By Dennis Raglin, Shareholder, Carlton Fields
Key takeaways:
- Most enforcers calculate exposure as daily consumption of the maximum serving size on the label, inflating violation claims, and the settlements based on them.
- The law instead requires exposure to be calculated using “typical consumer” and average consumption data, not a full daily serving.
- Defendants recently won Prop 65 food trials by rejecting plaintiffs’ inflated consumption assumptions, using the averaging method from the seminal Beech-Nut case. Averaging remains a strong defense: two trials are pending or underway, and multiple settlements allowing averaging have won court approval.
I have been involved in three California Proposition 65 chemical failure-to-warn food cases in recent years, and all three resulted in decisions for the defense. In each, the court agreed that exposure in these cases is decided by calculating the average consumer’s consumption of the food. These courts rejected the position plaintiffs’ lawyers routinely take to force settlements: that “average” means a full serving size as shown on the label, eaten every day, for the rest of the consumer’s life.
The regulations
Proposition 65 liability turns on exposure to the “typical user.” The concept sounds simple, but the regulatory language is dense. Here’s what governs exposure to chemicals listed as causing reproductive harm or birth defects:
“(b) The level of exposure to a chemical listed as causing reproductive toxicity shall be determined by multiplying the level in question (stated in terms of a concentration of a chemical in a given medium) times the reasonably anticipated rate of exposure for an individual to a given medium. The reasonably anticipated rate of exposure shall be based on the pattern and duration of exposure that is relevant to the reproductive effect which provided the basis for the determination that a chemical is known to the state to cause reproductive toxicity.
Clear as mud? Then:
“(C)(2) For exposures to consumer products, the level of exposure shall be calculated using the reasonably anticipated rate of intake or exposure for average users of the consumer product, and not on a per capita basis for the general population. The rate of intake or exposure shall be based on data for use of a general category or categories of consumer products, such as the United States Department of Agriculture Home Economic Research Report, Foods Commonly Eaten by Individuals: Amount Per Day and Per Eating Occasion, where such data are available.”
In plain English: measuring chemical exposure from a product or food must account for averages — the average user, and his or her average use.
The anticipated rate of exposure for average users
Whether the product is a plate, a wrench, or a food like protein powder or rice, a Prop 65 warning isn’t triggered by the amount of chemical in the product alone. It also depends on the reasonably anticipated rate of exposure for the average consumer’s use. For a wrench, that means asking how often the average person uses it, how they hold it (all five fingers? does it touch the palm?), and for how long, since Prop 65 covers three routes of exposure — dermal, oral, and inhalation. A wrench is mostly a dermal exposure, but hands also touch the mouth, so oral exposure matters too.
Courts require admissible evidence for these numbers: government data such as Labor Department statistics, or a survey commissioned from a reputable source to question people on their typical handling and use of the product. That “how and how long” data is multiplied by the chemical concentration to calculate exposure, which is measured in micrograms, not the parts-per-million or -billion used to measure concentration in the product itself.
So how do we calculate exposure for food?
The same concept applies to food. Take a case I’m working on involving lentil soup. Does the average person eat a full serving (a bowl) of lentil soup every day, forever? We are not talking about your Uncle Morty, who can’t get enough of the stuff. Just as we care about the average user of a wrench and not a commercial user who handles one eight hours a day, with food we care about how frequently an average consumer eats it, and how much he eats when he does. We take that “how often and how much” data and multiply it by the amount of the chemical detected in the soup.
Unlike Uncle Morty, most people don’t eat a bowl of soup daily, but treating “every day” as normal is exactly what plaintiffs’ lawyers do. Instead of looking at real consumption, they look at the label’s serving size, multiply it by the chemical’s concentration, and call that the only acceptable exposure figure. For example: if a soup tests at .02 parts per million lead with a 35-gram serving, they calculate exposure as .02 x 35 = 0.7 micrograms — over the 0.5 microgram Prop 65 safe harbor for lead, triggering a warning requirement.
NHANES and real-world consumption
That maximum-serving, every-day calculation isn’t the only option — the regulations reject it. Is there a trustworthy source for the average consumer’s actual food consumption? Yes: the regulation requires using “data for use of a general category” where such data is available, and for food consumption the required source is the CDC’s National Health and Nutrition Examination Survey (NHANES). NHANES has been around since before disco (1971) and is the best available data on what typical Americans eat, how often, and how much, drawn from thousands of responses that are gathered, adjusted, and properly weighted.
Take the lentil soup again. NHANES reports that in winter, 40.7% of respondents ate soup more than four times a month, while 27.3% ate it less than once a month; across all seasons, those figures were 18.6% and 39%, respectively. That’s a very different picture than “a bowl a day.” Plaintiffs’ lawyers say none of this matters, that we should ignore a 55-year-old survey conducted by trained professionals in favor of the full-serving-daily assumption. Why? This may surprise you: money. The plaintiffs’ calculation almost always produces an exposure level above the safe harbor, which in turn forces a settlement. Using their method, the Uncle Mortys of the world, who eat lentil soup on the Fourth of July, all summer, and Labor Day, become the “average” consumer. That’s not what voters intended when they passed Proposition 65 in 1986.
Defendants have succeeded in food cases using averaging
Defendants have fought back successfully against the artificial full-serving, daily-consumption assumption. Their main weapon is Environmental Law Foundation v. Beech-Nut Nutrition Corporation (2015) 235 Cal.App.4th 307, as modified on denial of reh’g (Apr. 16, 2015), involving lead in baby food and juice. Beech-Nut is the seminal case endorsing the use of NHANES average consumption data in Prop 65 food cases, along with averaging using the geometric mean, which excludes outlier results and focuses on the average user.
Recent trials have endorsed beech-nut averaging
Judges in three Prop 65 food cases decided between 2023 and 2026 followed Beech-Nut, finding NHANES the best data source for food consumption and approving averaging using the geometric mean: Consumer Advocacy Group, Inc. v. Gulf Pacific Rice Co. (Cal. Super. Ct. 2024, No. BC553427); Consumer Advocacy Group, Inc. v. Badia Spices, Inc. (Cal. Super. Ct. 2024, No. 19STCV38610); and Consumer Advocacy Group, Inc. v. Gel Spice Co. Inc. (Cal. Super. Ct. 2024, No. BC665798). All three decisions are on appeal. A fourth case was dismissed before reaching a court decision, but the rulings issued along the way aligned with the other three, rejecting plaintiffs’ proposed method of calculating exposure.
Though the regulation mandates NHANES data “where available,” it isn’t the only source that can be used. But any alternative has to be shown superior to NHANES, and no court has found one. In these cases, plaintiffs claimed the calculation should be based not on the average consumer but on the assumed higher consumption of certain ethnicities. In one case, plaintiffs claimed Asians eat more rice, so exposure should be calculated based only on their habits, not those of an average of all consumers. Notwithstanding the rank stereotyping, none of the courts found this subpopulation approach a superior way to calculate exposure — a critical win not just for these defendants, but for the food industry in general.
Subsequent trials and settlements continue to show courts agree with averaging
Beech-Nut remains good law, though the state Attorney General’s office isn’t happy about it. The AG’s office routinely objects to proposed settlements based on averaging, arguing it permits exposure on a single day that’s higher than the safe harbor. In Blue Water Consaint, LLC v. Bumble Bee Foods, LLC, San Diego County Superior Court, Case No. 24CU003388C, involving mercury in fish steaks, the AG appealed a consent judgment on those grounds, but limited the appeal to whether the settlement served the public benefit.
Most recently, a court ruled for the defendant in a case involving cadmium in spinach: The Chemical Toxin Working Group v. The Pictsweet Company, Los Angeles Superior Court Case No. 24STCV08902. As in the other cases, the judge found exposure should be based on NHANES averaging. This was the first food trial involving cadmium, a chemical increasingly targeted by enforcers, particularly in seafood. Plaintiffs argued Beech-Nut averaging shouldn’t apply outside of lead cases; the judge disagreed, a result that should give defendants more confidence in cadmium cases going forward.
Notably, the Pictsweet court found that OEHHA itself listed cadmium as a Prop 65 chemical based on science that averaged exposure over time, not a single day. Plaintiffs had planned to argue, based on post-listing studies, that cadmium is a teratogen requiring single-day exposure measurement. The court noted OEHHA’s own data showed sperm count and motility effects only after long-term exposure. And while OEHHA didn’t define “long term,” it clearly wasn’t a single day.
A fifth food trial is on the horizon, involving Korean foods: Consumer Advocacy Group, Inc. v. Korean Farm, Inc., Alameda County No. 22CV018811, the first food trial outside Los Angeles. Will NHANES go 5-0?
Dennis Raglin brings more than 30 years of experience counseling and defending clients regarding California’s Proposition 65 right-to-know chemical warnings law. He advises clients on a broad spectrum of state consumer chemical regulations, including PFAS, recycling, packaging, producer responsibility, tools, cosmetics, food, and personal care products.
