
Going straight at the non-party doctor for documents left the door open to the insurer’s favorite move — the argument that financial information should come through a less intrusive means, a deposition. Serving interrogatories on the compensation and claim counts at the same time as the document requests forecloses that argument: once the insurer refuses to answer them, it cannot claim the claimant should have used the very route it stonewalled. The plaintiffs served none.
The fight was waged entirely on constitutional financial privacy. But insurers raise a second objection — California’s insurance-privacy statute (the IIPPA) — and that statute carries its own exceptions: disclosure pursuant to a subpoena or court order, and disclosure otherwise permitted by law. The plaintiffs never engaged the statute or its exceptions, leaving half the privacy terrain uncontested.
Brzezinski is a warning, not a wall. Match the discovery to the theory — a selection theory needs the outcome files, not just the income. Anchor the request in the insurer’s duty, never in “credibility.” Propound the interrogatories alongside the document requests. And answer both privacy objections, statutory and constitutional. Correct those four, and the same relevance finding that yielded only estimates yields the documents.
Read the ground-by-ground teardown → See the checklist → The bias-evaluation service →
See the issue this case study serves in The First Proof of Expert Bias Is in the Checkbook, and the moves that avoid these mistakes in Ten Moves Before You Ask What the Expert Was Paid.
Drawn from Brzezinski v. Allstate Ins. Co., No. 11-cv-2373-CAB (DHB), 2012 WL 12869522 (S.D. Cal. Aug. 28, 2012), read in full. Supporting authority — Egan v. Mutual of Omaha Ins. Co., 24 Cal.3d 809 (1979); Eastman v. Allstate Ins. Co. (IIPPA exceptions); Demer v. IBM Corp. LTD Plan, 835 F.3d 893 (9th Cir. 2016). The impeachment cases and full citations are in the subscriber deep-dive. Educational and informational only; not legal advice.