Three recent decisions from two judges in the Northern District of California provide us with a lot of information regarding where food labeling cases are headed in terms of class certification strategy. Notably, two of the decisions are from Judge Lucy Koh and granted class certification.  Of course, these losses from the defense perspective are disappointing, but it is important to understand them in order to develop effective defense strategies. The third decision is from Judge Charles Breyer and denied class certification. That opinion shows us strategies that work from the defense perspective. In addition, however, the opinion is also interesting because Judge Breyer recognizes the Northern District of California’s status as the epicenter of these cases and because he often expressly acknowledges that his reasoning differs from Judge Koh’s.  So bear with me as we go through three important class certification decisions before identifying some best practices for defendants in these types of claims.

Two Losses for the Defense, but Important Lessons to Learn.

In Werdebaugh v. Blue Diamond Growers, No. 12-CV-2724-LHK (N.D. Cal. May 23, 2014), Judge Koh granted class certification when the plaintiff alleged that he purchased the defendant’s almond milk that allegedly had misleading labels.  That plaintiff contended that using “evaporated cane juice” rather than “sugar” on the label was deceptive and that the phrase “all natural” was deceptive because the products contain potassium citrate.  I won’t address every aspect of this decision but will focus on the ones of most interest to defendants.

That plaintiff had standing to challenge those aspects of the labels because he contended he would not have purchased the almond milk had he known about the misbranding. He contended that the “all natural” label was a substantial reason why he bought the product as well. As to “evaporated cane juice,” Judge Koh distinguished to earlier decisions from the Northern District of California and found that it was plausible for this plaintiff to contend that he did not understand what “evaporated cane juice” meant when he purchased the products. At that point, the court turned to the requirements of Rule 23.

In a minor victory for the defendants, the court refused to certify and injunctive relief class under Rule 23(b)(2). That plaintiff did not allege or even attempt to establish that he intends or wants to purchase these almond milk products in the future. That meant there was no likelihood of future injury to him that could be redressed through injunctive relief.

Turning to the requirements of Rule 23(a), Judge Koh rejected the defendant’s argument that the class was not ascertainable. She concluded that a class is ascertainable so long as it is defined with “objective criteria” and if it is administratively feasible to determine whether a particular member is in the class. Class membership hinged on objective criteria, i.e., purchasing the almond milk products during the class period.  While other courts have held that a class is not ascertainable when there are no corporate records to identify product purchasers or when the purchases are so small that consumers are not likely to have records of them, Judge Koh rejected that approach. She relied on the facts that all purchasers of the almond milk products are within the class definition and all cartons of the products or the alleged mislabeling.  Moreover, “[t]he class period here is also far shorter than in [an earlier case], and inviting plaintiffs to submit affidavits attesting to their belief that they have purchased a carton of Blue Diamond almond milk in the past several years is much likelier to elicit reliable affidavits then asking potential class members to recall whether they had smoked 146,000 of a certain cigarette over the course of several decades.”  [Slip Op. at 20]

As to Rule 23(b)(3), the court refused to certify a nationwide class because it concluded that the law of each consumer’s state of residence would apply to his or her claims. Thus, it limited the class to California consumers.

A substantial issue was whether the plaintiff presented a damages model under California consumer protection statutes that is consistent with his liability case. The plaintiff relied on the testimony of an economist, Dr. Oral Capps, to present three damages models. The court rejected two of them. First, a full refund model was not appropriate because consumers received benefits from the almond milk even if it was mislabeled. Second, the court rejected a price premium model in which Dr. Capps tried to compare the price of the Blue Diamond products to allegedly comparable products that did not have the challenged label statements. He planned to attribute the entire price difference between the Blue Diamond products and the “comparable” products to the labeling. That theory also ignored, however, that the “comparable” product also contained potassium citrate, so it was not an appropriate comparator.  Last, this price premium theory also could not account for any other differences between the Blue Diamond products and the “comparable” products that may lead consumers to pay different prices (e.g., brand loyalty, generic vs. brand name).

The court, however, accepted Dr. Capps’ third theory, a regression model. He contended that he could isolate the relationship between a dependent variable (i.e., the price of Blue Diamond almond milk) and other variables (e.g., the alleged mislabeling). Plaintiff’s expert contended he could control for regional price differences, and Blue Diamond did not introduce any evidence about how regional price differences would affect regional price changes.  And it is that price change that the regression analysis purports to measure. That is, what price movement (if any) occurred after Blue Diamond stopped using the challenged labels? Under this theory, by ostensibly controlling for all other factors that could account for the change, all of the resulting change would be attributable to the labeling.

The court did not require the plaintiff to present a regression analysis that actually works at the class certification stage rather, it only required whether he established a workable model. Indeed, Dr. Capps had yet to run his regression analysis at the time of class certification. While two other courts had excluded Dr. Capps’ testimony in prior cases, Judge Koh found it important that Blue Diamond did not question the tool of regression analysis itself.

The second decision from Judge Koh is Brazil v. Dole Packaged Foods, LLC, No. 12-CV-01831-LHK (N.D. Cal. May 30, 2014).  That plaintiff challenged the “all natural” labeling on 10 packaged fruit products.  Each product contains both ascorbic acid (a naturally occurring form of Vitamin C) and citric acid (a natural preservative derived from citrus).  Coming only a week after the Werdebaugh class certification decision, it is not surprising that the analysis in Brazil is similar.  Once again, Judge Koh rejected the defendant’s ascertainability argument. Because the alleged misrepresentations appeared on the product packaging, there was no concern that the class includes individuals not exposed to the alleged misrepresentations. She also rejected the argument that the lack of company records identifying product purchasers affected ascertainability.  Judge Koh seemed satisfied that every purchaser received the same alleged misrepresentations and the class was limited to only 10 products in a specified timeframe.

The court certified a nationwide injunctive relief class under Rule 23(b)(2), rejecting the notion that the plaintiff’s damages were not incidental to the injunctive or declaratory relief requested. And although this plaintiff said he stopped buying Dole products months ago, he said he would remain willing to buy them now. That sufficed under Judge Koh’s analysis.

As to the Rule 23(b)(3) class, the court again refused to certify a nationwide class based on choice of law issues. These plaintiffs also used Dr. Capps as their damages expert, so the court once again rejected a full refund model and price premium model. The court accepted the regression analysis, however, just as it had in the earlier case. Dole’s position had another wrinkle because it had refused to produce certain economic data, contending that was not needed for class certification. Judge Koh viewed that as a bit of gamesmanship. She would not allow Dole to challenge Dr. Capps’ failure to actually create a working regression model when Dole had refused to provide the economic data he contended he needed for that process.

A Defense Victory Quickly Follows.

Soon after Judge Koh’s class certification decisions, Judge Breyer denied class certification in Jones v. ConAgra Foods, Inc., No. C 12-01633 CRB (N.D. Cal. June 13, 2014).  As he noted, “[t]his district has seen a flood of such cases, in which plaintiffs have challenged, with varying degrees of success, marketing claims on everything from iced tea to nutrition bars. This Order does not—and, given their multiformity, could not—speak to the merits of all such cases.”  [Slip Op. at 1 (footnote omitted)]  So it seems Judge Breyer may not be impressed with Northern District of California’s “food court” moniker.

These plaintiffs challenged a variety of Hunt’s tomato products labeled as “100% Natural” and “free of artificial ingredients & preservatives” because they contained citric acid and/or calcium chloride. They also challenged PAM cooking spray labeled as “100% natural” because that spray contains a propellant that is not natural. Last, they challenged a variety of Swiss Miss cocoa products labeled as being a source of antioxidants.

The opinion nicely analyzes standing and typicality, exploring different plaintiffs’ deposition testimony regarding why they bought certain products, what they read on the labels, what was important to them, and what they admitted was not deceptive. I want to focus on different portions of the decision in this article, however, so I will not discuss standing and typicality in detail.

Judge Breyer uses a more stringent ascertainability standard than Judge Koh. The plaintiffs proposed having class members identify products they purchased with photographs or affidavits/declarations. The court found it infeasible to believe that consumers would recall such purchases accurately, particularly considering that literally dozens of varieties of different can sizes with different ingredients existed during the class period.  While this is a useful conclusion for defendants, plaintiffs will try to distinguish this case by noting that the products’ ingredients and labels varied during the class period.

Judge Breyer continued rejecting class certification arguments with his Rule 23(b)(2) analysis. Each of the named plaintiffs disavowed any intent to buy the challenged products in the future. That was fatal to any requested injunctive relief.

Moving to Rule 23(b)(3) predominance was lacking because of different labeling statements and, more importantly, the plaintiffs’ failure to show uniform understandings regarding the challenged labels. In this instance, the plaintiffs used an expert who opined that the labeling statements were material. That expert, however, did not survey consumers and relied on circular reasoning. That is, she opined that the statements were material because defendants would not have included them on the labels if they were not. That did not suffice for the court. The lack of any established meaning of the word “natural” when used on food labeling truly undermined these claims. Plaintiffs’ expert did them no favors by failing to survey consumers and by admitting that some consumers do not read labels, do not care about labeling statements, and would purchase products regardless of their labels.

In the next collision with Judge Koh’s decisions, Judge Breyer rejected the same Dr. Capps’ regression analysis as Judge Koh had accepted. He believed there simply was no way to control all of the variables to conclude that the price difference was attributable to the labeling. Moreover, Dr. Capps could not identify an adequate compared for product. And, absent an accurate way to identify what each class member paid, no damages analysis could succeed. In essence, Judge Breyer would not accept the notion that Dr. Capps theoretically could create a feasible damages model using regression analysis. Rather, he wanted adequate proof that Dr. Capps had done so at this stage.

How These Decisions Guide Defendants’ Strategy.

Using Experts.

Neither the Werdebaugh nor Brazil decisions granting class certification mentions if the defendants presented survey evidence about consumers’ understandings of “all natural,” either generally or relating to the challenged ingredients specifically (which all are naturally occurring).  These food labeling claims typically rely on California consumer protection laws.  In broad strokes, those laws require that the label be likely to mislead the reasonable consumer.  The difficulty for judges and the parties is that phrases such as “all natural” do not have a set definition.   Class action plaintiffs will use this vagueness to argue that the potential for misleading consumers is a question of fact for the jury to resolve.  Defendants in these cases will benefit from hiring experts to survey consumers to determine their understandings of the phrases at issue.  If substantial percentages of consumers do not share the plaintiff’s understanding of a phrase or disagree with it, it should be untenable to contend that the a cohesive, class-wide understanding of the phrase exists.  The label will not deceive the hypothetical “reasonable consumer.”  Such surveys may also help establish that consumers bought the products for different reasons (e.g., brand loyalty, had a coupon, wanted to try something new).

While parties to these cases are using economists and other damages experts, retail grocery pricing experts/consultants may be useful, too.  Such experts could be outside consultants who advise retailers on pricing strategy or former pricing analysts for retailers.  Such experts may help establish that too many independent variables affect retail pricing for a regression analysis to work.  In addition to factors mentioned in the cases discussed above, several other variables exist.  For example, an individual store may need to reduce the price of a certain product due to inventory control issues during a particular time frame.  Similarly, certain consumers may pay different prices for the same product based on membership in a retailer’s “club card” program.  In fact, those types of discounts may result in the challenged product having a lower price than the comparator product at a specified time.  Those types of discounts may be the motivating factor for consumers as well.  That is, if they can purchase the challenged product at a lower price due to a “club card” membership, they may have done so.  Those types of consumers would not be misled by any labeling.

Another expert to consider is a food scientist.  This type of expert would explain that ingredients such as ascorbic acid and citric acid truly are natural ingredients, whether created in a laboratory or in nature.  There is no chemical difference.  Similarly, she can explain that research confirms the safety of genetically-modified organisms (another frequent target of such “all natural” litigation).  In theory, this type of evidence arguably relates to the merits and not class certification.  But you always want to give your judge the comfort of knowing that she is not allowing truly bad conduct to go without a remedy if she refuses to certify the class.

Disproving Your Plaintiffs’ Allegations.

It also is important to evaluate your plaintiff’s purchasing habits and history.  As we saw in Jones, those defendants effectively used testimony from the plaintiffs to defeat class certification.  Explore whether they buy other products with similar ingredients and whether they have continued to buy the challenged product even after filing suit.  See if your plaintiff will admit that the labeling statement wasn’t important to her.

Many grocery retailers offer membership programs to customers.  Typically, these programs track purchasing history at an individual level so the retailer may direct specific advertising and offers to that customer.  Online retailers such as Amazon likewise track customers’ order history.  Try to obtain those data to determine if your named plaintiff buys other products with ascorbic acid or citric acid despite now avowing that he tries to avoid those ingredients.  Does your plaintiff who now avows a passion for healthful food habitually purchase sugary beverages, processed snacks, or other junk food?  Such purchasing history also may have data regarding the prices that plaintiff paid for the challenged product.  Those data may show the dramatically varying prices from week to week as well as discounts that a consumer enjoyed because of membership in that type of program.

Food labeling class actions continue to be a thriving business for plaintiffs’ counsel.  Mounting strong defenses not only helps your client end the current case but also sends a message to the plaintiffs’ bar that bringing suit against your client will not be a good investment for them.

James Smith is a partner in the Phoenix office of Bryan Cave LLP.  He is a member of the Class & Derivative Actions Client Service Group and of the Food and Beverage Team.