Opinion
12-CV-09672
02-21-2023
JOE EAGLE, MICHAEL KEYS, JAMES ZOLLICOFFER, and EVAN FRANKLIN on behalf of themselves and other similarly situated individuals, Plaintiffs, v. VEE PAK, INC., VEE PAK, LLC d/b/a VOYANT BEAUTY, and STAFFING NETWORK HOLDINGS, LLC, Defendants.
MEMORANDUM OPINION AND ORDER
John J. Tharp, Jr., United States District Judge
The claim in this case is straightforward. The plaintiffs, who are African American, allege that Vee Pak, a beauty supply distributor, had a policy that favored hiring Latino workers over African Americans and instructed several staffing agencies to implement that policy when filling temporary positions in its warehouse. They have brought this putative class action, individually and on behalf of other African Americans similarly denied work, against Vee Pak and three staffing agencies that allegedly implemented Vee Pak's discriminatory policy. Two of the staffing agencies have settled the claims against them; Vee Pak and the third staffing agency, Staffing Network, remain in the case. The plaintiffs have moved to certify a class of African American workers who sought, but were denied, work assignments at the three staffing agencies from which they could have been referred to Vee Pak between 2011 and 2015. Vee Pak and Staffing Network oppose certification and have also moved to bar the testimony of Dr. Mark Bendick, a labor economist on whose proffered opinions the plaintiffs rely in their classcertification motion.
Because Bendick's report satisfies the admissibility standards articulated in Daubert and Rule 702, the Court denies Defendants' motion to strike Bendick's testimony. The Court finds that the defendants' objections to Bendick's report implicate its probative value, rather than its admissibility. The Court further finds that plaintiffs' proposed class satisfies the criteria of Federal Rule of Civil Procedure Rule 23. A class action is the most efficient way to determine whether a company-wide discriminatory policy caused the defendants to disfavor African Americans. Accordingly, the Court certifies the class with an amended definition.
BACKGROUND
A. Vee Pak Temporary Staffing Procedures
Vee Pak, a company located in Countryside and Hodgkins, Illinois, operates manufacturing facilities in which workers fill and cap bottles and tubes for personal-care and drug companies. Due to cyclical demand for its clients' products, Vee Pak hired temporary workers through staffing agencies Staffing Network Holdings LLC (“Staffing Network”), Personnel Staffing Group, LLC d/b/a Most Valuable Personnel (“MVP”), and Alternative Staffing, Inc. d/b/a ASI (“ASI”) to fulfill client orders during the class period.
Nominally, there are two Vee Pak defendants: Vee Pak, Inc., and Vee Pak LLC, d/b/a Voyant Beauty. Vee Pak LLC is named only as Vee Pak's successor and its liability, if any, is entirely derivative of that of Vee Pak, Inc. Because of the derivative nature of Vee Pak LLC's liability, no further mention of Vee Pak LLC is required.
As noted in Section G, infra, the proposed class period spans from January 1, 2011, up through and including December 31, 2015 for Staffing Network and ASI, and from January 1, 2011, up through and including October 21, 2013 for MVP.
During the class period, Vee Pak laborers worked in two shifts, one running from 5:30 a.m. to 3:30 or 4:30 p.m. and one running from 3:30 or 4:30 p.m. to midnight or 2:00 a.m. Nearly all of Vee Pak's temporary workers were assigned to Vee Pak's Countryside manufacturing facility in the far western side of the Chicago metropolitan area. Some staffing-agency employees testified that workers would sometimes decline assignments to Vee Pak. Workers' hesitance was due in part to the time and difficulty required to travel to Vee Pak from Chicago, the low wages, and the grueling nature of the temp positions, which required lifting and standing for long hours.
The parties dispute the level of skills and experience that Vee Pak required of its workers. According to testimony that plaintiffs cite, Vee Pak's production-floor jobs required no formal training or education. All necessary skills could be learned on the job through training. Vee Pak did not require its employees to submit criminal background checks, credit tests, or drug tests. For its part, Vee Pak presents testimony that its managers instructed staffing-agency workers to send experienced, or “regular” workers-workers who had consistently worked at Vee Pak in the past-to help run the production lines. These workers, according to Vee Pak, knew how to do different jobs more efficiently and effectively without supervision, and reduced Vee Pak's availability for less-experienced workers. According to testimony from a Staffing Network employee, dispatchers assigned workers who had prior warehouse experience to Vee Pak, worrying that employees who had never worked in a warehouse could not handle long, demanding shifts.
Plaintiffs present evidence that Vee Pak used the term “regular” worker as a code word for its preferred type of worker: one who is Latino, rather than African American. Some staffingagency management understood “regular” to mean non-African American. According to plaintiffs, Vee Pak kept a list of “regulars” that it preferred to be assigned to shifts. As certain staffing-agency employees testified, almost all the workers on Vee Pak's list of “regulars” happened to be Latino. Vee Pak's Executive Director, Tom Zwartz, however, stated in an email that Vee Pak “do[es] not order specific individuals, and . . . do[es] not provide long term assignments to specific individuals,” but instead requests individuals “with previous experience at Vee Pak.” Dec. 2, 2015 Email from Tom Zwartz, ECF No. 585-2. Zwartz later confirmed that his email concerning Vee Pak's assignments of specific individuals reflected official policy.
Unlike the disagreement about worker qualifications and eligibility, the parties dispute little about the logistics of scheduling. All three staffing agencies referred workers to Vee Pak in a similar manner. Vee Pak's master schedulers determined the number of laborers needed each shift, and Vee Pak would communicate those needs to the staffing companies each day. Until 2013, Vee Pak management called dispatchers directly to communicate staffing needs. After 2013, Vee Pak submitted its orders by email. Workers registered for work at the staffing agencies using sign-in sheets. At Staffing Network and ASI, workers would also fill out an application. Dispatchers then assigned the workers to a client company either in person or by calling the workers directly. If a position was available, then a worker would be assigned to a client company like Vee Pak.
During some portion of the class period, some staffing agencies would transport workers to client companies, including to Vee Pak, by van, either directly from the workers' homes or from the staffing agencies. Some workers had to travel directly to the client using their own vehicles or via public transportation. The parties dispute the extent and availability of van transportation at each agency.
As detailed below, multiple staffing-agency employees testified that they were given explicit instructions not to assign African American workers to Vee Pak. Some workers who did receive assignments initially, moreover, were turned away at Vee Pak's door. When workers could not comply with the demands of Vee Pak's manufacturing job or did not follow the instructions included with Vee Pak's training materials, they were designated “Do Not Return,” or “DNR.” Some testifying staffing-agency employees observed that workers whom Vee Pak designated DNR were disproportionately African American. Other staffing-agency employees testified that Vee Pak turned away African American laborers that the staffing agency brought to Vee Pak.
B. Testimony of MVP Employees
Until 2015, MVP operated an office in Cicero, from which it would staff temporary workers to clients, including Vee Pak. MVP referred workers to Vee Pak from 2008 until October 2013.
MVP dispatcher Yessenia Saucedo testified about Vee Pak's and MVP's hiring practices. According to Saucedo, her supervisors and other dispatchers instructed her not to send African American workers to Vee Pak because Vee Pak would send them back.
Other MVP employees testified that MVP also generally avoided referring African American workers to various non-Vee Pak clients. MVP supervisors told dispatchers and onsite employees not to send African American workers to certain clients, sometimes openly, and sometimes using certain code words, like “bilingual” to refer to Latinos and “lazy” or “not bilingual” to refer to African Americans. MVP dispatcher Saucedo testified that, following supervisors' instructions, she falsely told African Americans that MVP was not giving out applications, while making applications available to Latino workers. Another dispatcher, Pamela Sanchez, testified that she was similarly trained to withhold applications from African American workers.
C. Testimony of Staffing Network Employees
Staffing Network provided workers from its Cicero and Burbank offices to Vee Pak from 2009 to 2015. Most of these workers were referred from Staffing Network's Burbank office, the location of which changed during the class period. According to Staffing Network employees, Vee Pak's needs for workers fluctuated and was based on the type of work available at the time. Depending on the number of workers Vee Pak needed, Staffing Network would send workers from Vee Pak's list of “regulars,” then “back up” workers (i.e., workers ones that had previously worked at Vee Pak but not on the “regulars” list), then new applicants, in that order.
Staffing Network branch manager Judy Cruz testified that Staffing Network managers instructed her to refer only Latino workers to Vee Pak when she started working there in 2000. Although Cruz testified that she did not follow those instructions, another Staffing Network employee, Paloma Martinez, testified that Cruz condoned the policy after the employee and African American workers themselves raised concerns about Vee Pak's racial discrimination. Another Staffing Network witness, who was the branch manager before the class period, testified that Cruz had told him directly that Vee Pak only wanted Spanish-speaking people.
Other Staffing Network employees testified about Vee Pak's discriminatory practices. Former branch manager Dennis Medina testified that in 2007, a Vee Pak supervisor, Staffing Network's president, and another Staffing Network branch manager all told him not to send African Americans to Vee Pak. When Medina attempted to voice disagreement with this policy, Staffing Network management told him to maintain the status quo. This practice appeared to continue beyond Medina's tenure. A Staffing Network dispatcher, Adriana Armendariz, testified that in 2013 and 2014, a branch manager told her to tell a local organization to stop sending African Americans because Vee Pak wanted only Latino workers. Armendariz testified that she would try to send African American employees to Vee Pak despite instructions to the contrary.
Dispatcher Martinez testified that many other employees told her not to send African Americans to Vee Pak, and that a Vee Pak supervisor confirmed these rumors about Vee Pak's preferences. The Vee Pak supervisor told Martinez, in Spanish, that he didn't “want those black people,” calling African Americans “lazy.” Martinez Dep. at 32-35, ECF. No. 544-27. Martinez stated that she and other employees would defy these instructions and continued to refer African Americans to Vee Pak and other employers. Martinez further testified that employees “knew better” than to follow racist policies because Staffing Network provided them with antidiscrimination training. Martinez Dep. at 52-54, ECF. No. 565-5.
D. Testimony of ASI Employees
ASI provided referrals to Vee Pak from its Cicero office. ASI's Chairman, Steven Swerdloff, testified that Vee Pak supervisors explicitly told him that they only wanted Latino workers. Vee Pak shift supervisor Juan Montoya told Swerdloff that Vee Pak wanted “young, good-looking Hispanic females.” Swerdloff Dep. at 107, ECF No. 544-4. Another Vee Pak supervisor told Swerdloff that he “wanted Hispanic individuals because the African-Americans never stayed.” Id. at 108. Swerdloff testified that he followed Vee Pak's discriminatory instructions for fear of losing its business.
ASI employees at other levels of management also described discrimination at the agency. An ASI dispatcher testified that Montoya threatened to take his business elsewhere if ASI did not provide Vee Pak with “his kind of people.” Gonzalez Dep. at 26, ECF No. 544-29. The dispatcher reported these comments to Swerdloff, who told her to comply with Montoya's instructions. ASI's President during the class period, Kyle Carstensen, testified that Montoya asked “[w]hat [he was] supposed to do with all [these] people” after observing that a large number of African American workers referred from ASI checked in at Vee Pak. Carstensen Dep. at 153-54, ECF No. 544-31. An ASI Office Manager, Julie Tracey, also testified that she believed ASI was following a discriminatory practice to please Vee Pak.
E. Named Plaintiff Testimony
Joe Eagle and Michael Keys are African American workers who seek to represent the Staffing Network subclass. Eagle first sought work at Staffing Network in March 2011. During his first visit, he completed an application and was told to return the next morning. This began a pattern of Eagle arriving at Staffing Network early in the morning, signing in, and waiting for hours to no avail. While waiting, Eagle often observed that Staffing Network assigned Latino workers, who arrived at Staffing Network in vans that the staffing agency provided, to Vee Pak. Eagle did not, however, observe any African American workers on a van destined for Vee Pak. Eagle's work history and testimony reflects that from March to October 2011, he received some work assignments from other Staffing Network clients, though he was never assigned to Vee Pak. Eagle found full time employment in 2012.
Keys sought work from Staffing Network around the same time as Eagle, in September 2011. Even though Staffing Network told Keys that it only accepted new applications two days a week, Keys observed new Latino workers successfully obtain work on other days of the week. Like Eagle, Keys testified that he was consistently among the first workers to sign in at Staffing Network, but that the agency did not give him an assignment for two weeks. Like Eagle, Keys also testified that Staffing Network assigned him to a few of its clients but never to Vee Pak. Keys' work history records reflect that Staffing Network sporadically assigned him work between September 2011 and January 2012. Also like Eagle, Keys observed Staffing Network sending out Latino workers while he was waiting for work. He discussed with other African American workers his and their sense that they were not receiving assignments from Staffing Network at the same rate as Latino workers. Keys stopped looking for work at Staffing Network in January 2012, after he started attending vocational school and college.
James Zollicoffer testified that he experienced similar treatment to Eagle and Keys at Staffing Network, although outside of the class period (in 2009 and 2010). Seeking to represent the MVP subclass, Zollicoffer also testified to MVP's discriminatory treatment during the class period. That testimony parallels his, Eagle's and Keys' testimony about Staffing Network. After Zollicoffer moved to Chicago in the summer of 2009, he sought work at MVP's Cicero office. The first time he came into that office, he signed in and was told to come back the next day. Although Zollicoffer returned to the office early the next morning, Staffing Network did not assign him work. Zollicoffer testified that on multiple occasions in 2009, he would travel to MVP's Cicero office during the early morning hours on foot, seeking assignment for temporary work. He further testified that he observed MVP assign Latino workers to clients on days when he was waiting for work for multiple hours. Zollicoffer unsuccessfully sought work from MVP again in 2012 and 2013, visiting the office in 2013 and calling the office a couple of times in those years.
Evan Franklin testified to similar treatment from ASI. She first sought placements at ASI's Cicero office in September 2014. Franklin continued to seek assignments from ASI on numerous occasions, indicating that she was available and that she had no work limitations. When Franklin was actively seeking work, she would call ASI or visit the office. Like the other plaintiffs, Franklin received sporadic work placements over a period of several months, but was never placed at Vee Pak. And like the other plaintiffs, she observed that ASI assigned Latino workers, but not African Americans, to work when she visited the ASI offices in person.
F. Bendick Declaration
Plaintiffs have submitted the testimony of Dr. Marc Bendick, Jr., Ph.D., to supplement the witness testimony concerning MVP's hiring practices. Bendick performed a statistical analysis of the relevant labor market that purports to find a shortfall of staffing-agency placements of African American workers at Vee Pak. Specifically, Bendick compared his expected representation of African Americans eligible to work at Vee Pak to Vee Pak's actual hiring practices. Analyzing each staffing agency separately, Bendick first determined the expected representation of Non-Hispanic African Americans (“NHAA”s), measured by the expected proportion of laborer pay that each staffing agency would invoice to Vee Pak.
To find the expected representation of NHAAs, Bendick first determined the “reasonable recruitment area,” or relevant labor market, for each staffing agency. From staffing-agency invoices, Bendick obtained the addresses of all the workers whom each staffing agency employed during the class period, whether those workers were assigned to Vee Pak or another client. Bendick then located each address in a Public Use Micro Data Area (“PUMA”), the smallest continuous geographic unit containing an average of 100,000 people for which the U.S. census bureau releases data. Bendick followed Census bureau data practice by using PUMAs based on the 2000 decennial census for all census data collected before 2012 and PUMAS based on the 2010 decennial census for all census data collected beginning in 2012 or later.
Next, for both the 2000 and 2010 decennial census PUMAS, Bendick ordered each PUMA in descending order, beginning with the PUMA with the highest number of addresses and ending with the PUMA with the lowest number of addresses. Bendick included in his analysis only the PUMAs that accounted for 95 percent of the addresses that he collected; he stopped counting PUMAs once 95 percent of addresses were accounted for. The remaining PUMAs constituted each staffing agency's reasonable recruitment area.
Bendick then determined the percentage of all the workers living in each PUMA who were likely to be qualified for and interested in Vee Pak referrals. To do that, Bendick counted individuals who answered in the yearly census that they were in the Civilian Labor Force (i.e., actively seeking work), had no college degrees, and earned less than median earnings for the Chicago metropolitan area. Bendick then calculated the percentage of those workers that were NHAAs. According to Bendick, this number gave him the expected representation of NHAAs that would be interested in working for Vee Pak in each PUMA for each year. After adding up these percentages for each class-period year using both the 2000 and 2010 PUMAs, Bendick weighted the proportion of data collected under each type of PUMA. Following standard statistical practice, Bendick calculated a confidence interval, or range in which a figure can be considered accurate with 95 percent certainty, for each of the figures reflecting NHAA representation.
The Census Bureau's American Community Survey only releases data reflecting five-year periods each year. To analyze each class period year, Bendick used the census data set for which the year was the median. For example, in 2011, the relevant census data that Bendick examined was the data set spanning 2009-2013, which was issued in 2013.
Finally, Bendick determined the actual percentage of workers assigned to Vee Pak from the staffing agencies who were NHAAs. Because the workers' races were not provided with the invoices from which Bendick procured the addresses, he employed a “geocoding” technique to estimate actual race representation for each agency. Through geocoding, Bendick used Census data to determine the probability that a person living in an address within a Census block was a certain race (based on the racial demographic of the block). Bendick then applied the probability that a worker was a certain race to the pay invoiced for that worker. Using that number, Bendick estimated the percentage of pay that each staffing agency invoiced to Vee Pak which was attributable to NHAAs.
To determine whether there was a shortfall in Vee Pak's hiring of African Americans, Bendick calculated the difference between the actual representation and the expected representation of African Americans for each year (as measured by percentage of worker pay). Bendick found a 20.7% shortfall in MVP NHAA worker pay, a 29.4% shortfall in Staffing Network NHAA worker pay, and a 27.4% shortfall in ASI NHAA worker pay during the class period. Bendick contends that the shortfall for the three agencies was “statistically significant beyond a shadow of a doubt.” Bendick Decl. ¶ 9(e), ECF No. 570-2. The number of standard deviations, or a measure of how far a value set differs from an expected value, for all five years, is 177. According to Bendick, the probability of observing such a shortfall by chance is less than one in a trillion.
G. Procedural History
Plaintiffs originally sued Vee Pak, ASI, MVP, and Staffing Network in 2012, alleging that the defendants violated Title VII of the Civil Rights Act of 1964 and 42 U.S.C. § 1981 by engaging in racially discriminatory hiring practices. Since then, some of the named plaintiffs have changed, and plaintiffs are no longer pursuing a claim based on defendants' violation of Title VII. Beyond that, plaintiffs' allegations remain largely the same as they were at the start of the lawsuit. In 2014, this Court denied defendants' motion to strike class allegations and to dismiss the complaint, holding that the proposed classes facially met the requirements of Rule 23. Lucas v. Vee Pak, Inc., 68 F.Supp.3d 870, 880 (N.D. Ill. 2014). This Court approved a settlement between plaintiffs and ASI in 2018, holding that the proposed ASI settlement class also met Rule 23's requirements. Order of Final Approval of Partial Class Settlement between Plaintiffs and Defendant Alternative Staffing, Inc. Only, ECF No. 355; see also Lucas v. Vee Pak, Inc., No. 12-CV-09672, 2017 WL 6733688, at *1 (N.D. Ill.Dec. 20, 2017) (granting preliminary certification to ASI settlement class).
This Court has also recently issued a preliminary settlement approval order between Plaintiffs and Personnel Staffing Group, the company doing business as MVP. Preliminary Approval Order of Partial Settlement with PSG, ECF No. 637. That order was issued in conjunction with Judge Ellis's recent preliminary approval between Personnel Staffing Group and plaintiffs in Zollicoffer v. Gold Standard Baking, No. 13-cv-01524. In that case, a class represented by Zollicoffer alleged that MVP and another client company engaged in similar discriminatory hiring practices. Judge Ellis certified that class in 2020. Zollicoffer v. Gold Standard Baking, Inc., 335 F.R.D. 126, 133 (N.D. Ill. 2020). A final approval hearing in this case is scheduled on April 7, 2023.
Plaintiffs move to certify three subclasses pursuant to the below definitions, with the following named plaintiff representatives:
Staffing Network subclass (represented by Joe Eagle and Michael Keys): African American laborers who sought work assignments at Staffing Network from which they could have been referred to Vee Pak, but on one or more occasions were not assigned to work at Vee Pak during the period of January 1, 2011, up through and including December 31, 2015.
MVP subclass (represented by James Zollicoffer): African American laborers who sought work assignments at MVP from which they could have been referred to Vee Pak, but on one or more occasion were not assigned to work at Vee Pak during the period of January 1, 2011, up through and including October 21, 2013.
ASI Subclass (represented by Evan Franklin): African American laborers who sought work assignments at ASI from which they could have been referred to Vee Pak, but on one or more occasion were not assigned to work at Vee Pak during the period of January 1, 2011, up through and including December 31, 2015.
Through discovery, the plaintiffs concluded that the last date that MVP issued checks to laborers for Vee Pak assignments was October 21, 2013. Because MVP stopped assigning workers to Vee Pak on that earlier date, the end of the MVP class period differs from the Staffing Network and ASI class periods.
DISCUSSION
Rule 23 requires the Court to perform a “rigorous analysis” when deciding whether to certify a class. Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350-51 (2011) (quoting General Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 161 (1982)). At the class-certification stage of litigation, the Court must thread the needle between two countervailing considerations. On one hand, “Rule 23 does not set forth a mere pleading standard,” and a party seeking class certification “must be prepared to prove” that the proposed class meets all of the Rule's prerequisites. Id. at 350. The class-certification inquiry may therefore overlap somewhat with the merits. Id. at 351. On the other hand, this Court may only examine the merits to the extent that they are relevant to Rule 23's requirements; the class-certification stage is not the time for “free-ranging merits inquiries.” Amgen Inc. v. Connecticut Ret. Plans & Tr. Funds, 568 U.S. 455, 465-66 (2013).
When faced with a challenge to expert testimony at the class-certification stage, the line between merits inquiries and proof of certification blurs. To guide district courts, the Seventh Circuit has held that “a district court must conclusively rule on any challenge to the expert's qualifications or submissions prior to ruling on a class-certification motion” when that report is “critical” to certification. Am. Honda Motor Co. v. Allen, 600 F.3d 813, 815-16 (7th Cir. 2010). Courts must err on the side of addressing expert testimony first; if any information is even relevant to any of Rule 23 requirements, a court must address a challenge to an expert's reliability. Id.; see also Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 812 (7th Cir. 2012). Here, Defendants move to strike Dr. Bendick's declarations and testimony. His report purports to establish a statistical shortfall between the expected and actual representation of African Americans in referrals to Vee Pak. Such information is undoubtedly relevant to the question of whether the defendants enforced a discriminatory policy; the policy's expected effect is probative of its existence. The Court therefore concludes-and the parties agree-that it should resolve the defendants' motions to strike Dr. Bendick's expert report before addressing class certification.
In a sur-reply to the class-certification motion, the defendants also objected to the plaintiffs' introduction of the Declaration of Christopher Williams and its supporting summary exhibits in their reply brief. The Court will not, however, consider Vee Pak's Federal Rule of Evidence 1006 objections at this juncture. The Williams Declaration is not “critical” to certification; indeed, the Court did not consider it when determining whether to certify the class. Am. Honda Motor Co., 600 F.3d at 815. The defendants may reassert these arguments should the plaintiffs seek to introduce the declaration and supporting summaries as evidence during a later stage of litigation. At this point, however, the Court finds it premature to resolve this admissibility issue.
I. Motions to Strike Dr. Bendick's Declaration and Testimony
Both the Supreme Court's opinion in Daubert v. Merrell Dow Pharmaceuticals Inc., 509 U.S. 579 (1993) and Federal Rule of Evidence 702 govern the admissibility of expert witness testimony. See Gopalratnam v. Hewlett-Packard Co., 877 F.3d 771, 778 (7th Cir. 2017). Pursuant to that authority and its role as a gatekeeper of evidence, the Court must determine “whether the witness is qualified; whether the expert's methodology is scientifically reliable; and whether the testimony will assist the trier of fact to understand the evidence or to determine a fact in issue.” Id. at 779 (cleaned up). The first and third inquiries are not at issue in this case; the defendants do not meaningfully contest Bendick's qualifications, and his findings of a shortfall in hiring African Americans, if reliable, would render it more likely that a discriminatory policy existed. Like the parties, the Court therefore focuses its inquiry on the reliability of Bendick's methodology.
In its reply brief, Staffing Network points out that, contrary to plaintiffs' representations, some courts have previously excluded Dr. Bendick's testimony. But the Court does not construe this passing reference as a meaningful challenge to Bendick's qualifications.
Rather than scrutinizing the accuracy of the expert's conclusions, the Court evaluates “the soundness and care” of the expert's analysis. C.W. ex rel. Wood v. Textron, Inc., 807 F.3d 827, 834 (7th Cir. 2015) (quoting Schultz v. Akzo Nobel Paints, LLC, 721 F.3d 426, 431 (7th Cir. 2013)). “Reliability . . . is primarily a question of the validity of the methodology employed by an expert, not the quality of the data used in applying the methodology or the conclusions produced.” Manpower, Inc. v. Ins. Co. of Pennsylvania, 732 F.3d 796, 806 (7th Cir. 2013). The proponent of the testimony bears the burden of showing that its proffered testimony meets the admissibility requirements by a preponderance of evidence. Lewis v. CITGO Petroleum Corp., 561 F.3d 698, 705 (7th Cir. 2009). Reliability is a flexible inquiry and is not geared toward any fixed evaluative factors. See Gopalratnam, 877 F.3d at 780.
Bendick's basic methodology is not controversial. “Studies in employment discrimination cases typically begin by defining the relevant labor market, and then ask what the results would be for the salient variable . . . if there were no discrimination.” Adams v. Ameritech Servs., Inc., 231 F.3d 414, 424 (7th Cir. 2000). That is exactly what Bendick did here. Bendick determined the expected representation of African Americans at Vee Pak by defining the relevant labor market. To do so, he determined the percentage of persons in the reasonable recruitment area who fit certain criteria. Of those persons, Bendick calculated the percentage that were African American. Next, Bendick compared that expected representation to the actual representation at Vee Pak. He then computed the “[t]he ‘standard deviation' from the expected norm and indicate[d] how likely it is that race played no part in the decisionmaking” (in this case, extremely unlikely). Id. Nevertheless, the defendants argue that Bendick's methodology was unreliable because he failed to consider certain factors it claims were critical to his analysis. The Court considers those arguments in turn.
A. Consideration of Workers' Commuting Distance
The defendants' principal qualm with Bendick's statistical analysis is his purported failure to consider the commuting distance, and associated costs, of workers' commutes when defining the defendants' reasonable recruitment area. In selecting the PUMAs that constituted his reasonable recruitment area, Bendick used the addresses of employees who had been staffed at any of ASI's, MVP's, or Staffing Network's employer partners during the class period, regardless of whether they had ever been staffed at Vee Pak. Because Bendick's analysis included workers who did not embark upon the commute to Vee Pak, Defendants argue that Bendick wholly failed to consider workers' actual commuting distance, rendering his methodology unreliable.
Courts have repeatedly recognized that “the selection of the variables to include in a regression analysis is normally a question that goes to the probative weight of the analysis rather than to its admissibility.” Manpower, 732 F.3d at 808 (collecting cases). While reaffirming this general principle, the Seventh Circuit has recognized that one variable is “crucial” to defining a market from which an employer would be expected to draw workers. See E.E.O.C. v. Chicago Miniature Lamp Works, 947 F.2d 292, 303 (7th Cir. 1991). An expert must consider a worker's interest the job, which is often informed by the difficulty of a worker's commute, when defining a relevant labor market. See id. at 302; accord Bennett v. Roberts, 295 F.3d 687, 697 (7th Cir. 2002); Bolden v. Walsh Grp., No. 06 C 4104, 2012 WL 1079898, at *2 (N.D. Ill. Mar. 30, 2012). For that reason, courts have rejected statistical analyses that draw a labor market in an arbitrary manner without factoring in workers' commuting distance. See Miniature Lamp Works, 947 F.2d at 302; Bennett, 295 F.3d at 697; Bolden, 2012 WL 1079898, at *2. The defendants argue that Bendick's construction of his reasonable recruitment area analysis is an example of such arbitrary line-drawing.
The Court disagrees that Bendick drew his reasonable recruitment area arbitrarily. In Miniature Lamp Works and Bennett, the plaintiffs' experts drew their labor markets based on nothing more than educated guesses. Miniature Lamp Works, 947 F.2d at 301-02 & n.8 (“boundaries of the relevant labor market were arbitrarily limited to the City of Chicago” and “considerations of applicant preference were totally ignored”); Bennett, 295 F.3d at 697 (rejecting whole state as relevant labor area based on expert's unverified assertion that “it seemed reasonable to expect that the School District would hire applicants from within the state”). By contrast, Dr. Bendick drew his reasonable recruitment area based on staffing-agency invoices, which meant that every data point constituting his recruitment area stemmed from a real person who embarked upon a commute to a staffing agency implicated in this case. Bendick also understood that staffing agencies often provided vans for workers, which supported his choice of data. His reasonable recruitment area may have been somewhat overinclusive, especially given defendants' evidence about the sporadic availability of van transportation, but that does not mean that it was arbitrarily drawn. That Bendick's choice to define the reasonable recruitment area using the addresses of staffing-agency workers “could be challenged as incomplete or faulty does not make it any less grounded in real data.” See Manpower, 732 F.3d at 809.
Bendick also culled the far-away outliers from his data set by cutting off PUMAs once he accounted for ninety-five percent of worker addresses, a procedure he testified was “typical” among economists, professional peers, and the reference works that he consulted. Bendick Decl. ¶ 17 n.14, ECF No. 570-2; Bendick Dep. at 169-70, ECF No. 568-2. Defendants argue that Bendick's choice of ninety-five percent (as opposed to some other number), rendered his whole analysis arbitrary, but Bendick explained why he chose that number. The point of the cut-off, moreover, was to render the data set more accurate by narrowing the area from which employers “usually” draw workers See Zollicoffer, 335 F.R.D. at 150 (“Including the PUMAs that account for less than 5% of referrals would do little to define the area from which an employer usually draws workers.”). No statistical analysis, especially one that measures a geographic area as nebulous as a labor market, can perfectly define where an employer “usually” recruits. As Bendick testified, he had to “draw a line somewhere” and in doing so, he relied on professional consensus. Bendick Dep. at 173, ECF No. 568-2. Whether he drew that line correctly is a matter for the jury, not the Court to determine. Manpower, 732 F.3d at 809.
The defendants in Gold Standard Baking also moved to strike Bendick's expert testimony, although on different grounds. Zollicoffer, 335 F.R.D. at 145-52. Judge Ellis denied that motion, reasoning that his estimates were sufficiently reliable for the court to consider as part of plaintiffs' class-certification motion. Id. at 152.
The Court further disagrees that Bendick entirely failed to consider commuting distance in this analysis. Instead of disregarding distance altogether, Bendick described why considering distance as a matter of mileage from Vee Pak was not an appropriate proxy for interest in temporary work. First, Bendick explained that any account of commuting times without considering race would underestimate the willingness of African Americans to accept positions with far-away employers. In support, Bendick referenced his own independent research, which examines the disproportionately long commuting times of African Americans to staffing agencies located in Chicago. He also explained, citing the “professional standard” codified in the Office of Federal Contract Compliance Programs' regulations, that considering distance from Vee Pak would bake racial inequities into the analysis. Bendick Decl. ¶¶ 13 n.10, 26, ECF No. 570-2 (quoting 41 CFR § 60-2.14(e)). Assuming plaintiffs' allegations of racial discrimination are true, if Bendick based his findings only on PUMAs reflecting addresses of Vee Pak employees, those addresses would be disproportionately drawn from areas in which African Americans do not live. See 41 CFR § 60-2.14(e) (“[A] contractor may not draw its reasonable recruitment area in such a way as to have the effect of excluding minorities or women.”); see also Mister v. Illinois Cent. Gulf R. Co., 832 F.2d 1427, 1433-34 (7th Cir. 1987) (noting that “[d]iscrimination itself produces a statistical finding of no discrimination” if model based on distance ignores differences in racial demographics in different areas).
Finally, Bendick explained that accounting for a worker's commute would also require the introduction of too many variables, such as cost, travel time, and the availability of public transportation and ridesharing. Merely considering distance as a matter of mileage, according to Bendick, would reduce, rather than increase accuracy, so excluding the variable altogether would render more reliable results. In his rebuttal declaration, Bendick provided examples in which staffing-agency workers living in PUMAs that were located further from Vee Pak supplied fewer wages invoiced to Vee Pak than ones that were closer to Vee Pak (and vice versa). With specific examples, Bendick explained why measuring distance as a matter of milage is an inartful measure of worker interest in Vee Pak.
Defendants also argue that Bendick's rebuttal improperly advances plaintiffs' “case in chief.” Staffing Network Memo in Supp. of Mot. to Strike at 13-14, ECF No. 568; Vee Pak Memo in Supp. of Mot. to Strike at 26-27, ECF No. 570. The Court disagrees. The Court already gave Plaintiffs leave to file Bendick's rebuttal because it found that the report responded to defendants' experts' report. Order Granting Plaintiff's Leave, ECF No. 540. Bendick's rebuttal attempted to address the flaws in defendants' experts' criticisms and apply those criticisms to his own analysis, both proper issues to address in a rebuttal.
Far from basing his opinion “only on his ipse dixit,” Bendick connected the data he employed (or in this case, declined to employ), to the opinion he offered. Manpower, 732 F.3d at 806 (quoting General Electric Co. v. Joiner, 522 U.S. 136, 146 (1997)). Instead of “totally ignor[ing]” worker interest as measured by distance from Vee Pak, Bendick backed up his reasoning with facts and data. Cf. Miniature Lamp Works, 947 F.2d at 301. To be sure, none of Bendick's explanations for rejecting the incorporation of distance from Vee Pak in his analysis are ironclad. As the defendants' experts show, Bendick's analysis includes PUMAs from which workers would have to embark upon unreasonably long commutes. But Bendick's explanation for his choices is nonetheless rationally related to his opinion, rendering his opinion admissible, notwithstanding an opposing expert's disagreement. See Walker v. Soo Line R. Co., 208 F.3d 581, 589 (7th Cir. 2000). Defendants may attack Bendick's explanations for his exclusion of distance on cross examination, but it is not for the Court to reject them at this stage. See Schultz, 721 F.3d at 432.
The defendants' experts offer competing variables that they claim would better account for workers' interest in Vee Pak as measured by their commuting costs. That defendants' experts thought of other methodologies does not render Bendick's methodology automatically unreliable. “[I]t is often the case that experts reach conflicting conclusions based on applying different but nevertheless reliable methodologies to a set of partially known facts.” Baugh v. Cuprum S.A. de C.V., 845 F.3d 838, 847 (7th Cir. 2017). Tellingly, none of the defendants' experts cite to any professional consensus or reference point that guides how an expert must determine a worker's interest in the labor market.
For example, Vee Pak's experts Dr. Edward Anderson and Dr. Catherine Massey measured the expected “intensity of interest” in commuting to Vee Pak and to each staffing agency in each PUMA. Anderson and Massey Decl. ¶ 30 & n.12, ECF No. 570-6. This “intensity” was based on ACS survey response data and measured by the percentage of respondents who expected to choose a commute from the centroid of a PUMA to Vee Pak or the relevant staffing agency as compared to alternative work or a current job. Although Anderson and Massey reference undefined “economic principles” that “require” the inclusion of their chosen variables in the analysis, that does not give the Court a sufficient basis to conclude that their method is the economic gold standard, or even that it is more reliable than Bendick's. Id. ¶ 23. Anderson and Massey's testimony certainly does not go so far as to show that Bendick's exclusion of the variables that they chose was based on mere say-so or speculation. See Metavante Corp. v. Emigrant Sav. Bank, 619 F.3d 748, 761 (7th Cir. 2010).
To this point, Bendick explained in his rebuttal why he thought Anderson and Massey's approach was wrong-namely, that it failed to incorporate racial differences in willingness to travel and excluded unemployed people, the very people most likely to seek temporary work, from the analysis. The Court is thus left with a battle of expert conclusions, one that the defendants cannot win using Daubert as a weapon. See In re Dealer Mgmt. Sys. Antitrust Litig., 581 F.Supp.3d 1029, 1055 (N.D. Ill. 2022) (denying Daubert motion, where “both sides offer[ed] seemingly reasonable arguments, supported by testimony from well qualified and respected economists, concerning why [their conclusions were] appropriate”). Without a firm basis for concluding that Bendick must account for certain factors that pertain to a worker's commute to render his methodology reliable, defendants' attacks on his methodology ring hollow.
To be clear, the Court does not impose upon defendants the burden of presenting a competing statistical model to refute Bendick's conclusion. The plaintiffs make much of Mister v. Illinois Central Gulf Railroad Company, 832 F.2d 1427 (7th Cir. 1987). In that case, a plaintiff class of African American railroad workers established, through statistical evidence, a high probability of disparate treatment in violation of Title VII. Id. at 1429. The Seventh Circuit held that the defendant did not meet its burden of proving, at trial, that its own race-neutral explanation for hiring, that it wanted to hire workers close to the railroad, explained the disparity. Id. That case does not, as the plaintiffs argue, state that a party moving to strike an expert bears the burden to show that an omitted variable accounts for an opposing expert's result. Indeed, the Seventh Circuit has squarely held, after Mister, that a defendant “need not in include in its own statistical analysis variables that it asserts are missing from the plaintiff's analysis.” Miniature Lamp Works, 947 F.2d at 301. The Court's point is rather that the variable that defendants assert is crucial-measuring the relative cost of workers' commute to Vee Pak-was not ignored, but rather considered, in Bendick's analysis.
B. Other Factors
Defendants' remaining criticisms of Dr. Bendick's analysis focus on the quality of the data that Bendick used and other variables that Bendick omitted. As discussed below, all of these arguments implicate the weight, rather than the admissibility, of Bendick's testimony.
The defendants argue that Bendick used data from the wrong agency offices in his analysis. According to Staffing Network, Bendick included data from Staffing Network's Cicero office, even though witness testimony reflects that Staffing Network almost never sent workers from that office to Vee Pak. And Vee Pak argues that Bendick erroneously included workers from ASI's Aurora office in his model even though ASI's President testified that workers were not sent to Vee Pak from that office either. Plaintiffs counter that exclusion of this data either would not have made a difference in Bendick's analysis (Aurora) or would have incorporated racial discrimination into the analysis by excluding an office located near a predominantly African American neighborhood (Cicero). These arguments are precisely the type that are well suited for a jury's consideration; “[w]hether [the expert] selected the best data set to use . . . is a question for the jury, not the judge.” Manpower, 732 F.3d at 809.
Defendants' arguments that Bendick used the addresses of workers from the wrong time periods to build his PUMAs are similarly unpersuasive. Staffing Network argues that Bendick included some data from residents working at Staffing Network outside the class period in fashioning his PUMAs. It also challenges Bendick's use of addresses of workers who were invoiced late in the class period when determining the reasonable recruitment area for earlier years. Those choices may be questionable, but they are matters of data selection and do not bear upon admissibility. Indeed, Judge Ellis rejected almost identical arguments in the Gold Standard Baking case. Although Judge Ellis recognized defendants' valid criticisms about Bendick's failure to vet the data files that counsel provided, she noted that those “critiques go to the weight of Dr. Bendick's analysis, not its admissibility.” Zollicoffer, 335 F.R.D. at 148-49.
Finally, the defendants argue that in fashioning the expected representation of African Americans in each PUMA, Bendick failed to consider a host of factors that would influence a worker's interest in Vee Pak. Bendick erred, they argue, in assuming that every worker with a certain level of education in the work force making below the median income would be interested and available for a job at Vee Pak. For instance, Staffing Network argues, that individuals with certain occupations making less than a median income, such as nurses and librarians, would not consider low wage temporary work at Vee Pak desirable, citing its expert, Dr. Paul White. Similarly, Vee Pak argues that Bendick should have incorporated other factors such as a worker's access to and availability of public transportation, the cost for gas or bus tickets, and employees' shift preferences.
Notably, White did not explain whether “restrict[ing] the population to occupations that may be more likely to be interested in a laborer job” is an accepted methodology or whether it is standard practice to do so. White Decl. at 24-25, ECF No. 568-10. He simply opined that Bendick's expected representation included occupations that would be unlikely to be interested in Vee Pak. Id. at 3.
It can be argued, to be sure, that Bendick's data inflates the number of workers who might be interested in a temporary position in Vee Pak's warehouse, but that argument is also best reserved for a jury. Even though Bendick's estimate of each PUMA's expected representation of NHAA's is, to put it mildly, “rough,” the Court must “let the jury determine how the uncertainty about [the expected representation of African Americans in each PUMA] affected the weight of [Bendick's] testimony.” Stollings v. Ryobi Techs., Inc., 725 F.3d 753, 76667 (7th Cir. 2013). Defendants' criticisms challenge Bendick's choice of variables in determining worker interest, a choice that, again, concerns the probative value rather than the admissibility of testimony. Chicago Tchrs. Union, Loc. 1 v. Bd. of Educ. of City of Chicago, No. 12 C 10311, 2020 WL 914881, at *16 (N.D. Ill. Feb. 25, 2020).
The defendants' criticisms of Dr. Bendick's interest criteria also overstate the significance of the criteria used to define the pool of people in each PUMA who were interested in temporary work at Vee Pak. Those criticisms address only the effect of the criteria used on the total number of persons interested in Vee Pak rather than on the percentage of NHAAs within that group. Further, to the extent that the defendants complain that Dr. Bendick's criteria are overinclusive, they may understate the degree of discrimination. The defendants complain, for example, that Bendick's pool is too broad because it would include “nurses and librarians” who presumably would not be interested in temporary work at Vee Pak. But if, hypothetically, “nurses and librarians” are disproportionately white, excluding them from the criteria used to determine the pool of interested workers would increase the expected percentage of “interested” NHAAs in each PUMA, consequently increasing the delta between expected and actual representation of NHAAs at Vee Pak.
This is so even considering the Seventh Circuit's emphasis on the interest of potential applicants in a labor-market analysis. See, e.g., Bennett, 295 F.3d at 697. An expert cannot wholly ignore an applicant's interest in the job, but Bendick did not do so, and nothing in the caselaw dictates a formula an expert must follow or which specific individual variables an expert must use. For instance, the court in E.E.O.C. v. Consolidated Service Systems, 777 F.Supp. 599, 606 (N.D. Ill. 1991), aff'd, 989 F.2d 233 (7th Cir. 1993), found an expert's statistical evidence lacking where the expert “failed to take into account those persons who would not desire a position at Consolidated given the particular working conditions offered.” But here, Bendick did consider the desirability of a position at Vee Pak, reasonably if imperfectly, by limiting the pool of expected representation through median income, position in the labor force, and level of education.
The Court notes, moreover, that Consolidated Systems tested the sufficiency of evidence after a bench trial, rather than ruling on the admissibility of evidence on a Daubert motion. 777 F.Supp. at 609-10. The Court does not imply, by ruling Bendick's declaration admissible, that it would sufficiently or even substantially prove plaintiffs' case on the merits.
It would be impossible for an expert to include every conceivable variable pertaining to worker interest-such as the vagaries of public transportation schedules or workers' level of comfort with waiting alone at a bus stop-in a labor market analysis. This impossibility does not, as defendants imply, mean that a labor market that considers workers' interests can never be reasonably approximated. See Bazemore v. Friday, 478 U.S. 385, 400 (1986) (“[I]t is clear that a regression analysis that includes less than ‘all measurable variables' may serve to prove a plaintiff's case.”). Instead, as long as an expert uses a reliable methodology, whether the expert used the right variables and the conclusions that can be drawn from those limited variables is a question for the jury. Smith v. Ford Motor Co., 215 F.3d 713, 718-19 (7th Cir. 2000).
In sum, Bendick considered African Americans' potential interest in the job, drew his relevant labor market based upon real data, and applied a regression methodology that courts have repeatedly endorsed. Although his results are not bulletproof, the Court finds that any deficiencies can be addressed on cross examination. Defendants' motion to strike Bendick's testimony is therefore denied.
II. Motion for Class Certification
To certify a class, the Court must first consider whether it meets the requirements of Rule 23(a). That section requires that “(1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class.” Fed.R.Civ.P. 23(a). In interpreting the requirement that a class be “defined,” Courts have added that a class must be ascertainable, or “defined clearly and based on objective criteria.” Mullins v. Direct Digital, LLC, 795 F.3d 654, 659 (7th Cir. 2015). If a proposed class meets Rule 23(a)'s prerequisites, a court must examine whether one of Rule 23(b)'s provisions is satisfied. Here, plaintiffs seek certification pursuant to Rule 23(b)(3), which requires that “questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.”
Rule 23(c)(5) further provides that, “[w]hen appropriate, a class may be divided into subclasses that are each treated as a class under [Rule 23].” Plaintiffs accordingly seek certification of each proposed subclass under Rules 23(b)(3) and (c)(5).
A. Rule 23(a) Requirements
1. Numerosity
Defendants do not dispute that the proposed subclasses satisfy Rule 23(a)'s numerosity requirement. Plaintiffs estimate that the MVP class will have 2,896 members, the Staffing Network class will have 5,889 members, and the ASI class will have 4,791 members. Those numbers far exceed the threshold that courts have held would make “joinder impracticable.” Fed.R.Civ.P. 23(a)(1); see, e.g., Swanson v. Am. Consumer Indus., Inc., 415 F.2d 1326, 1333 n.9 (7th Cir. 1969) (forty people sufficiently numerous); Phillips v. Waukegan Hous. Auth., 331 F.R.D. 341, 350 (N.D. Ill. 2019) (same). The Court finds the numerosity requirement satisfied.
2. Commonality
Rule 23(a) next requires that there be “questions of law or fact common to the class.” Fed.R.Civ.P. 23(a)(2). “Where the same conduct or practice by the same defendant gives rise to the same kind of claims from all class members, there is a common question.” Suchanek v. Sturm Foods, Inc., 764 F.3d 750, 756 (7th Cir. 2014) (collecting cases). Plaintiffs allege that employees at the staffing-agency offices declined to assign African Americans, adhering to the same assignment practices and pursuant to discriminatory policies over which the employees had no discretion. See Lucas, 2017 WL 6733688, at *5. When this Court preliminarily approved the ASI settlement class and when it denied the defendants' motion to strike class allegations, it held that the question of whether a discriminatory policy of steering African Americans away from Vee Pak existed satisfies the commonality requirement. Id. (“A company-wide discriminatory practice that was applied to all class members is sufficient to establish commonality.”); Lucas, 68 F.Supp.3d at 881 (“[A discriminatory] policy would be a common cause of all the class members' injuries.”). Judge Ellis concluded the same in the Gold Standard Baking case. Zollicoffer, 335 F.R.D. at 157. The defendants' arguments here do not persuade the Court to conclude otherwise.
Defendants argue that the Supreme Court's recent holding in Comcast Corp. v. National Association of African American-Owned Media, 140 S.Ct. 1009 (2020), changes the calculus of this Court's previous decisions in this case and Judge Ellis's decision in Gold Standard Baking. In Comcast, the plaintiffs argued that at the pleading stage of a § 1981 case, they needed only to allege that racial discrimination was a “motivating factor” rather than the “but-for” cause of the defendant's challenged decision. 140 S.Ct. at 1014. The Supreme Court rejected that argument, holding that plaintiffs alleging discrimination under § 1981 “must initially plead and ultimately prove that, but for race, [they] would not have suffered the loss of a legally protected right.” Id. at 1019. This burden, the Court held, “remains constant” throughout the lawsuit. Id. at 1014-15.
This Court decided to certify ASI before the Supreme Court issued Comcast, and the parties did not brief the impact of Comcast before Judge Ellis in the Gold Standard Baking case.
According to the defendants, Comcast precludes § 1981 plaintiffs, like the ones in this case, from proceeding under the framework that the Supreme Court established in International Brotherhood of Teamsters v. United States, 431 U.S. 324 (1977). Under that approach, a class of plaintiffs alleging a pattern or practice of discrimination may submit for class-wide resolution the question of liability, or whether the defendants indeed operated under a discriminatory policy. Teamsters, 431 U.S. at 360-61. Courts then determine the scope of individual relief in later, individualized proceedings in which the defendant would bear the burden to show that “the individual applicant was denied an employment opportunity for lawful reasons.” Id. at 361. The plaintiffs propose a similar approach in this case: a class-wide action to determine whether the defendants had a policy of steering African American workers away from Vee Pak, followed by later, individualized proceedings to determine the scope of plaintiffs' damages.
After Comcast, the defendants contend, a Teamsters-based approach to § 1981 cases impermissibly shifts the burden of proof to them. The defendants argue that because each individual plaintiff must prove that the employer would still have hired that plaintiff if not for her race, the question of the defendant's liability does not rise and fall with proof of a common policy. And defendants argue that they cannot bear the burden of proving that they still would have hired the plaintiffs at the second step of the Teamsters framework; that burden falls to plaintiffs.
The Court is skeptical that Comcast, a case about mixed-motive pleading standards for § 1981 cases, reaches so far as to foreclose application Teamsters framework to establish defendants' liability. Cf. Chicago Tchrs. Union, Loc. 1, Am. Fed'n of Tchrs., AFL-CIO v. Bd. of Educ. of City of Chicago, No. 12 C 10311, 2021 WL 1020991, at *16 (N.D. Ill. Mar. 17, 2021) (applying Teamsters framework to § 1981 claim post-Comcast). After all, when the Supreme Court decided Teamsters, a Title VII case, in 1977, the “but-for” standard, rather than the “motivating factor” standard, governed such claims. Comcast, 140 S.Ct. at 1017 (explaining that the Supreme Court introduced the motivating factor standard in 1989). The Court in Teamsters was not concerned that its burden shifting framework could run afoul of Title VII's then existing but-for causation standard.
Regardless, the Court need not decide whether Comcast forecloses a bifurcated approach to establishing but-for causation. Proving this case at trial will not require any burden shifting to establish the defendants' liability. Apart from the common question of the existence of a discriminatory policy, the other important common question in this case is whether there were any real qualifications for the positions open at Vee Pak. Plaintiffs assert, and have provided some evidence to show, that the answer to that question is no: the work available at Vee Pak was unskilled and required no prior experience. And any of Vee Pak's claims that they desired “experienced” or “regular” workers, Plaintiffs, assert, were merely pretexts for racial discrimination.
Of course, the defendants vigorously maintain that a workers' level of experience mattered to Vee Pak and have provided their own evidence to refute plaintiffs' claims. Vee Pak's allegedly pretextual labeling of workers as “experienced” or “regulars,” however, is part of the discriminatory policy that plaintiffs claim defendants perpetuated. As discussed below, the plaintiffs need not definitively prove that the policy existed to show that the requirements for class certification are met. At this stage, plaintiffs need only show that they could prove their case using common evidence. See Messner, 669 F.3d at 817. True, even absent a discriminatory policy that used experience as a proxy for race, individual plaintiffs could argue that they were subject to discrimination. But such individual claims rooted in idiosyncratic discrimination “would be based on an entirely different legal theory.” Bell v. PNC Bank, Nat. Ass'n, 800 F.3d 360, 378 (7th Cir. 2015).
Thus, if a jury finds true plaintiffs' claims that (1) defendants operated under a discriminatory policy and (2) every worker was equally qualified for a position at Vee Pak, it is no defense to liability to assert that a plaintiff was not hired due to that plaintiff's qualifications. Defendants' liability can be resolved in one centralized proceeding. As the Court previously noted when preliminarily certifying the ASI settlement class, there may be a need for subsequent individualized inquiries to determine which class members were actually adversely affected by the discriminatory practice. Lucas, 2017 WL 6733688, at *5. “Plaintiffs need not prove that every member of the proposed class has been harmed before the class can be certified;” if each member of the class could have been harmed by a class-wide policy, certification is appropriate. Bell v. PNC Bank, Nat. Ass'n, 800 F.3d 360, 380 (7th Cir. 2015). Because plaintiffs' proposed class would allow for class-wide resolution of liability issues under any reading of Comcast, defendants' arguments about the effect of that case are unpersuasive.
The defendants also argue that plaintiffs have not proffered “significant proof” of a uniform, general policy that would link their claims of discrimination together. Wal-Mart, 564 U.S. at 353. At the class-certification stage, plaintiffs must offer more than allegations about a discriminatory practice. See id. at 350. Instead, they must offer evidence to show that the existence of the policy in the case is “capable of proof at trial through evidence that is common to the class.” Bell, 800 F.3d at 375 (quoting Messner, 669 F.3d at 818). For example, evidence that company decisionmakers impute their own biases when making discretionary employment decisions is not “significant proof” of such a policy. See Walmart, 564 U.S. at 356-57. Because the plaintiffs in Wal Mart offered only proof that employers' actions were discretionary, it would be impossible to rely on that evidence to show that the company engaged in discriminatory conduct. Id. at 357. By contrast, if plaintiffs offer enough evidence to show that questions common to the class exist, they need not prove the “truth of their allegation that [the defendant] has an unofficial policy.” Bell, 800 F.3d at 374, 377 (noting that district court must resolve only material factual disputes that bear on the requirements for class certification”) (emphasis in original).
The plaintiffs have satisfied their burden here. Employees at all three staffing agencies testified to the defendants' discriminatory practices. This testimony included first-hand accounts of discriminatory conduct at all levels of management, including that Vee Pak told staffingagency managers not to refer African Americans, that staffing-agency employees perpetuated these policies, that staffing agencies used derogatory code words to refer to African American workers, and that the few African Americans who did manage to by-pass the staffing agencies were turned away at Vee Pak's door. Some of these employees testified that Vee Pak shift supervisors directly and repeatedly instructed them not to send African Americans to Vee Pak. The named plaintiffs, moreover, testified to the staffing agencies' passing them over for jobs due to their race. The testimony about specific mechanisms of discrimination, from the upper echelons of staffing-agency management to the staffing-agency dispatchers to the employees facing the effects of discrimination, undoubtedly speaks to the existence of a discriminatory policy among the defendants. See id., 800 F.3d at 375 (testimony from company branch managers about discriminatory policy constituted enough evidence to satisfy commonality requirement).
The defendants protest that plaintiffs' anecdotal evidence is not substantial enough, citing Wal-Mart. In that case, 120 employees' testimony about their own experiences of discrimination was not enough to establish a wide-ranging policy spanning 3,400 stores employing millions of workers in all fifty states. Wal-Mart, 564 U.S. at 358. Without any testimony that gave color to the alleged policy, the employees' scattered testimony about the discrimination they suffered was of course not enough to form the “glue” holding together the alleged reasons for the discrimination. Id. Here, the plaintiffs furnish testimony from defendant employees, some of whom were in managerial positions, about the discriminatory practices and the uniform ways in which they were implemented. This testimony stands in addition to the testimony of the named class members and proposed class members about their own experiences of discrimination. Class members' testimony about their personal experiences of discrimination was not enough, standing alone, to establish a discriminatory policy in Wal-Mart, but first-hand accounts about the policy's contents and implementation would, if true, allow a jury to find a policy existed in this case.
Defendants also argue that some of the witness testimony predates the class period and is irrelevant to the existence of a policy during the alleged period. Outside-the-class period evidence alone would not suffice to show that discriminatory conduct common to all class members occurred within the class period. Combined with other within-class-period testimony, however, testimony about a policy that began before the class members' claims accrued and persisted throughout the class period is relevant. See Nat'l R.R. Passenger Corp. v. Morgan, 536 U.S. 101, 112-13 (2002) (evidence of discriminatory policy outside of statute of limitations relevant as “background evidence to support a timely claim”). For similar reasons, the Court finds unpersuasive the defendants' arguments that one staffing agency's practices of steering workers away from Vee Pak are irrelevant to another agency's same practices. And the staffing agency's discriminatory policies at the behest of non-Vee Pak clients are also relevant. Plaintiffs allege that Vee Pak asked staffing agencies not to refer African Americans and that staffing agencies followed those discriminatory policies. Although not dispositive, that one staffing agency or client perpetuated this discriminatory policy renders it more plausible that the policy extended to other agencies or other clients.
The defendants' remaining objections to plaintiffs' witness testimony rest on disputes about witness credibility and the sources of the witness's knowledge. For example, some staffing-agency employees who testified about Vee Pak's discriminatory instructions, such as Cruz, Martinez, and Armendariz, also testified that they did not follow those instructions. But plaintiffs have offered testimony that casts doubt upon those statements. See Medina Dep. at 43, ECF No. 544-25 (testifying that Cruz told him that he could not refer African American workers to Vee Pak); Martinez Dep. at 45, 54-55, 70-71, ECF No. 544-27 (describing Staffing Network's practice of hiring “regulars,” and testifying that, although Cruz did not want her to “find more Latinos,” that Cruz knew about the policy and told her “there's nothing we can do”); Armendariz Dep. at 39-44, ECF No. 551-1 (noting same “regulars” policy and noting regulars were predominately Latino). The Court cannot resolve a textbook credibility dispute by reading deposition transcripts. At the class-certification stage, courts may not “weigh[ ] . . . evidence to determine whether the plaintiff class will prevail on the merits.” Bell, 800 F.3d at 377.
It is true that a court must sometimes delve into the merits when considering class certification. See Szabo v. Bridgeport Machs., Inc., 249 F.3d 672, 676 (7th Cir.2001). Such inquiries are only appropriate, however, to the extent they are necessary to determine whether the class certification requirements are met (in this case, whether the issue can be resolved on a class-wide basis through evidence common to the class). See Bell, 800 F.3d at 376-77; see also Schleicher v. Wendt, 618 F.3d 679, 685 (7th Cir. 2010). The plaintiffs have presented enough evidence to show that the resolution of defendants' claims of discrimination will rise and fall with the truth of witness testimony about the existence of a discriminatory policy. In other words, if a jury does not believe the witness testimony about the policy, the “certified class can go down in flames on the merits.” Schleicher, 618 F.3d at 685. The plaintiffs' have proffered enough witness testimony to support a legal theory common to the entire class; that is precisely what the Supreme Court meant by “significant proof” in Wal-Mart.
The plaintiffs, moreover, have corroborated their anecdotal evidence with Dr. Bendick's statistical analysis, which as explained above, the Court deems admissible. As this Court has previously explained, a statistical analysis alone does not provide an answer to the cause of a racial disparity. Jones v. Nat'l Council of Young Men's Christian Ass'ns of the United States of Am., 34 F.Supp.3d 896, 909 (N.D. Ill. 2014). But such an analysis “may go a long way to establishing that there is a race-based disparity, ruling out the possibility that an observed disparity is simply the product of chance.” Id. Bendick's analysis, therefore, renders it more likely that something caused a racial disparity, and witness testimony suggests that a discriminatory policy was that something. Id. Using that evidence-which is common to every member of the class-the plaintiffs seek to prove their case. A jury does not have to find Bendick's numbers convincing, and the defendants may rely, on the merits, on all of the deficiencies that they raised in their Daubert motion. Similarly, a jury does not have to find convincing the testimony of plaintiffs' witnesses. In any event, however, a jury can weigh this evidence and decide whether all of the plaintiffs win or lose. As Judge Ellis put it, “the statistical evidence and anecdotes together provide ‘significant proof' of a ‘uniform employment practice' that caused a common injury to the class.” Zollicoffer, 335 F.R.D. at 157 (citing Walmart, 564 U.S. at 353-55). Plaintiffs have satisfied the commonality requirement.
3. Typicality
Rule 23(a) requires that “the claims or defenses of the representative parties are typical of the claims or defenses of the class.” “A claim is typical if it ‘arises from the same event or practice or course of conduct that gives rise to the claims of other class members and . . . [the named plaintiffs'] claims are based on the same legal theory.'” Oshana v. Coca-Cola Co., 472 F.3d 506, 514 (7th Cir. 2006) (quoting Rosario v. Livaditis, 963 F.2d 1013, 1018 (7th Cir. 1992)). Factual differences between the plaintiffs do not destroy typicality as long as plaintiffs proffer similar legal theories. See De La Fuente v. Stokely-Van Camp, Inc., 713 F.2d 225, 232 (7th Cir. 1983). It is the legal theories supporting those injuries that must be typical of plaintiffs' claims; the plaintiffs may have claims typical of other class members even if the extent of class members' injuries differs. See id. at 232-33.
Here, named plaintiffs Zollicoffer, Eagle, Keys, and Franklin, allege that they were denied work after waiting unsuccessfully for assignments and that the staffing agencies gave work to Latino laborers instead. They allege that defendants similarly denied class members work opportunities at Vee Pak due to their race pursuant to a discriminatory policy. That is, the named plaintiffs have alleged the same injury as the proposed class members: racial discrimination. The named plaintiffs also assert the same legal theory as the proposed class members, that defendants violated § 1981. Plaintiffs have therefore satisfied the typicality requirement. Chicago Tchrs. Union, Loc. 1 v. Bd. of Educ. of the City of Chicago, 307 F.R.D. 475, 481 (N.D. Ill. 2015) (typicality satisfied where “all of the proposed class representatives base[d] their claims on the same legal theory: racial discrimination”).
Defendants nonetheless argue that the named plaintiffs have not accrued the same injuries as the proposed class members. Staffing Network argues that named plaintiffs' injuries are atypical because the staffing agencies eventually assigned Eagle and Keys to customers other than Vee Pak. This argument confuses the kind of injury that named plaintiffs allege with the extent of their injuries. See Pruitt v. Pers. Staffing Grp., LLC, No. 16-CV-5079, 2020 WL 3050330, at *8 (N.D. Ill. June 8, 2020) (“Defendants' argument about the short amount of time the Named Plaintiffs actually sought assignments at MVP and their availability to take assignments from MVP may go to the Named Plaintiff's damages, but those are not defenses to claims of racial discrimination.”). Regardless of the extent of pecuniary losses that named plaintiffs may have ultimately incurred, all the named plaintiffs allege that they were denied a job due to racial discrimination. Zollicoffer, 335 F.R.D. at 157 (rejecting defendants' contention that Zollicoffer suffered atypical injuries because he was “unavailable to work for significant segments of the class period”). Defendants' argument that some of the named plaintiffs only sporadically sought assignments during the class period are similarly relevant only to the extent of plaintiffs' harms. The named plaintiffs and proposed class members therefore all suffered a typical injury.
Vee Pak argues that Franklin cannot serve as a named plaintiff for ASI because she was denied a job in 2014, a period after which, defendants claim, ASI took steps to correct discriminatory practices. In support, defendants cite the testimony of Steven Swerdloff, ASI Chairman who testified that ASI threatened to drop the Vee Pak account after a lawsuit challenged its discriminatory practices. But Swerdloff did not give a clear answer as to whether ASI made those changes in 2013, 2014, or 2015. Swerdloff 30(b)(6) Dep. at 103, ECF No. 573-9 (“I think it would have to be around 2015, '14 '15 . . . somewhere in there.”). Plaintiffs, in turn, argue that ASI continued to discriminate against African Americans well beyond 2014. Swerdloff had previously testified that Vee Pak continued to ask ASI to discriminate against African Americans until 2015. And an ASI employee testified that a Vee Pak employee told her in no uncertain terms to turn Black employees away in the same month that Franklin applied for work, September 2014. Plaintiffs have proffered enough evidence to show that Vee Pak's alleged discriminatory policy continued when Franklin sought work from ASI. Franklin corroborates this testimony through her statements that she sought and did not receive work from ASI during the same period. Because Franklin testified that defendants injured her in the same manner-through a discriminatory policy-as proposed class members during the relevant period, her injuries are typical. Defendants' attacks on typicality of injury do not persuade the Court.
4. Adequate Representation
Defendants next challenge the adequacy of named plaintiffs' representation. Rule 23(a)(4) requires that “the representative parties . . . fairly and adequately protect the interests of the class.” Typicality and adequacy operate in tandem; a named claimant who suffers injuries typical of the class will presumably be incentivized to vindicate its effort. See Amchem, 521 at 626 n.20. Therefore, to meet the adequacy requirement, “[a] class representative must be part of the class and possess the same interest and suffer the same injury as the class members.” Id. at 625-26 (cleaned up). The rationale animating the adequacy requirement is to ensure that the named plaintiffs will vigorously advocate for a positive outcome for class claimants. See Lacy v. Cook County, 897 F.3d 847, 866-67 (7th Cir. 2018).
Rule 23(a)(4) also requires that class counsel be adequate representatives for the class. Defendants do not challenge the adequacy of class counsel. This Court continues to find that “class counsel are experienced and competent employment discrimination and class action plaintiffs' attorneys” and are therefore adequate representatives. Lucas, 2017 WL 6733688, at *7.
Staffing Network argues that Eagle and Keys cannot represent the Staffing Network class because they lack a sufficient interest in the case. In support, they repeat their arguments contesting the typicality of their injuries: because Eagle and Keys obtained alternative employment, Staffing Network argues, they did not suffer an injury, and have an insufficient stake in this case's outcome. Staffing Network cites Culver v. City of Milwaukee, in which a named plaintiff who sought permanent employment alleged that the defendant discriminatorily refused to hire him and other proposed class members. 277 F.3d 908, 910 (7th Cir. 2002). That plaintiff could not serve as an adequate class representative because, among other things, he had found a new, permanent position shortly after his “perfunctory” efforts to seek employment with defendant. Id. at 912. Culver, however, is inapposite; that Eagle and Keys sought temporary work meant that they were subject to repeated occurrences of racial discrimination. Even if they found other employment or stopped seeking work during the class period, Eagle and Keys both suffered injuries, and the extent of those injuries is a damages question. Pruitt, 2020 WL 3050330, at *8. Moreover, unlike Culver, who stopped seeking any relief from defendants and agreed his case was moot, Eagle and Keys both continue to seek recovery for injuries they suffered. 277 F.3d at 912. Like the proposed class members, Eagle and Keys have an interest in recovering damages for the discrimination they repeatedly suffered. Therefore, the Court finds that they are adequate class representatives.
Staffing Network also argues that Zollicoffer would not be an adequate representative for the Staffing Network class. Plaintiffs, however, do not propose Zollicoffer as a class representative, but rather submit his testimony as additional evidence of Staffing Network's discriminatory policy. The Court therefore declines to address any arguments about Zollicoffer's ability to represent the Staffing Network class.
Vee Pak, in turn, contests the adequacy of Zollicoffer as a class representative for MVP and Franklin as a class representative for ASI. According to Vee Pak, Zollicoffer and Franklin are insufficiently informed about the instant litigation to adequately pursue class members' interests. “[A]n adequate class representative must maintain only an ‘understanding of the basic facts underlying the claims, some general knowledge, and a willingness and ability to participate in discovery.'” Wahl v. Midland Credit Mgmt., Inc., 243 F.R.D. 291, 298 (N.D. Ill. 2007) (quoting Murray v. E*Trade Fin. Corp., 240 F.R.D. 392, 396, (N.D. Ill. 2006)). Vee Pak argues that Franklin has not met those minimal requirements because she lacks personal knowledge about Vee Pak and was unfamiliar with certain Vee Pak employees that counsel listed in her interrogatory responses. As for Zollicoffer, Vee Pak emphasizes that Zollicoffer testified to having no knowledge of Vee Pak, including what it did, where it was located, whether staffing agencies assigned workers to Vee Pak or why he was suing Vee Pak, “leav[ing] that up to [his] attorney.” Zollicoffer Dep. at 26-28, ECF No. 573-31.
Courts have emphasized that the bar for class representatives' knowledge about the litigation subject is modest. See, e.g., Wahl, 243 F.R.D. at 298; Murray v. New Cingular Wireless Svc., Inc., 232 F.R.D. 295, 300 (N.D. Ill. 2005); Sebo v. Rubenstein, 188 F.R.D. 310, 316 (N.D. Ill. 1999); see also Zollicoffer, 335 F.R.D. at 160 (collecting cases). On the other side of the spectrum, the instances where class members have been deemed inadequate representatives are particularly egregious. For example, class members cannot get away with submitting false discovery responses by feigning lack of knowledge about their responsibilities. See Physicians Healthsource, Inc. v. Allscripts Health Sols., Inc., 254 F.Supp.3d 1007, 1025-26 (N.D. Ill. 2017). And a named plaintiff cannot be “less than willing to participate in discovery fully and in good faith.” Pruitt, 2020 WL 3050330, at *6. Otherwise, “[f]or a plaintiff who wishes to be a class representative, “(g)eneral knowledge [of the case] and a participation in discovery are enough.” Sebo, 188 F.R.D. at 316 (quoting Kriendler v. Chemical Waste Mgmt., Inc., 877 F.Supp. 1140, 1159 (N.D. Ill. 1995)).
Franklin and Zollicoffer clear that modest bar. Franklin testified that she continuously sought work from ASI, was repeatedly denied assignments, and observed that African Americans waited for work at ASI offices while ASI sent Latino workers out on assignments. Franklin read case materials, responded to all discovery requests, and explained that she understood the substance of the lawsuit: that African Americans were denied work assignments due to their race. Similarly, Zollicoffer testified that when he applied for temporary work at MVP, he observed that MVP selected Latino workers while he was denied work. Zollicoffer acquired “general knowledge” about the case: he understood that he was suing MVP and that his lawsuit was about work-related discrimination against Black people. Sebo, 188 F.R.D. at 316. The fact that named plaintiffs do not know much about Vee Pak or its employees does not raise eyebrows. It is enough that the named plaintiffs know that they were passed over for work at an agency due to their race. There would be no way for them to ascertain, through personal knowledge, that Vee Pak or its employees was the ultimate source of discrimination. Zollicoffer, 335 F.R.D. at 159 (noting that named plaintiffs' “lack of knowledge [about GSB] is at least mitigated by the fact that they never worked at GSB, and so have no personal knowledge about GSB's conduct”).
Certainly, the named plaintiffs' minimal knowledge about this litigation and their maximal deference to their attorneys warrants consideration. And the Court recognizes that Judge Dow has deemed inadequate a named plaintiff with a lack of subject-matter knowledge or knowledge about other defendants in a similar case against MVP. Pruitt, 2020 WL 3050330, at *6. But as Judge Ellis noted in the Gold Standard Baking case, the named plaintiffs' actions have shown that they are committed to vigorously pursuing their claims against defendants. As in his case against Gold Standard Baking, Zollicoffer has been involved in this case for “several years,” has “completed discovery requests, submitted to multiple depositions, completed interrogatories, and stayed in contact with [his] counsel.” Zollicoffer, 335 F.R.D. at 160. The same goes for Franklin. And as in Gold Standard Baking, the reality of this case bears upon the Court's findings. The Court declines to deny certification of this more-than-decade-long litigation solely because the named plaintiffs-low wage workers who understandably lack legal knowledge-do not know enough about a defendant that discriminated against them through another agent. See id. The Court therefore finds that the named plaintiffs are adequate representatives.
B. Rule 23(b)(3) Requirements
1. Predominance
To meet the requirements of a 23(b)(3) class, plaintiffs must show that the common questions it identifies predominate over individual claims. Importantly, predominance does not require that individual questions be absent; “[t]he rule requires only that those questions not predominate over the common questions affecting the class as a whole.” Messner, 669 F.3d at 815. Plaintiffs demonstrate predominance “if ‘a failure of proof on the [common questions] would end the case' and the whole class ‘will prevail or fail in unison.” Bell, 800 F.3d at 376 (quoting Amgen, 568 U.S. at 460). The key question to ask when considering predominance is whether “adjudication of questions of liability common to the class will achieve economies of time and expense.” Chicago Tchrs. Union, Loc. No. 1 v. Bd. of Educ. of City of Chicago, 797 F.3d 426, 444 (7th Cir. 2015).
As with commonality, this Court has already found when preliminarily certifying the ASI's settlement class, that the common issues in this case predominate over individualized inquiries. The Court continues to find that the existence of a staffing agency's company-wide policy of steering African American workers away from Vee Pak is a key part of every class member's discrimination claim. Lucas, 2017 WL 6733688, at *7. “Although individualized issues may impact the amount of damages each class member is entitled to, that does not ‘undermine the fact that the fundamental question that predominates is a common question.'” Id. (quoting Brand v. Comcast Corp., Inc., 302 F.R.D. 201, 223-24 (N.D. Ill. 2014)). It would clearly save resources to litigate in one proceeding the questions of whether a common discriminatory policy existed and whether Vee Pak required that its workers have any qualifications, rather than having to make that determination in hundreds of cases. See Chicago Tchrs. Union, 797 F.3d at 444.
Defendants argue otherwise, again invoking Comcast. According to the defendants, individualized inquiries as to liability cannot, by definition, predominate when a plaintiff must prove that discrimination was a but-for cause of an employer's decision. 140 S.Ct. at 1014. As explained in the commonality section, that is wrong, at least in the context of this case. If plaintiffs show that a discriminatory policy existed and that jobs at Vee Pak required essentially no qualifications, “the same evidence will suffice for each [class] member to make a prima facie showing” that the class member was not hired due to race. Messner, 669 F.3d at 815 (quoting Blades v. Monsanto Co., 400 F.3d 562, 566 (8th Cir. 2005)).
Defendants also argue that the Teamsters presumption of liability once a discriminatory policy is proven is only available in Rule 23(b)(2), rather than in Rule 23(b)(3), class actions. Because plaintiffs in Rule 23(b)(2) class actions seek only equitable relief, the argument goes, a district court following Teamsters would not have to make the kind of particularized injury determinations inherent in damages actions. That argument understates the individualized inquiries that the Court in Teamsters envisioned. The district court in Teamsters was not simply instructed to hand plaintiffs certificates stating they were victims of discrimination after a discriminatory policy was found; instead, the Court recognized that the district court had to make “a substantial number of individual determinations” with respect to the scope of the remedy in step two. Teamsters, 431 U.S. at 371-72. That process would “necessarily involve a degree of approximation and imprecision.” Id. at 372. Such individualized inquiries are similarly present when determining the scope of a plaintiff's injuries and measuring them in the form of damages. The Court sees nothing inherent in individualized damages inquiries that would foreclose its application to a Teamsters-like framework, especially when Teamsters also envisioned individualized inquiries in fashioning equitable remedies.
Plaintiffs' argument, moreover, ignores Seventh Circuit precedent. The Seventh Circuit has never held that classes of plaintiffs seeking damages in employment discrimination cases cannot be certified under Rule 23(b)(3) because those cases raise too many individualized inquiries. To the contrary, “[i]t is well established that the presence of individualized questions regarding damages does not prevent certification under Rule 23(b)(3).” Messner, 669 F.3d at 815 (collecting cases). In particular, if an employment practice “is determined to be discriminatory, individualized remedies and damages may have to be determined for each plaintiff or perhaps for subclasses of plaintiffs,” but that does not preclude certification. Chicago Tchrs. Union, 797 F.3d at 444. The Court finds that predominance is satisfied.
2. Superiority and Manageability
Under Rule 23(b)(3), a claimant must also show that a class proceeding is superior to another approach. If “resolving Plaintiffs' claims through a class action would be more efficient than proceeding with hundreds of individual suits,” this requirement is satisfied. Porter v. Pipefitters Ass'n Loc. Union 597, 208 F.Supp.3d 894, 912 (N.D. Ill. 2016)
As this Court has held previously when preliminarily certifying the ASI class, “separate lawsuits by the individual class members “would needlessly require multiple courts to resolve the same liability issues.” Lucas, 2017 WL 6733688, at *8. The Court adheres to that assessment. It would be extremely inefficient for the Court to determine, in separate individualized proceedings, whether the defendants engaged in a discriminatory policy that applied to all class members. See Chicago Teachers Union, 797 F.3d at 444. And as the Court has also previously noted, a class action device is superior when, as here, class members' individual claims are too small to incentivize class members or their counsel to vindicate them. Lucas, 2017 WL 6733688, at *8. This is so even considering, as defendants argue, that various types of damages and attorney's fees are available in § 1981 cases. Proving the existence of an unwritten rule is no easy feat. Class counsel have gone to great lengths during this years-long litigation, deposing scores of witnesses and engaging an expert, to unearth evidence of a discriminatory practice. Defendants cannot seriously contend that individual recovery on a § 1981 action-even with no backpay damages cap-would sufficiently incentivize an individual plaintiff to pursue such a difficult task.
Finally, the defendants argue that individualized damages inquiries will render this case unmanageable. The Court rejects that argument again. First, the Court is hopeful that the parties will be able to finagle a method of optimizing the damages inquiry, should the defendants' liability be established. “Although the extent of each class member's personal damages might vary, district judges can devise solutions to address that problem if there are substantial common issues that outweigh the single variable of damages amounts.” Arreola v. Godinez, 546 F.3d 788, 801 (7th Cir. 2008). Of course, a district court may always, pursuant to its discretion, decide that when individualized issues common to a class are resolved, the remaining damages determinations become unruly. See Riffey v. Rauner, 910 F.3d 314, 319 (upholding district court's decision not to certify class of union workers seeking dues refunds after liability issue was resolved by Supreme Court). But at this point, especially since plaintiffs claim that assignments to Vee Pak were regularly available and did not depend on eligibility requirements, the Court expects the damages determinations to be tame enough to allow for certification.
The Court maintains its optimism even though the plaintiffs have not yet determined whether a formulaic mitigation offset to damages in this case is possible. See Order, Jan. 16, 2018, ECF No. 297. The Court's facilitation of a damages determination at the parties' behest and its rejection of one method of calculating damages does not change the fact that it would be most efficient to determine whether defendants are liable in one proceeding.
In any event, even if the Court must conduct an individualized damages inquiry as to every class member, “at least it wouldn't be necessary in each of those trials to determine whether the challenged practices were unlawful.” McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482, 491 (7th Cir. 2012). The class members' claims involve the same practice by all of the defendants-the failure to assign African Americans work at Vee Pak due to their race. Plaintiffs' allegations survive only if plaintiffs prove that defendants acted pursuant to an unlawful discriminatory policy The Court sees no downside to determining, on a class-wide basis, whether that practice existed. Id. Plaintiffs have therefore satisfied the requirements of Rule 23(b)(3).
C. Ascertainability and Class Period
In addition to meeting Rule 23's explicit requirements, a class must be ascertainable. Mullins, 795 F.3d at 657. To satisfy this requirement, the class must be defined clearly, using objective criteria that are not based upon success on the merits. See id. at 659-60. Vee Pak argues that plaintiffs' proposed class definitions are vague because they include “African American laborers who sought work assignments at [one of the staffing agencies] from which they could have been referred to Vee Pak, but on one or more occasion were not assigned to work at Vee Pak during [the class period].” Vee Pak argues that the inclusion of the word “could” adds “significant uncertainty as to who is actually included in the class.” Vee Pak Opp'n at 55, ECF No. 572. This is because, as Vee Pak has extensively argued, whether a plaintiff was qualified for work at Vee Pak is a highly individualized inquiry that would involve a significant number of variables.
As the Court has explained, plaintiffs have provided enough evidence for a jury to determine that the Vee Pak positions had no required qualifications, rendering Vee Pak's concern about individualized inquiries of no consequence. In any case, whether it would be difficult administratively to determine whether plaintiffs “could” work at Vee Pak is not the relevant question. See Mullins, 795 F.3d at 657. Instead, when determining whether a class is ascertainable, courts ask whether the definition is based on objective, clear criteria. Id. at 659-60.
As Plaintiffs correctly note, this Court determined in the context of certifying the settlement class included in the settlement between plaintiffs and ASI, that a class definition almost identical to the ones at issue here was ascertainable. Lucas, 2017 WL 6733688, at *4. The Court's approved definition, however, did not contain the clause “from which [plaintiffs] could have been referred to Vee Pak,” instead defining the class as “[t]he Class Representative and all African Americans who sought a work assignment through ASI but, on one or more occasion, were not assigned or hired to work at Vee Pak during [the class period].” Id. The Court declined to include in the class definition the term “otherwise eligible to work at Vee Pak,” noting that the class definition did not set forth any objective eligibility criteria. Id.
The phrase “could have been referred,” at first glance, seems to be another way of including undefined eligibility criteria into the class definition. This Court has already determined that the inclusion of such eligibility criteria is “essentially meaningless” given the plaintiffs' assertion that no such qualifications existed for Vee Pak jobs. Id. Plaintiffs, however, explain that they included “could have been referred” in the class definition to clarify that the definition applies to African American workers who applied to relevant branch offices. For example, only MVP's Cicero office, rather than its Waukegan office, assigned workers to Vee Pak. Only a worker applying for MVP's Cicero's office, according to plaintiffs, “could have been referred” to Waukegan.
The Court disagrees that “could have been referred” clearly denotes office location. That phrase could refer, as plaintiffs guessed, to defendants' eligibility, or the availability of open positions at Vee Pak at the time. This ambiguity could be amended, as plaintiffs propose, by clarifying that the class is limited to plaintiffs who visited the relevant offices during the class period. See Wilson v. City of Evanston, Ill., 14 C 8347, 2017 WL 3730817, at *6 (N.D. Ill. Aug. 30, 2017) (amending overbroad class definition by more precisely defining term). The Court therefore amends the class definition to read, “African American laborers who sought work assignments at [staffing agency] from offices that referred workers to Vee Pak during [the class period], but on one or more occasions were not assigned to work at Vee Pak during the class period.” These class definitions would ascertain the group of people that defendants' policies harmed while addressing the concerns about branch locations that plaintiffs raised.
Finally, Vee Pak contests the plaintiffs' class period, arguing that the class must be limited to a time during which the named plaintiffs were seeking work. Vee Pak cites Bolden v. Walsh Constr. Co., 688 F.3d 893, 895 (7th Cir. 2005), in which the Seventh Circuit held that a class could not be extended to the indefinite future when none of the named plaintiffs' employment extended past 2002. But in Bolden, the Court's problem was not intrinsically with the length of the class period. Rather, the plaintiffs had alleged that they were subjected to a hostile and racist work environment, a practice that manifested in myriad ways. Id. at 893-95. The Court explained that it could not certify such an “across-the-board” class. Id. at 895. Namely, plaintiffs who had experienced one kind of discrimination earlier in the class period could not represent plaintiffs who may have been subjected to a different kind of discrimination later in the class period. Id. In this case, by contrast, the plaintiffs allege that all persons in the class were subject to the same discriminatory hiring policy, perpetuated by Vee Pak, throughout the class period (which corresponds to the times plaintiffs contend the staffing agencies sent workers to Vee Pak). During the time that staffing agencies refused to send workers to Vee Pak pursuant to Vee Pak's policy, both named plaintiffs and proposed class members were treated alike. Vee Pak's challenge to the class period is not persuasive.
Ultimately, the Court finds unpersuasive the defendants' attacks on class certification. A jury's finding that a discriminatory policy, rather than Vee Pak's insistence on experienced workers, motivated defendants' hiring and referral decisions, would “eliminate the need for repeat adjudication of this question for determinations of damages or individual injunctive relief. Chicago Tchrs. Union, 797 F.3d at 445. The Court finds that the class definitions, as amended by the Court, meet Rule 23's certification requirements. The amended subclasses are certified.
For the reasons set forth above, the Court denies defendants' motions to strike Dr. Bendick's declaration and testimony [568] [569] and grants plaintiffs' motion for class certification [544]. The Court appoints Plaintiffs Joe Eagle and Michael Keys as Staffing Network subclass representatives, Plaintiff James Zollicoffer as MVP subclass representative, and Plaintiff Evan Franklin as ASI subclass representative and appoints plaintiffs' counsel, Joseph M. Sellers, Harini Srinivasan, Christopher J. Williams, Christopher J. Wilmes, and Caryn C. Lederer as class counsel.