Unreported / Non-Citable
Background
In 2017, Food & Water Watch and a group of co-plaintiffs petitioned the EPA under Section 21 of the Toxic Substances Control Act (TSCA) to ban the practice of adding fluoride to public drinking water on the ground that it presents an unreasonable risk of neurodevelopmental harm. After EPA denied the petition, the plaintiffs sued for de novo review under Section 21(b)(4)(B). The case proceeded through years of litigation, two trials (delayed in part to allow finalization of underlying scientific studies), and ultimately a November 20, 2024 judgment in plaintiffs’ favor — the first time a TSCA citizen group had ever taken a Section 21 petition to a successful federal trial.
Plaintiffs then sought fees and costs as the prevailing party. They asked for roughly $10 million, supported by detailed declarations on hours, hourly-rate methodology (using the Real Rate Report for San Francisco), and a 2.0 lodestar multiplier for lead counsel Michael Connett. EPA opposed both the rate methodology and any multiplier, arguing that the case was “environmental,” that no exceptional result was achieved beyond the statutory minimum, and that no costs should be awarded because plaintiffs had recouped litigation expenses through targeted fundraising.
The Court’s Holding
Judge Edward M. Chen awarded $5,263,705 in fees and $496,745 in costs, for a combined $5,760,450, with payment stayed pending appeal.
On reasonable hourly rates, the court applied In re Prison Legal News v. Schwarzenegger and Stonebrae and held that the relevant comparator market was San Francisco complex civil litigation generally, not just environmental litigation. Section 21 TSCA citizen-suit work — with a full trial, heavy expert use, and novel science — looks much more like complex civil litigation than typical record-review environmental cases.
The court adopted Real Rate Report rates, with a key adjustment for lead counsel Michael Connett: third-quartile (rather than median) rates for his early-career associate years, on the ground that he carried lead-counsel responsibility from the outset, brought distinctive subject-matter expertise from prior fluoride research, and tried the case effectively. The court declined to bracket him at higher experience tiers than he actually had, and once he became partner it used median, not third-quartile, partner rates. Year-by-year, Connett’s rates climbed from $530 (2016) to $1,033 (2025).
The court found a 1.3x multiplier appropriate for Connett’s lodestar — but not the requested 2.0x. Under Perdue v. Kenny A. and Chambers v. Whirlpool, lodestar multipliers are reserved for rare and exceptional circumstances. Plaintiffs’ historic, first-of-its-kind TSCA trial victory and the nearly decade-long delay (including an unexpected two-stage trial) supported some enhancement, but factors such as Connett’s subject-matter expertise and quality of representation were already baked into the third-quartile associate rates and could not be double-counted. The 1.3x multiplier mirrors the modest enhancements approved in earlier TSCA fee-litigation precedent like EDF v. EPA.
On costs, the court rejected EPA’s argument that fundraising-recouped expenses are non-recoverable. Following Pickering v. Holman, the court held the windfall, if any, belongs to the prevailing party rather than the wrongdoer, and that the rule applies even when the third-party source is targeted donor fundraising rather than insurance or an indemnitor. The remaining costs — travel, testifying-expert fees, Westlaw, and transcripts — were reasonable and were awarded in full after plaintiffs voluntarily abandoned a few items in reply.
Key Takeaways
- Even in a TSCA citizen-suit case, San Francisco district judges will use complex-litigation comparator rates rather than narrow environmental-bar rates when fee-shifting calls for “the prevailing rate in the community for similar work.”
- Real Rate Report data continues to drive Northern District lodestar calculations, with third-quartile rates available where individual circumstances (early lead-counsel responsibility plus subject-matter expertise) justify them.
- Multipliers remain rare. Even a historic first-of-its-kind statutory trial win, secured against intense agency opposition, drew only a 1.3x enhancement, applied to one lawyer’s hours.
- Factors built into the lodestar — novelty of issues, complexity, and quality of representation — cannot also drive multiplier enhancement under Perdue.
- Costs recouped through donor fundraising remain recoverable from the losing party. Public-interest plaintiffs do not lose access to statutory cost-shifting just because they raised matching dollars from supporters.
Why It Matters
The underlying case is one of the most consequential environmental-law decisions in years: the first successful TSCA Section 21 citizen-petition trial, brought against a federal agency on a politically charged public-health practice. The fee award has its own significance. Awarding more than $5.7 million in fees and costs against EPA validates the economic feasibility of TSCA citizen suits and signals to the public-interest bar that complex science-driven trials against the federal government can be undertaken without economic ruin.
For California fee-litigation practitioners, the order is also a clean modern application of the post-Perdue lodestar regime. The careful, year-by-year rate construction — including the rationale for third-quartile associate rates with a step-down to median partner rates — is a useful template. So is the rejection of EPA’s recouped-costs argument, which closes off a potential avenue for defendants to use plaintiffs’ fundraising prowess against them.