Opinion handed down May 1, 2015
Employees of Gilster-Mary Lee Corporation (“Gilster”) filed a class action lawsuit in the Circuit Court of Jasper County against their employer alleging that workplace exposure to butter flavoring products caused lung impairment.[1] Gilster removed the case to federal court under the Class Action Fairness Act (“CAFA”).[2] However, the U.S. District Court for the Western District of Missouri ordered a remand to the state trial court, noting CAFA’s local controversy exception applied.[3] The exception requires the district court to decline jurisdiction when more than two-thirds of the members of the plaintiff class are citizens of the state where the action was originally filed.[4] Gilster appealed the district court’s decision, contending that the plaintiff’s class did not sufficiently establish two-third Missouri citizenship because the district court’s calculation included potentially unreliable last known addresses of unresponsive members.[5] The U.S. Court of Appeals for the Eighth Circuit reversed the district court’s holding and remanded the case to reevaluate the number of Missouri citizens in the class, noting that the last-known address evidence was insufficient in establishing the local controversy exception.[6]
I. Facts and Holding
Former and current packaging plant employees of Gilster filed a class action lawsuit against the corporation, claiming they suffered or would suffer lung impairment caused by an exposure to natural and artificial butter flavoring products and ingredients present at the workplace.[7] The asserted class included anyone who worked at the Gilster plant in Jasper, Missouri, prior to January 1, 2008 – when Gilster stopped using an especially harmful butter flavoring product.[8] This pre-2008 class was comprised of approximately 372 potential class members characterized by four different groups: 40 current employees, 246 former employees, 61 temporary staff workers, and 25 prior litigants.[9] Attempting to avoid removal to federal court under CAFA, employees sought evidence that two-thirds of its class were Missouri citizens, which would qualify the suit for state court under the local controversy exception to CAFA.[10] However, when the employees requested affidavits confirming Missouri citizenship from the class, most did not respond.[11] Only 41% of the potential class members were clearly identified as Missouri citizens by their response to the affidavits.[12]
The employees stressed that 126 of the non-responsive former employees had last-known addresses in Missouri and that their Missouri citizenship should be presumed.[13] Conversely, Gilster argued that these last known addresses were not reliable because they were outdated – some as old as 27 years old – and did not identify U.S. citizenship.[14] The U.S. Court of Appeals for the Eighth Circuit reversed the district court’s decision that the two-thirds class citizenship requirement was established, holding that last-known addresses did not sufficiently establish current Missouri citizenship necessary for the local controversy exception.[15]
II. Legal Background
CAFA requires district court jurisdiction for all civil actions exceeding $5,000,000 in damages when any member of the class of plaintiffs is a citizen of a state different than the defendant.[16] However, there are exceptions: a district court must decline jurisdiction when more than “two-thirds of the members of all proposed plaintiff classes . . . are citizens of the state in which the action was originally filed.”[17] Determining this two-thirds citizenship can be slightly difficult considering the large classes and frequently changed addresses. It is established that the party seeking remand bears the burden of proof for the CAFA exception and any doubt about applying the exception is resolved against the party seeking remand.[18]
In In re Sprint Nextel Corp., the U.S. Court of Appeals for the Seventh Circuit held that consumer phone numbers and addresses were insufficient to establish citizenship necessary for the local controversy exception.[19] In Sprint, cell phone consumers alleged that Sprint conspired to impose artificially high text message service prices.[20] Consumers argued that the class’s definition as “all Kansas residents” who purchased text messaging through Sprint and its sole use of Kansas cell phone numbers and mailing addresses were sufficient proof that two-thirds of the proposed class members were Kansas citizens.[21] Though the court noted the apparent logic in this conclusion, it stressed that the conclusion was still merely conjecture as it relied on possibly inaccurate assumptions about citizenship.[22] Further, the court noted the class definition of “all Kansas residents” was insufficiently narrow because residents are not always citizens of the state.[23] Conclusively, the court held that inferences about citizenship of class members may not rely on things like phone numbers and mailing addresses.[24]
This refusal to merely infer necessary citizenship was reiterated by the Ninth Circuit. In Mondragon v. Capital One Auto Finance, California consumers filed a class action against an automobile financer for violating the California Consumers Legal Remedy Act.[25] In asserting application of the local controversy exception, the consumers relied solely on the class’s definition, which limited the class members to “consumers who purchased and registered cars in California.”[26] Though the court noted that it was extremely likely most of the prospective class members were California citizens, the consumers must present some type of evidence that the citizenship requirement was satisfied and cannot merely assume citizenship.[27] The court noted that some members in the proposed class purchased vehicles five years before the case was removable; absent any evidential support of citizenship the conclusion would be based solely on guesswork.[28]
III. Instant Decision
The Eighth Circuit reversed the district court in holding that the employees did not provide sufficient evidence to establish that the local controversy exception to CAFA applied because the use of last-known addresses was inadequate proof of two-thirds Missouri citizenship in the prospective class.[29] Despite 41% of the class being clearly identified as Missouri citizens and an additional – yet unconfirmed – 33% of the class having last-known addresses in Missouri, the court determined that the employees did not meet their burden of proof required for the exception and remand.[30]
The court noted three specific ways plaintiffs seeking remand could meet their burden of proof: (1) affidavit evidence; (2) statistically significant surveys; or (3) redefining the class as only local citizens.[31] Because the employee’s affidavit evidence only accounted for 41% of the required two-third citizenship, the court stressed the standard for statistically significant surveys necessary to establish the remaining citizenship.[32] While random samples of potential class members can sometimes be extrapolated to bridge the gap, the court expressly stated that the limited results being extrapolated must “yield a lopsided result.”[33] The court noted that the extrapolation of unresponsive and unconfirmed last know addresses lacked the necessary “sample, sampling methodology, or other . . . disciplined approach” to be a statistically significant survey.[34] Instead, the court revealed a statistical fallacy in the district court’s extrapolation: “[T]hose still at the last-known address were more likely to respond, and those not at the last-known address were less likely to respond.”[35] The court also emphasized the significant gap between the actually confirmed Missouri residents and the two-thirds citizenship required by the local controversy exception.[36] The court held that the existence of 126 last-known addresses was not sufficient to bridge the 26% gap between confirmed and required Missouri citizens.[37]
IV. Comment
The court’s reluctance to simply infer two-thirds citizenship of potential class members from indefinite address information is supported by the burden of proof placed on plaintiffs by the CAFA exception. Excessive extrapolations would directly conflict with the notion that any doubt in applying CAFA exceptions must be resolved against the party seeking remand. CAFA intends to remove cases with large prospective damages and diverse parties because it provides a larger jury pool and protects defendants from potentially biased local juries – who might be more willing to award larger judgments to local plaintiffs. Though requiring concrete proof of facially or likely true assertions creates a slightly unnecessary judicial inefficiency, the requirement is clearly consistent with the statutory burden and it furthers the purpose of the federal code. This presumptive codified removal and its purpose reinforce the court’s seemingly stringent standard placed on plaintiffs seeking an exception to the CAFA removal.
Though district courts are given discretion to make reasonable inferences when interpreting the significance of the statistical evidence, it is important that established statistical standards are still considered. The court appears to allow some wiggle room for judges to extrapolate available statistics in concluding that currently unavailable data would satisfy the required citizenship. However, the gap must not be excessively wide between the established citizenship and required citizenship. Further, there must be some disciplined statistical approach used to extrapolate the data. This leniency provides future class action plaintiffs the opportunity to provide sufficient sampling methodology and representative sampling while still incentivizing the discovery of clear proof of citizenship to minimize the gap between the proven and required citizenship. Hood illustrates the importance of the gap between established and required citizenship. The employees needed 94 more Missouri plaintiffs to qualify for the local controversy exception, but attempted to satisfy this by asserting evidence of a plaintiff group of 126 unproven Missouri citizens. Absent some strong statistical methodology or sampling, it is difficult to infer that 94 of 126 citizens are still living in Missouri. Future class action plaintiffs would be more likely to receive a CAFA exception if they initially provide some form of accepted statistical concept that supports an extrapolation of their already proven data. The court’s approach serves the judicial intent of CAFA, while still providing plaintiffs a reliable and reasonable opportunity to satisfy the burden necessary for an exception.
- Mark Ohlms
[1] Hood v. Gilster-Mary Lee Corp., 785 F.3d 263, 264 (8th Cir. 2015).
[7] Hood v. Gilster-Mary Lee Corp., No. 3:14-CV-05012-MDH, 2015 WL 328409, at *1 (W.D. Mo. 2015), rev’d and remanded, 785 F.3d 263 (8th Cir. 2015).
[16] 28 U.S.C. § 1332(d)(2) (2012).
[17] 28 U.S.C. § 1332(d)(4) (2012).
[18] Westerfield v. Independent Processing, LLC., 621 F.3d 819 (8th Cir. 2010).
[25] Mondragon v. Capital One Auto Fin., 736 F.3d 880, 882 (9th Cir. 2013).
[29] Hood v. Gilster-Mary Lee Corp., 785 F.3d 263, 264 (8th Cir. 2015).
