Opinion
1:20-cv-00682
06-08-2021
ROBERT SUTTON, On behalf of himself and those similarly situated Plaintiffs, v. DIVERSITY AT WORK GROUP INC. D/B/A UNITED COURIER et al., Defendants.
ORDER GRANTING DEFENDANTS' REQUEST TO CONDUCT DISCOVERY OF 17 RANDOMLY-SELECTED PLAINTIFFS
Timothy S. Black Judge.
This civil action is before the Court on the parties' briefs (Docs. 48 and 50) regarding the scope of discovery pursuant to the Order of April 28, 2021.
BACKGROUND
This Fair Labor Standards Act (“FLSA”) collective action involves one lead Plaintiff and 20 total Plaintiffs who assert they were misclassified as independent contractors and were not paid overtime or compensated for work-related expenses. (Doc. 1 at ¶ 3, ¶ 5). Plaintiff contends that drivers are required to provide their own vehicles to complete deliveries and maintain their vehicles in safe and working condition, including purchasing gasoline, vehicle parts and fluids, insurance, licensing and registration, and other necessities to complete their job duties. (Id. at ¶¶ 130-133). Because these expenses were not reimbursed, Plaintiff claims that after deducting vehicle expenses, the drivers were paid less than minimum wage. (Id. at ¶ 136).
Congress enacted the FLSA in 1938 as a remedial statue “designed to correct labor conditions detrimental to the maintenance of the minimum standard of living necessary for health, efficiency, and general well-being of workers.” Dunlop v. Carriage Carpet Co., 548 F.2d 139, 143 (6th Cir. 1977). The intent of Congress in developing the collective action procedure was to limit the burden on similarly situated employees by allowing them to rely on representative testimony. As the Supreme Court noted in Hoffman-La Roche, Inc. v. Sperling, 493 U.S. 165, 170 (1989), a collective action under Section 216(b) allows employees “the advantage of lower individual costs to vindicate rights by pooling resources” as well as benefitting the judicial system by “efficient resolution in one proceeding of common issues of law and fact.” The FLSA provides that employers may not require employees to work more than forty hours per workweek unless those employees receive overtime compensation at a rate of not less than one-and-one half times their regular pay. 29 U.S.C. § 207(a)(1).
Unless otherwise indicated, and as reflects the parties' own usage, the Court will refer to the lead plaintiff as a singular “Plaintiff” and use “Plaintiffs” or “opt-in Plaintiffs, ” as appropriate, when referring to the larger collective.
Defendant Diversity at Work Group, Inc. d/b/a United Courier operates a delivery and courier service across Ohio, Indiana, Michigan, Pennsylvania, and elsewhere. (Id. at ¶¶ 3, 78). Defendant Lynn Myers is the President of United Courier and Director of Diversity at Work Group. (Id. at ¶¶ 27-28). Her husband, Defendant Jim Meyers, is the Vice President of United Courier. (Id. at ¶¶ 46-47). Her son, Defendant Scott Laminack, is the General Manager of United Courier. (Id. at ¶¶ 61-62). Defendants maintain its drivers were independent contractors, pursuant to agreements. (Doc. 19 at ¶ 99).
The Court conditionally certified a class and Plaintiff sent court-approved notices to 187 drivers. (Doc. 31). At the end of the opt-in period, 20 drivers belonged to the opt-in Plaintiff class. (Doc. 48 at 2). The record does not reflect where these 20 drivers reside.
After Defendants had already served some discovery demands, and at the request of the parties, on April 26, 2021, the Court held a conference to resolve a discovery dispute about the number of Plaintiffs who would respond to discovery and how those Plaintiffs ought to be selected. The Court asked for briefing, which the parties have duly provided. (Docs. 48 and 50).
II. STANDARD OF REVIEW
The Federal Rules of Civil Procedure grant courts broad discretion in determining the scope and method of discovery based upon the circumstances of each case. Fed.R.Civ.P. 26(b)(2). Specifically, Rule 26(b)(2) permits a court, upon its own initiative, to limit "the frequency or extent of use of discovery methods" if:
(i) the discovery sought is unreasonably cumulative or duplicative, or is obtainable from some other source that is more convenient, less burdensome, or less expensive; (ii) the party seeking discovery has had ample opportunity by discovery in the action to obtain the information sought; or (iii) the burden or expense of the proposed discovery outweighs its likely benefit, taking into account the needs of the case, the amount in controversy, the parties' resources, the importance of the issues at stake in the litigation, and the importance of the proposed discovery in resolving the issues.
III. ANALYSIS
Both parties agree to use representative discovery. This means the parties will conduct discovery on some, rather than all, of the opt-in Plaintiffs. The dispute arises around how many Plaintiffs will respond to discovery and how those Plaintiffs will be selected. The Court will first look at method of selection and then turn to the number of Plaintiffs who will respond to discovery.
A. Random Selection
Plaintiff argues Defendants should not be able to handpick the opt-in Plaintiffs to complete discovery. (Doc. 48 at 12-13). According to Plaintiff, the parties should randomly generate the discovery-producing Plaintiffs. (Id.).
While random sampling is not a hard-and-fast rule in class-based discovery, courts have stood behind the proposition that a randomly selected sample helps reduce the likelihood of bias. Hostetler v. Johnson Controls, Inc., No. 3:15-CV-226 JD, 2016 WL 3662263 at *12. (N.D. Ind. July 11, 2016). On the other hand, “[h]and-picking is almost certain to introduce a substantial amount of selection bias into the sample.” Id.
The Court concludes the parties will select Plaintiffs to respond to discovery using random sampling. Hand-picking would unnecessarily introduce the risk of bias. Hostetler 2016 WL 3662263 at *12. Defendants seem to oppose randomized selection only conditionally, if at all. Thus, the parties will proceed with random selection as the method to identify discovery-producing Plaintiffs.
To the extent selection method was ever in dispute, it seems to arise from a misunderstanding. Plaintiff believed he had a deal with Defendants to limit the representative sample to 10. (Doc. 48 at 4.). However, since this dispute has come before the Court, Defendants have made clear they would agree to 10 only if Defendants could choose the Plaintiffs. (Doc. 50 at 6). Defendants do not appear to be against random selection in all cases.
B. Size of the Sample
Defendants argue 17 of the 20 opt-in Plaintiffs ought to produce discovery. According to Defendants, 17 plaintiffs are necessary to create a “statistically significant”and representative sample. (Doc. 50 at 3). A sample of fewer than 17, Defendants claim, would run the risk of an unrepresentative discovery response. (Id. at 3-4). Plaintiff claims 10 opt-in Plaintiffs are sufficient. (Doc. 48 at 13). In support, Plaintiff relies on cases that have limited the representative Plaintiff samples to a percentage of the whole opt-in population. When compared on a percentage basis, Plaintiff argues, Plaintiff's proposal that half the opt-in Plaintiffs respond to discovery is similar or better than most other FLSA cases that employ representative discovery. Plaintiff also points to authority suggesting a small amount of evidence can fairly represent the experience of a larger group.
Assuming a 10% margin of error and a 95% confidence interval.
As the briefing here highlights, courts do not apply a uniform rule of statistical reasoning. Rather, “[t]he fairness and utility of statistical methods … will depend on facts and circumstances particular to those cases.” Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 460, (2016). Courts have recognized the value of “statistically significant” samples in representative discovery among FLSA Plaintiffs. See Smith v. Lowe's Home Centers, Inc., 236 F.R.D. 354, 358 (S.D. Ohio 2006). On the other hand, it is a well-established principle that “[individualized discovery undermines the purpose and utility” of FLSA cases. Gentrup v. Renovo Servs., LLC, No. 1:07-CV-430, 2010 WL 6766418, at *7 (S.D. Ohio Aug. 17, 2010). While individualized discovery is not an option before the Court, it stands to reason that overly broad or burdensome representative discovery may have a similar effect.
Defendants' argument about the benefits of a statistically significant sample is well-taken. A statistically significant sample allows greater assurance the sample will be representative of the whole population. See, e.g., Andreas-Moses v. Hartford Fire Ins. Co., 326 F.R.D. 309, 318 (M.D. Fla. 2018). A representative sample, in turn, provides more confidence in inferences drawn from the sample. When it comes to statistically significant sampling, as the Court in Rosenbohm v. Cellco Partnership makes clear, size matters. 2:17-CV-731, 2019 WL 2141901, at *2 (S.D. Ohio May 16, 2019). According to the court in that case, a sample must be “of sufficient size so as to achieve statistical significance to the desired level of confidence in the result obtained.” Id. To arrive at the same 95% confidence level as that achieved in Rosenbohm, and assuming a 10% margin of error, Defendant proposes a representative sample of 17. Id.
The Court is persuaded that a statistically-sample of 17 Plaintiffs is desirable. But that does not answer the question of whether it is workable. For that reason, the Court must turn to Plaintiff's arguments and determine whether the imposition of a sample of 17 runs the risk of undermining this FLSA collective action by making Plaintiff's case “unnecessarily laborious” or “unworkable in practice.” Gentrup v. Renovo Servs., LLC, No. 1:07-CV-430, 2010 WL 6766418, at *5 (S.D. Ohio Aug. 17, 2010) (citing Wilson v. Guardian Angel Nursing, Inc., No. 3:07cv069, 2008 U.S. Dist. LEXIS 92250 (M.D.Tenn. Nov. 12, 2008)).
Opposing a sample of 17 and arguing instead for a sample of 10, Plaintiff raises many points that are undoubtedly correct. Plaintiff argues, and the Court agrees, that statistical significance is not a “legal requirement.” Stillman v. Staples, Inc., No. CIV.A.07-849(KSH), 2009 WL 1437817, at *20 (D.N.J. May 15, 2009). Plaintiff is also correct that the approved representative sampling in other FLSA cases would not meet Defendants' proposed standards for statistical significance. (Doc. 48 at 8). To reach a confidence interval of 95%, for example, the sample at issue in this Court's decision in Gentrup would have had to be at least 51 Plaintiffs-far and above the 22-Plaintiff sample ultimately ordered. 2010 WL 6766418, at *8 (S.D. Ohio Aug. 17, 2010).
Finally, Plaintiff makes the important point that the inquiry into the sample size ought to turn on factual considerations. Plaintiff cites again to this Court's decision in Gentrup wherein the Court analyzed, among other things, geographic isolation of the Plaintiffs and other factors which may show the “burdensomeness of discovery.” Id. at *3. Another case cited by Plaintiff puts forward a three-pronged inquiry into appropriate limits on FLSA discovery based on “….the size of the class, the costs and burdens associated with discovery, and the type of discovery sought.” Martinez v. First Class Interiors of Naples, LLC, No. 3:18-CV-00583, 2020 WL 7027504, at *3 (M.D. Tenn. Nov. 30, 2020).
The problem is Plaintiff points to few factual circumstances that would compel the Court to further limit discovery in the present case. Plaintiff does not assert that the opt-in class is geographically diffuse or that discovery demands are anticipated to be more intensive than usual. On the record before it, the Court can discern no unique burdens faced by Plaintiffs. Cf. Martinez, 2020 WL 7027504, at *4 (M.D. Tenn. Nov. 30, 2020) (noting the “need to translate the requests from English to Spanish (and, the Court assumes, the opt-in class members' responses from Spanish to English), many opt-in class members' low literacy levels, and the fact that many opt-in class members have moved away from Nashville where Plaintiffs' counsel are located.”).
Nor should the opt-in class size present much of a burden to Plaintiff. At 17 opt-in Plaintiffs, Defendants' proposed sample here is smaller than most of the representative samples approved in the cases relied on by Plaintiff. Based on the sample sizes in those cases, it cannot be true that requiring 17 opt-in Plaintiffs to respond to discovery, on its own, undermines the utility of the FLSA as a collective vehicle. Simply put, Plaintiff does not demonstrate that a 17-plaintiff sample for discovery would be “unnecessarily laborious” or “unworkable in practice” in this case. Gentrup 2010 WL 6766418, at *5.
In Gentrup, the sample was 22, exclusive of the nine named Plaintiffs. Monroe v. FTS USA, LLC, 860 F.3d 389, 395 (6th Cir. 2017) (50 Plaintiffs); Ross v. Jack Rabbit Servs., LLC, No. 3:14-CV-00044-TBR, 2014 WL 2219236, (W.D. Ky. May 29, 2014) (24 Plaintiffs). Halleen v. Belk, Inc., No. 4:16-CV-00055, 2017 WL 1495085, at *2 (E.D. Tex. Apr. 26, 2017) (25 Plaintiffs); Lloyd v. J.P. Morgan Chase & Co., No. 11 CIV. 9305 LTSHBP, 2015 WL 1283681, at *6 (S.D.N.Y. Mar. 20, 2015) (written discovery on “100 opt-ins who have not signed arbitration agreements and [Defendant] may depose 20 of those opt-ins.”); Strauch v. Computer Scis. Corp., No. 3:14 CV 956 JBA, 2015 WL 540911, at *3 (D. Conn. Feb. 10, 2015) (32 Plaintiffs). Scott v. Bimbo Bakeries USA, Inc., No. CIV.A. 10-3154, 2012 WL 6151734, at *6 (E.D. Pa. Dec. 11, 2012) (written discovery on 65 Plaintiffs and depositions on 20).
Instead of making an affirmative case to limit discovery to 10 Plaintiffs based on the circumstances of this case, Plaintiff relies on comparative percentage benchmarking from other cases. Plaintiff proposes the sample is composed of half the opt-in class, or 10 out of 20. As Plaintiff notes, 50% is a higher percentage of the total opt-in Plaintiffs than is approved in many FLSA cases. (Doc. 48 at 7).
The Court finds the comparative percentage analysis engaged in by Plaintiffs of little value. (Doc. 48 at 7). It is true that in other FLSA cases courts have approved representative samples that constitute a much smaller percentage of the whole opt-in class. See e.g., Ross v. Jack Rabbit Servs., LLC, No. 3:14-CV-00044-TBR, 2014 WL 2219236, (W.D. Ky. May 29, 2014) (ordering representative discovery of a 10% sample). Without the context of the total opt-in class size or a comparison of relative plaintiff burdens, those percentages mean very little, because they do not provide insight into either the representational value of the sample or the risk of undermining the FLSA collective. Martinez, 2020 WL 7027504, at *3 (M.D. Tenn. Nov. 30, 2020).
Plaintiff further argues that limiting discovery to 10 Plaintiffs is warranted even for a small opt-in class because, as the Court held in Reich v. Southern New England Telecommunications, “the weight to be accorded evidence is a function not of quantity but of quality.” 121 F.3d 58, 67 (2d Cir. 1997) (quoting Anderson v. Mt. Clemens Pottery 328 U.S. 680, 682 (1946)). Reich, as Plaintiff points out, takes guidance from Anderson v. Mt. Clemens Pottery, which held that the testimony of 8 was amply representative evidence to award damages to 300 Plaintiffs in an FLSA case. 328 U.S. 680, 682 (1946).
In its “quality-over-quantity” argument, Plaintiff's reliance on Reich and Mt. Clemens is misguided. Both those cases address representational evidence at trial rather than in certification-oriented discovery. The difference is important. The discovery at issue here is likely to support a motion to decertify the class wherein the critical question will be whether the Plaintiffs are indeed similarly situated. See Frye v. Baptist Mem'l Hosp., Inc., 495 Fed.Appx. 669, 670 (6th Cir. 2012). Once an FLSA plaintiff class is presenting prima facie evidence, on the other hand, similarity has already been established. For that reason, a fully representative sample is more important when a case, as here, in in the certification phase.
Both Reich and Mt. Clemens are also distinguishable because they deal with the employer's failure to maintain records. Those cases have an additional justification for representative testimony that is not present here. As the Court in Reich explains “where the employer's records are inaccurate or inadequate and the employee cannot offer convincing substitutes ... [t]he solution ... is not to penalize the employee by denying him any recovery on the ground that he is unable to prove the precise extent of uncompensated work. Reich v. S. New England Telecommunications Corp., 121 F.3d 58, 66 (2d Cir. 1997) (quoting Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680, 687-88 (1946)).
Plaintiff's remaining arguments concern Defendants' access to records and a proposed way forward where the Court would initially adopt Plaintiff's proffered limit of 10 Plaintiffs to produce discovery and then let Defendants move for wider discovery later. Neither is enough to overcome Plaintiff's failure to offer an affirmative, fact-based justification to limit discovery to 10 Plaintiffs.
Plaintiff anticipates a sample of 10 Plaintiffs would suffice for Defendants because “the primary evidence in the case-the deliveries completed by the drivers, the hours worked by the drivers, the money paid to the drivers, etc.-has been maintained in the Defendants' records.” (Doc. 48 at 13). The Court finds that it is premature to determine what constitutes “primary evidence.” That is an issue to be refined by discovery. Furthermore, each side has information advantages over the other. While Defendants' counsel has access to Defendants' records, Plaintiff's counsel has access to the opt-in Plaintiffs. For these reasons, the Court cannot accept a mere allegation that Defendants have better access to evidence as a justification to limit discovery to 10 opt-in Plaintiffs.
Finally, Plaintiff proposes Defendants can take discovery from 10 opt-in Plaintiffs and then, if Defendants are unsatisfied with the discovery production, move the Court for further discovery. (Doc 48 at 13). It is the Court's task, however, to determine the correct number of Plaintiffs to produce discovery in the first instance.
The Court concludes that such number is 17. While statistical significance is not the rule, the benefit of a statistically significant sample of 17 opt-in Plaintiffs is clear. Against this clear benefit, Plaintiff fails to demonstrate that a greater limitation on discovery is warranted.
IV. CONCLUSION
For the reasons stated, the Court finds that:
1. Defendants may conduct discovery upon 17 Plaintiffs selected at random;
2. The parties shall meet-and-confer to generate the list of 17 Plaintiffs through an agreed upon randomization procedure; and
3. Within 14 days of the date of this order, the parties shall submit a joint proposed order re-setting deadlines for discovery, decertification and summary judgment motions.
IT IS SO ORDERED.