Brown v. Nucor Corporation

Filed: October 8, 2009 UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT No. 08-1247 (2:04-cv-22005-CWH-GCK) QUINTON BROWN; JASON GUY; RAMON ROANE; ALVIN SIMMONS; SHELDON SINGLETARY; GERALD WHITE; JACOB RAVENELL, individually and on behalf of the class they seek to represent, Plaintiffs - Appellants, v. NUCOR CORPORATION; NUCOR STEEL BERKELEY, Defendants - Appellees. O R D E R The court amends its opinion filed August 7, 2009, as follows: On page 19, Section IV., lines 7 through 10, the third sentence of text and the case citation sentence immediately following are deleted; footnote 11 on page 19 is also deleted. For the Court – By Direction /s/ Patricia S. Connor Clerk PUBLISHED UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT QUINTON BROWN; JASON GUY;  RAMON ROANE; ALVIN SIMMONS; SHELDON SINGLETARY; GERALD WHITE; JACOB RAVENELL, individually and on behalf of the  class they seek to represent, No. 08-1247 Plaintiffs-Appellants, v. NUCOR CORPORATION; NUCOR STEEL BERKELEY, Defendants-Appellees.  Appeal from the United States District Court for the District of South Carolina, at Charleston. C. Weston Houck, Senior District Judge. (2:04-cv-22005-CWH-GCK) Argued: January 27, 2009 Decided: August 7, 2009 Before MICHAEL, GREGORY, and AGEE, Circuit Judges. Vacated and remanded with instructions by published opin- ion. Judge Gregory wrote the opinion, in which Judge Michael joined. Judge Agee wrote a separate opinion concur- ring in part and dissenting in part. 2 BROWN v. NUCOR CORP. COUNSEL ARGUED: Robert L. Wiggins, Jr., WIGGINS, CHILDS, QUINN & PANTAZIS, P.C., Birmingham, Alabama, for Appellants. Earle Duncan Getchell, Jr., MCGUIREWOODS, L.L.P., Richmond, Virginia, for Appellees. ON BRIEF: Armand Derfner, D. Peters Wilborn, Jr., DERFNER, ALT- MAN & WILBORN, Charleston, South Carolina; Ann K. Wiggins, Susan Donahue, WIGGINS, CHILDS, QUINN & PANTAZIS, P.C., Birmingham, Alabama, for Appellants. Cary A. Farris, Marc Patton, ALANIZ & SCHRAEDER, L.L.P., Houston, Texas; Erin M. Sine, MCGUIREWOODS, L.L.P., Richmond, Virginia; John S. Wilkerson, III, Nosizi Ralephata, TURNER, PADGET, GRAHAM & LANEY, P.A., Charleston, South Carolina, for Appellees. OPINION GREGORY, Circuit Judge: This case involves allegations of racial discrimination at a steel manufacturing plant in Huger, South Carolina, owned by Nucor Corporation and Nucor Steel Berkeley (collectively, "Nucor"). We find that the district court abused its discretion and erred as a matter of law in denying class certification to the plaintiffs-appellants. We therefore vacate the order and remand the case to the district court for certification. I. The allegations that the appellants present in support of their racial discrimination and hostile work environment claims speak for themselves: white supervisors and employ- ees frequently referred to black employees as "nigger," "bolo- gna lips," "yard ape," and "porch monkey." White employees frequently referred to the black employees as "DAN," which BROWN v. NUCOR CORP. 3 stood for "dumb ass nigger." These racial epithets were broad- cast over the plant-wide radio system, along with "Dixie" and "High Cotton." Monkey noises were also broadcast over the radio system in response to the communications of black employees. The display of the Confederate flag was pervasive throughout the plant, and items containing Nucor’s logo alongside the Confederate flag were sold in the plant’s gift shop. Additionally, several e-mails that depicted black people in racially offensive ways, such as by showing them with nooses around their necks, were circulated by various employees. Once, an employee held up a noose and told a black co-worker that it was for him. The plant is organized into six production departments: beam mill, hot mill, cold mill, melting, maintenance, and shipping. When a job opening at the plant becomes available, the position is advertised over a plant-wide posting and bid- ding system controlled by the central personnel department. Employees are allowed to bid on positions in any department. Although, by policy, the plant’s general manager approves all promotions and handles discrimination and harassment inves- tigations, the record suggests that each department manager has unbridled discretion to make promotions within his department utilizing whatever objective or subjective factors he wishes. There were no black supervisors until after the institution of the Equal Employment Opportunity Commis- sion charges that preceded this litigation. Indeed, a white supervisor testified that his department manager—who wore a Confederate flag emblem on his hardhat—told him that he would never promote a black employee to supervisor. (J.A. 1066, 1885-86.) The present litigation arose on August 25, 2004, when seven black employees at the plant, along with employee- plaintiffs at plants owned by Nucor in other states, brought suit under 42 U.S.C. § 1981 (2000) and Title VII of the Civil Rights Act of 1964 in the U.S. District Court for the Western District of Arkansas on behalf of themselves and approxi- 4 BROWN v. NUCOR CORP. mately one-hundred other past and present black employees at the plant. At the time the litigation commenced, there were 611 employees working at Nucor’s South Carolina plant, of whom seventy-one were black. The Western District of Arkansas severed the case and transferred the claims brought by the seven plaintiffs in South Carolina to the District of South Carolina. The appellants seek a permanent injunction, back pay, compensatory and punitive damages, and attorney’s fees. On May 7, 2007, the appellants filed a motion for class cer- tification alleging: (1) A pattern or practice of disparate treatment against African-American employees with respect to promotion opportunities at the plant; (2) Nucor’s promotion procedure, which allows white managers and supervisors to use subjective criteria to promote employees, has a disparate impact on African- American employees who apply for promotions, and (3) Nucor requires African-American employees to work in a plant-wide hostile work environment. (J.A. 8980.) The district court denied class certification, and the would-be class plaintiffs now appeal. II. We review the district court’s certification decision for abuse of discretion. Doe v. Chao, 306 F.3d 170, 183 (4th Cir. 2002), aff’d on other grounds, 540 U.S. 614 (2004). "[A] Title VII class action, like any other class action, may only be certi- fied if the trial court is satisfied, after a rigorous analysis, that the prerequisites of Rule 23(a) have been satisfied." Gen. Tel. Co. of the Sw. v. Falcon, 457 U.S. 147, 161 (1982). III. Rule 23(a) of the Federal Rules of Civil Procedure provides the following: BROWN v. NUCOR CORP. 5 One or more members of a class may sue or be sued as representative parties on behalf of all members only if: (1) the class is so numerous that joinder of all mem- bers is impracticable, (2) there are questions of law or fact common to the class, (3) the claims or defenses of the representative par- ties are typical of the claims or defenses of the class, and (4) the representative parties will fairly and ade- quately protect the interests of the class. In our review of these factors, we bear in mind that the final three requirements of Rule 23(a) "tend to merge," with commonality and typicality "serving as guideposts for determining whether . . . maintenance of a class action is economical and whether the named plaintiff’s claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence." Broussard v. Meineke Discount Muffler Shops, 155 F.3d 331, 337 (4th Cir. 1998) (quoting Falcon, 457 U.S. at 157 n.13) (alteration in original). Indeed, "[c]ertification is only con- cerned with the commonality (not the apparent merit) of the claims and the existence of a sufficiently numerous group of persons who may assert those claims." Lilly v. Harris-Teeter Supermarket, 720 F.2d 326, 332-33 (4th Cir. 1983). The district court below considered each of the Rule 23(a) requirements in turn, and we shall do the same. 6 BROWN v. NUCOR CORP. A. First, regarding numerosity, the district court found that the appellants satisfied the requirement because there were ninety-four black employees who worked at the plant from 2001 through 2004. Nucor argued that because only nineteen of these employees bid on positions from 2001 to 2006, the numerosity requirement had not been met. However, the dis- trict court held that the other black employees should be counted in the class because "potential applicants are eligible to prove that they would have applied for a promotion but for the discriminatory practice." (J.A. 8994.) Appellees do not challenge this finding, and we therefore presume it to be cor- rect. B. Second, on commonality, the district court ruled against the appellants. The court reasoned that subjectivity in decision- making alone was insufficient to establish a disparate impact claim, discredited the statistical evidence presented in support of the disparate impact claim, and rejected the hostile work environment claim. We find that the district court improperly discounted the appellants’ direct evidence, which, alone, was sufficient to establish commonality. Moreover, the district court improperly excluded the appellants’ alternative bench- mark for missing employment data and therefore erred in finding their statistical calculations to be insufficient to estab- lish commonality. Finally, the district court erred in finding an insufficient basis for commonality on the hostile work environment claim. 1. This Court has noted before that allegations of "a practice of disparate treatment in the exercise of unbridled discretion . . . rais[es] questions of law and fact common to all [subject] black employees." Lilly, 720 F.2d at 333. The Fifth Circuit BROWN v. NUCOR CORP. 7 reasoned similarly in Shipes v. Trinity Industries, 987 F.2d 311, 316 (5th Cir. 1993): The threshold requirements of commonality and typ- icality are not high; Rule 23(a) requires only that res- olution of the common questions affect all or a substantial number of the class members. Allega- tions of similar discriminatory employment prac- tices, such as the use of entirely subjective personnel processes that operate to discriminate, satisfy the commonality and typicality requirements of Rule 23(a). (internal citation omitted). In Lilly, we further concluded that statistical evidence is unnecessary to establish commonality: [P]laintiffs need not, at the time of the motion for class certification, demonstrate by statistical evi- dence that blacks have been terminated [or pro- moted] at a higher rate than have whites, or any other differential. Certification is only concerned with the commonality (not the apparent merit) of the claims and the existence of a sufficiently numerous group of persons who may assert those claims . . . . Lilly, 702 F.2d at 332-33. The district court determined that the appellants’ statistics were insufficient to establish commonality. Yet, the appel- lants have certainly presented compelling direct evidence of discrimination, such as denials of promotions when more junior white employees were granted promotions (J.A. 1004, 1017), denial of the ability to cross-train during regular shifts like their white counterparts (J.A. 1000, 1023), and a state- ment by a white supervisor that he would never promote a black employee (J.A. 1885-86). This evidence alone estab- lishes common claims of discrimination worthy of class certi- fication. See Stastny v. S. Bell Tel. & Tel. Co., 628 F.2d 267, 8 BROWN v. NUCOR CORP. 278 (4th Cir. 1980) (noting that class certification for a pattern-or-practice claim might be based on inferential, statis- tical, or direct evidence of discrimination). That the appellants also presented statistical data only strengthens their case. The appellants have presented five statistical calculations in support of their discrimination claims,1 but we need focus on only one. In order to demonstrate a disparity in job promo- tions, the appellants compared the estimated percentage of blacks who sought promotions between December 1999 and December 2003 with the estimated percentage of blacks who received promotions during that period. The former figure was 19.24%, and the latter figure was 7.94%. However, the appellants could not rely entirely on actual applicant data in making this calculation because Nucor has destroyed such data for promotion periods prior to 2001.2 Instead, in an effort to establish an alternative benchmark, the appellants, using change-of-status forms, identified twenty-seven positions 1 First, the appellants presented a calculation of the disparity in utiliza- tion rates at Nucor by comparing the average percentage of blacks in skill positions at the plant from 2000 to 2004 with the percentage of blacks in the surrounding population in 2000 who were qualified for the skill posi- tions at the plant. In the surrounding area, blacks made up 38.2% of the population qualified to hold skilled positions. However, only 13.42% of the employees at the plant from 2000 through 2004 were black. The sec- ond calculation was the same as the first, except the appellants extended the time period to 2006 to demonstrate that the disparity continued. To demonstrate a disparity in the number of job applications by black applicants, the third calculation compared the percentage of qualified blacks in the surrounding population in 2000 with the percentage of blacks who bid on plant promotions between 2001 and December 2003. The for- mer percentage was 38.2, and the latter was 16. The fourth calculation is discussed infra. Finally, the fifth calculation compared the percentage of qualified blacks in the surrounding community in 2000 with the estimated percentage of blacks who received promotions from December 1999 to December 2003. The former figure was 38.2%, and the latter figure was 7.9%. 2 We make no judgment regarding the propriety of Nucor’s destruction of the data. BROWN v. NUCOR CORP. 9 filled during the 1999-2000 time period and extrapolated from that data using the assumption that the racial composition of the bidding pool for those jobs was the same as the weighted average of the racial composition of the bidding pools for the subsequent, post-2000 period. Nucor provided selected promotions data for 2001 through 2006, and it argues that this data should have been used instead of the extrapolated 1999-2000 data because the appel- lants gathered the extrapolated data from the change of status forms, which do not indicate whether a position was open for bidding. The district court decided to exclude the appellants’ pre-2001 data and Nucor’s post-2003 data and rely solely on the data that Nucor provided for January 2001 through December 2003.3 Although the appellants argued that the 80% rule was the proper standard by which to evaluate their statis- tical evidence,4 the district court determined that standard deviation analysis was the proper method by which to do so.5 3 Nucor complains that the appellants themselves have created an unduly restricted data set by excluding the promotions that occurred after the fil- ing of their suit in 2004. However, post-lawsuit promotions would not enhance the accuracy of the data on Nucor’s promotion practices. Indeed, such promotions would likely detract from their accuracy, since it is no secret that the institution of litigation often prompts companies to change their practices. These post-suit changes have only minimal weight with regard to the merits of the plaintiffs’ claims. See Holsey v. Armour & Co., 743 F.2d 199, 214 n.5 (4th Cir. 1984) ("We find no error in the fact that the district court minimized the significance of evidence of Armour’s post- complaint hiring and promotion of black employees."). The district court found in the appellants’ favor on this issue, which is why it did not utilize Nucor’s post-2003 data, and we find no abuse of discretion in its doing so. 4 The 80%, or four-fifths, rule "is an administrative rule of thumb used by agencies concerned with Title VII cases. It offers a definition of what is a serious difference in the passing rates for protected classes. If the selection rate for a protected class is less than 80% of the selection rate for the group selected at the highest rate, that constitutes adverse impact." Chisholm v. U.S. Postal Serv., 665 F.2d 482, 495 n.22 (4th Cir. 1981). 5 Standard deviation analysis considers the differences between expected and observed values—in this case, for example, the number of blacks pro- 10 BROWN v. NUCOR CORP. In its subsequent assessment of the statistics, the district court found that the appellants had not proven a statistically signifi- cant disparity, and it therefore ruled against them on the com- monality prong of their claims. The appellants argue that the district court’s decisions were improper and negatively affected their statistics. In another Title VII suit involving destroyed data, United States v. County of Fairfax, 629 F.2d 932 (4th Cir. 1980), Fairfax County destroyed its job applications for the 1974-77 period. "As a consequence, the government extrapolated the figures for 1974-77 from the 1978 applicant pool by assuming that the applicant pool for 1974-77 contained the same pro- portions of blacks and women as did the 1978 applicant pool." Id. at 940. The Court approved the use of this alternative benchmark and concluded that it was "the most salient proof of the County’s labor market." Id.6 In this case, since Nucor moted versus the number of blacks one would expect to be promoted based on their percentage in the relevant population (e.g., job applicants or the plant workforce). Standard deviation analysis is often viewed as superior to the 80% rule because it takes into account the natural fluctua- tions from the expected value that come with any random sample. In other words, for any given sample of black applicant data, an observed disparity might be the result of chance. The smaller the standard deviation, the more closely the data points are clustered around the mean (expected value), and the more likely the data points are the result of chance. "For example, if a coin were tossed ten times . . . and came up heads four times, no one would think the coin was biased (0.632 standard deviations), but if this same ratio occurred for a total of 10,000 tosses, of which 4,000 were heads, the result could not be attributed to chance (20 standard devia- tions)." Lilly, 720 F.2d at 336 n.17. 6 The dissent does not argue with the underlying reasoning of County of Fairfax: when an employer destroys relevant employment data, the plain- tiffs may utilize alternative benchmarks to make up for this lost data. Cer- tainly, the benchmarks will not be as good as the destroyed data themselves—that would be next to impossible to achieve. Nevertheless, the plaintiffs should not be penalized by the destruction (however inno- cent) of such data. Any factual differences between the cases do not over- come this underlying reasoning: BROWN v. NUCOR CORP. 11 destroyed the pre-2001 job promotions data, the appellants likewise were free to attempt to utilize an alternative bench- mark in order to form their calculations, and the district court abused its discretion in ruling that only the data that Nucor provided could be used.7 The appellants should not be penal- ized because Nucor destroyed the actual pre-2001 job promo- tions data.8 The question before the district court was not whether the appellants have definitively proven disparate treatment and a disparate impact; rather, the question was whether the basis of appellants’ discrimination claims was sufficient to support class certification. See Lilly, 720 F.2d at 332. In excluding the First, the fact that there is only one year of missing data in this case is inapposite—the point is that a year of relevant data has been destroyed, and that destruction should not be used to weaken the appellants’ case. Likewise, the absence of a statutory duty upon Nucor to maintain any data does not imply a judicial bar from allowing the appellants to approximate the destroyed data in making their claim. Finally, we certainly do not agree with the dissent’s insinuation that, because this case involves pro- motions and not hiring, the appellants are less entitled to data (and approx- imations thereof) to support their claims. The dissent is correct, and should itself remember, that we are not reviewing the merits of this case, and the district court was simply incorrect to exclude all evidence of the 1999- 2000 promotions. 7 The district court made the following determination in excluding the appellants’ pre-2001 data: "Nucor provided the plaintiffs with records of promotions occurring in the plant from January 2001 to December 2003. Statistics based on actual data is [sic] more probative than statistics based on assumptions. Consequently, the Court relies on statistics resulting from an analysis of this data." (J.A. 8985.) 8 In response to appellants’ argument that the district court should have compelled Nucor to provide more data during discovery, Nucor argues that the appellants cannot appeal a discovery order under FRCP 23(f). We agree with the Eleventh Circuit that "jurisdiction granted by Rule 23(f) does not extend to [a] separate [discovery] order." DeLeon-Granados v. Eller & Sons Trees, Inc., 497 F.3d 1214, 1218 n.1 (11th Cir. 2007). We therefore express no opinion at this time regarding the district court’s deci- sion not to compel Nucor to produce additional data. 12 BROWN v. NUCOR CORP. appellants’ data, the district court ruled, "These assumptions [regarding the 1999-2000 data] may be reasonable and the statistics based thereon may be relevant to prove discrimina- tion at the plant. However, the necessity of the assumptions diminishes their probative value." (J.A. 8985.) Yet, evidence need not be conclusive to be probative, and even evidence that is of relatively weak probative value may be useful in meeting the commonality requirement. With the 1999-2000 data included, the record indicates that the appellants’ statistics would be significant at greater than two standard deviations.9 The appellants have therefore pre- sented valid statistical evidence that independently indicates a disparate impact and disparate treatment in job promotions at Nucor, and we reiterate that an in-depth assessment of the merits of appellants’ claims at this stage would be improper. See Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 177 (1974); Thorn v. Jefferson-Pilot Life Ins. Co., 445 F.3d 311, 319 (4th Cir. 2006). We therefore find that the appellants’ calculations 9 The Supreme Court has indicated that anything greater than two or three standard deviations in racial discrimination cases is suspicious. Hazelwood School Dist. v. United States, 433 U.S. 299, 308 n.14 (1977); Castaneda v. Partida, 430 U.S. 482, 496 n.17 (1977). We will presume that two standard deviations is the proper threshold for this case, but we reserve the unanswered question of whether this rule should be limited to large sample sizes. With the 1999-2000 data excluded, the appellants’ pro- motion data demonstrated a disparity resulting in only 1.48 standard devi- ations—below the Supreme Court’s threshold. With the data included, however, the appellants indicate that their calculations yielded 2.54 stan- dard deviations. For our purposes, a threshold of two standard deviations corresponds roughly to a 95% confidence level or a .05 level of significance, i.e., there is only a 5% probability that the result is due to chance. Three standard deviations would equate to a 99.7% confidence level. See generally Bruce J. Chalmer, Understanding Statistics 97-98 (1987); Wikipedia, Standard deviation, http://en.wikipedia.org/wiki/Standard_deviation (last visited June 19, 2009). BROWN v. NUCOR CORP. 13 based on their alternative benchmark were adequate to estab- lish commonality.10 In summary, because the appellants’ direct evidence alone was sufficient to demonstrate common claims of disparate treatment and disparate impact, their statistical data did not need to meet a two-standard-deviation threshold, and the dis- trict court erred as a matter of law in requiring them to do so. Yet, we further find that the district court abused its discretion when it excluded the appellants’ alternative calculations of the destroyed pre-2001 promotions data. With this data included, the appellants’ statistics were independently sufficient to meet the Rule 23 commonality requirement. We therefore conclude that the appellants satisfied the commonality requirement for their discrimination claims, and the district court should have found in the appellants’ favor on this portion of their motion. See Caridad v. Metro-North Commuter R.R., 191 F.3d 283, 293 (2d Cir. 1999) ("More detailed statistics might be required to sustain the Plaintiffs’ burden of persuasion, but [these statistics], in conjunction with the anecdotal evidence, satisf[y] the Class Plaintiffs’ burden of demonstrating com- monality for purposes of class certification." (internal citation omitted)). 2. The district court also discounted the sixteen affidavits that 10 The dissent takes issue with our crediting the appellants’ statistics. However, we emphasize that at this stage, we are dealing only with whether the appellants have presented sufficient statistical information to establish commonality. We must walk a fine line between a facial class certification assessment and an assessment on the merits, and the dissent has stepped to the other side. The dissent’s critiques might very well dis- credit the appellants’ statistics later, upon a full review of the merits, but the information that the appellants have presented is enough to allow them to get to that point. Our crediting the statistics certainly does not portend the opening of any class certification floodgates. There are many require- ments for class certification, and our decision today respects that fact. 14 BROWN v. NUCOR CORP. the appellants presented in support of their pattern-or-practice and hostile-work-environment claims because only two of the affiants were in departments other than the beam mill depart- ment. The district court found that this concentration pre- cluded the establishment of a pattern or practice of discrimination. It reached a similar conclusion on the appel- lants’ hostile work environment claim: "The plant’s produc- tion departments can be classified as separate ‘environments.’ A class members [sic] claim of a hostile work environment in the hot mill will vary significantly from a class member’s claim of a hostile work environment in the beam mill." (J.A. 8989.) The appellants challenge the district court’s ruling by argu- ing that the court’s "separate environments" analysis was flawed. The district court noted that "[t]he plaintiffs have presented plant-wide racist acts potentially experienced by every African-American employee working at the plant when the acts occurred. These acts include: (1) racist e-mails, (2) display of the confederate flag, and (3) racist remarks over the plant radio." (J.A. 8988-89.) Yet, it then proceeded to classify the departments as unique environments. The Supreme Court has held, "When the workplace is per- meated with ‘discriminatory intimidation, ridicule, and insult,’ that is ‘sufficiently severe or pervasive to alter the conditions of the victim’s employment and create an abusive working environment,’ Title VII is violated." Harris v. Fork- lift Sys., Inc., 510 U.S. 17, 21 (1993) (quoting Meritor Sav. Bank, FSB v. Vinson, 477 U.S. 57, 65, 67 (1986)). This is essentially what the appellants argue: despite the formal clas- sification of the plant into six production departments, the rac- ist acts had plant-wide repercussions and affected all black employees. A hostile environment determination can be made "only by looking at all the circumstances." Harris, 510 U.S. at 23. Such circumstances include "the frequency of the dis- criminatory conduct; its severity; whether it is physically threatening or humiliating, or a mere offensive utterance; and BROWN v. NUCOR CORP. 15 whether it unreasonably interferes with an employee’s work performance." Id.; accord Nat’l R.R. Passenger Corp. v. Mor- gan, 536 U.S. 101, 116 (2002). The allegations made in the affidavits that the district court discredited should have fac- tored into the totality-of-the-circumstances assessment of the plant as a whole. In Hill v. Western Electric Co., 596 F.2d 99, 102 (4th Cir. 1979), the Court held: A person who has been injured by unlawful, discrim- inatory promotion practices in one department of a single facility may represent others who have been injured by the same discriminatory promotion prac- tices in other departments of the same facility. In such a case, the representatives of the class all have the same interests in being free from job discrimina- tion, and they have suffered injury in precisely the same way in the denial of promotion. In Hill, the Court determined that employees in one depart- ment could not be included in the class because they did all of their work off-site. Id. That is not the case here. The appellees argue that Nucor’s management provided "clear evidence" of "the decentralized nature of Nucor’s prac- tices." (Appellees’ Br. 18.) However, the "clear evidence" they cite is itself contradictory on this matter. While the man- ager of the beam mill indicated that his department is "sepa- rate from the buildings and areas of other Nucor Steel Berkeley departments" (J.A. 7885), the melt shop manager and supervisor noted that their building "is attached to Beam Mill and Hot Mill" (J.A. 7896, 7917). Moreover, "Nucor Berkeley employees share a locker room with caster, caster maintenance, hot mill, melt shop maintenance, hot mill main- tenance, as well as environmental." (J.A. 7896, 7917.) And as noted previously, racial slurs, monkey noises, and other offen- sive statements were broadcast over the plant-wide radio. 16 BROWN v. NUCOR CORP. Thus, there is scant, if any, evidence that each of the depart- ments is so autonomous as to justify classifying them as sepa- rate environments. In light of the foregoing cases and facts, it was an abuse of discretion for the district court to find that the employees at the plant were separated into different environments. All of the employees worked at a single facility, in departments that were, at minimum, connected to each other, and employees shared several common areas. There is therefore sufficient evidence to indicate that all of the black employees were affected by the comments and actions of the white employees and supervisors in other departments. Thus, the affidavits of employees in one department are admissible to prove a plant- wide hostile environment that affected employees in other departments, and the plaintiffs have satisfied the commonality requirement for their hostile work environment claim. See Holsey, 743 F.2d at 216-17. C. The district court determined that in order for the appellants to prove typicality for their disparate treatment claim, they would have to satisfy the framework set forth in McDonnell Douglas, 411 U.S. 792. However, because the court found that the plaintiffs had not demonstrated a pattern or practice of discrimination, it held that the attempts by the individual plaintiffs to demonstrate disparate treatment would not be typ- ical of the attempts of the other class members to do so. Moreover, the district court found that because the plaintiffs had not demonstrated a disparity in black promotion rates, their disparate impact claims also would not be typical of the class. Finally, the district court found that the named plain- tiffs’ hostile work environment claims were not typical of the claims of class members outside of the beam mill. As noted previously, the district court abused its discretion in its rulings on the appellants’ disparate treatment, disparate BROWN v. NUCOR CORP. 17 impact, and hostile work environment claims; therefore, it was also an abuse of discretion for the district court to base its typicality conclusion on these rulings. Moreover, as we noted supra, commonality and typicality "tend to merge," Broussard, 155 F.3d at 337, and the appellants have presented sufficient evidence to satisfy the threshold typicality require- ment of Rule 23(a). D. Finally, regarding adequacy, the district court found the plaintiffs not to be adequate representatives for the disparate treatment and disparate impact claims. However, the court could "discern no conflicts of interest among the plaintiffs and class members regarding hostile work environment claims." (J.A. 8993.) Given the above conclusions that we have reached regarding the appellants’ disparate treatment and dis- parate impact claims, we also find that the district court’s assessment of the adequacy factor with regard to these claims was an abuse of discretion. To the extent that the district court was correct that the putative class representatives have a conflict with the class in terms of competition for promotions, this conflict should not defeat class certification. Indeed, if this were true, how might a class action challenging promotion practices ever be brought—unless the EEOC deems fit to do so—when the plaintiffs seek instatement into previously denied positions? The appellees point to General Telephone Co. of the North- west v. EEOC, 446 U.S. 318 (1980), but that case is to no avail. There, the Supreme Court simply "note[d]" as an "ex- ample" that "[i]n employment discrimination litigation, con- flicts might arise, for example, between employees and applicants who were denied employment and who will, if granted relief, compete with employees for fringe benefits or seniority. Under Rule 23, the same plaintiff could not repre- sent these classes." Id. at 331. Setting aside the question of whether this language was meant to be controlling, it plainly 18 BROWN v. NUCOR CORP. does not apply to this case, which deals with promotions and not the competition for "fringe benefits or seniority" posed by new hires. The district court can address plaintiffs’ claims for injunc- tive or other relief after liability and other common issues are determined. See Fed. R. Civ. P. 23(c)(5); Int’l Bhd. of Team- sters v. United States, 431 U.S. 324, 361-62 (1977) ("[A]s is typical of Title VII pattern-or-practice suits, the question of individual relief does not arise until it has been proved that the employer has followed an employment policy of unlawful discrimination. The force of that proof does not dissipate at the remedial stage of the trial."); Hill, 672 F.2d at 387 ("Bifurcation of Title VII class action proceedings for hear- ings on liability and damages is now commonplace . . . ."). If, at the second stage of the proceeding, conflicts need to be resolved with regard to promotions, the district court can do so then. See Gunnells v. Healthplan Servs., Inc., 348 F.3d 417, 429 (4th Cir. 2003) ("[T]he need for individualized proof of damages alone will not defeat class certification."); Pett- way v. Am. Cast Iron Pipe Co., 576 F.2d 1157, 1178 (5th Cir. 1978). Of course, certification is conditional, and should the conflicts prove intractable at the second stage, the district court may simply decertify the class. See In re Sch. Asbestos Litig., 789 F.2d 996, 1011 (3d Cir. 1986). But at this stage of the proceedings, we find that the appellants are adequate rep- resentatives for the disparate impact and treatment claims of the putative class. As noted previously, the district court found the appellants to be adequate representatives for the hostile work environ- ment claim. The appellees do not dispute this finding, and we find no abuse of discretion in the district court’s so conclud- ing. BROWN v. NUCOR CORP. 19 IV. Our review of the district court’s assessment of the Rule 23(a) factors leads us to conclude that the court abused its discretion in denying class certification. Given our consideration above, we find that the class certification requirements of Rule 23(a) have been satisfied with regard to the appellants’ disparate impact, disparate treatment, and hostile work environment claims. We therefore vacate the district court’s denial of the appellants’ motion for class certification, and we remand the case to the district court with instructions to certify the appellants’ class action and to engage in further proceedings consistent with this opinion. VACATED AND REMANDED WITH INSTRUCTIONS AGEE, Circuit Judge, concurring in part and dissenting in part: I. I concur in the majority opinion as to Section III(A) (the dis- trict court’s determination on the numerosity factor), that portion of Section III(B)(2) reversing the district court’s judgment as to the hostile work environment claim concerning the commonality factor,1 that portion of Section III(C) as to the 1To the extent the district court’s judgment was based upon a determin- ation with regard to Sections III(B)(2) and (C) that the appellants’ hostile 20 BROWN v. NUCOR CORP. typicality factor for the hostile work environment claim only, and Section III(D) reversing the district court’s judgment as to the adequacy of representation.2 However, because I believe the majority opinion fails to adhere to the standard of review and usurps the role of the district court, I respectfully dissent as to sections III(B)(1), (2), and III(C) regarding the commonality and typicality factors for the disparate treatment and disparate impact claims. For the following reasons, I would affirm the district court’s judgment not to grant class certification as to those claims. II. "We review the district court’s certification decision for abuse of discretion." Gregory v. Finova Capital Corp., 442 F.3d 188, 190 (4th Cir. 2006) (citing McClain v. S. C. Nat. Bank, 105 F.3d 898, 902 (4th Cir. 1997)). It is the party seek- ing class certification who bears the burden of proving the requirements of Rule 23. Lienhart v. Dryvit Sys., Inc., 255 F.3d 138, 146 (4th Cir. 2001); Int’l Woodworkers of Am. v. Chesapeake Bay Plywood Corp., 659 F.2d 1259, 1267 (4th Cir. 1981). "A court has broad discretion in deciding whether to allow the maintenance of a class action." Roman v. ESB, Inc., 550 F.2d 1343, 1348 (4th Cir. 1976). work environment claim did not have common questions of law or fact or was not typical of the class because the evidence showed separate, uncon- nected work environments, that determination is not supported by the record. While Nucor’s evidence shows its operation of the plant depart- ments was distinct and separated among the departments, that evidence is not probative as to the effects of the alleged acts on the employees’ work- ing environment across the whole facility. However, should Nucor adduce relevant evidence on the merits showing distinct and unconnected work environments as affects the appellants and other employees, the district court may take such further action as Rule 23(c)(1)(C) permits. 2 I also concur in the holdings represented by footnote 4, that the district court did not abuse its discretion in excluding post-2003 promotion data, and footnote 8, that Rule 23(f) did not permit interlocutory appeal of the discovery order. BROWN v. NUCOR CORP. 21 A district court has abused its discretion if its decision "is guided by erroneous legal principles" or "rests upon a clearly erroneous factual finding." Westberry v. Gislaved Gummi AB, 178 F.3d 257, 261 (4th Cir. 1999). We do not ask whether we would have come to the same conclusion as the district court if we were examining the matter de novo. Fell- heimer, Eichen & Braverman, P.C. v. Charter Techs., 57 F.3d 1215, 1223 (3d Cir. 1995). Rather, after reviewing the record and the reasons the district court offered for its decision, we reverse for abuse of discretion if we form "a definite and firm conviction that the court below committed a clear error of judg- ment in the conclusion it reached upon a weighing of the relevant factors." Westberry, 178 F.3d at 261. Morris v. Wachovia Sec., Inc., 448 F.3d 268, 277 (4th Cir. 2006); see also Thorn v. Jefferson-Pilot Life Ins. Co., 445 F.3d 311, 317 (4th Cir. 2006) ("A district court per se abuses its discretion when it makes an error of law or clearly errs in its factual findings."). In the class certification context of Rule 23, our decisions have specifically acknowledged that there is no abuse of dis- cretion, as a matter of law, when a reasonable jurist could have made the decision at issue based on the evidence in the record. [A] great deal of deference must be afforded to a District Court’s decision to deny certification of a plaintiff class for a class action. Indeed, even if rea- sonable persons can disagree as to whether the class certification was proper, the District Court’s decision certainly does not amount to an abuse of discretion. Simmons v. Poe, 47 F.3d 1370, 1381-82 (4th Cir. 1995). "[W]e cannot hold that the district court’s refusal to expand the class . . . was an abuse of discretion, though reasons 22 BROWN v. NUCOR CORP. clearly existed for taking the other course, and taking it surely would not either have abused discretion." Lewis v. Bloomberg Mills, Inc., 773 F.2d 561, 564 (4th Cir. 1985). On appeal from an order granting or denying certification, the appellate court does not reweigh the evidence de novo, but is to accord the district court’s decision "broad discretion." See Lienhart, 255 F.3d at 146 (Within the framework of Rule 23 "[a] district court has broad discretion in deciding whether to certify a class." (internal quotations omitted)). At its core, the district court determined the appellants could not meet either the commonality or typicality factors for certification under Rule 23(a) because they failed to meet their burden of proof as to the direct or statistical evidence of discriminatory promotions. The record in this case reflects the district court’s judgment was neither "guided by erroneous legal principles" nor grounded "upon a clearly erroneous fac- tual finding." Thus, when the full context of the appellants’ proffered evidence is examined and the standard of review is observed, the district court cannot be said to have abused its discretion in denying certification because a reasonable jurist could have reached that decision on this record. While the majority opinion is correct that a certification determination is not a judgment on the merits of the underly- ing claims, neither is it a pleadings-based determination that can be verified by any evidence, no matter how deficient. Cer- tainly since the United States Supreme Court’s decision in General Telephone Co. v. Falcon, 457 U.S. 147 (1982), dis- trict courts have been required to undertake a "rigorous analy- sis" not just of the plaintiffs’ claims as pled, but of the evidence to support those claims in order to make an appro- priate judgment on Rule 23 certification. 457 U.S. at 161 ("[A] Title VII class action, like any other class action, may only be certified if the trial court is satisfied, after a rigorous analysis, that the prerequisites of Rule 23(a) have been satis- fied."). "[S]ometimes it may be necessary for the court to BROWN v. NUCOR CORP. 23 probe behind the pleadings before coming to rest on the certi- fication question." Id. at 160. In the case at bar, the district court followed its duty and performed that rigorous analysis, although with a result that the majority may not have reached if they were sitting as the trier of fact. This Court has been clear in explaining that an objection to examining the merits of a class certification claim, as reflected by the majority opinion’s citation to Eisen v. Carlisle & Jacquelin, 417 U.S. 156 (1974), supra at 12-13, was rendered invalid after Falcon. The district court’s role, and ours, as explained in Gariety v. Grant Thornton, LLP, 368 F.3d 356 (4th Cir. 2004), is telling: If it were appropriate for a court simply to accept the allegations of a complaint at face value in making class action findings, every complaint asserting the requirements of Rule 23(a) and (b) would automati- cally lead to a certification order, frustrating the dis- trict court’s responsibilities for taking a "close look" at relevant matters, Amchem, 521 U.S. at 615, 117 S.Ct. 2231, for conducting a "rigorous analysis" of such matters, Falcon, 457 U.S. at 161, and for mak- ing "findings" that the requirements of Rule 23 have been satisfied, see FED. R. CIV. P. 23(b)(3). More- over, if courts could only consider the pleadings, then "parties would have wide latitude to inject friv- olous issues to bolster or undermine a finding of pre- dominance." Robert G. Bone & David S. Evans, Class Certification and the Substantive Merits, 51 Duke L.J. 1251, 1269 (2002). When Rule 23(c), which originally required certifi- cation orders to be made "as soon as practicable after commencement of [the] action," was amended in 2003 to require the court to determine class certifica- tions "at an early practicable time," the Advisory 24 BROWN v. NUCOR CORP. Committee on Civil Rules explained the preexisting and longstanding practice that prompted the change: Time may be needed to gather information necessary to make the certification deci- sion. Although an evaluation of the proba- ble outcome on the merits is not properly part of the certification decision, discovery in aid of the certification decision often includes information required to identify the nature of the issues that actually will be presented at trial. In this sense, it is appro- priate to conduct controlled discovery into the "merits," limited to those aspects rele- vant to making the certification decision on an informed basis. FED. R. CIV. P. 23 advisory committee’s note to 2003 amendments (emphasis added). The Eisen decision, upon which the district court relied, does not require a court to accept plaintiffs’ pleadings when assessing whether a class should be certified. In Eisen, the Supreme Court held that the district court’s preliminary hearing on the merits of the case—concluding that the plaintiff was "more than likely" to prevail—was inappropriate for the purpose of determining whether a class action could be maintained. 417 U.S. at 177-78, 94 S.Ct. 2140. Eisen simply restricts a court from expanding the Rule 23 certification analysis to include consider- ation of whether the proposed class is likely to pre- vail ultimately on the merits. See Castano, 84 F.3d at 744; 5 Moore’s Federal Practice ¶ 23.84[2][a] (3d ed. 2003). As the Supreme Court itself stated in a post-Eisen case, "sometimes it may be necessary for the [district] court to probe behind the pleadings before coming to rest on the certification question." BROWN v. NUCOR CORP. 25 Falcon, 457 U.S. at 160; see also Coopers & Lyb- rand v. Livesay, 437 U.S. 463, 469, 102 S.Ct. 2364 (1978) ("[T]he class determination generally involves considerations that are enmeshed in the fac- tual and legal issues comprising the plaintiff’s cause of action" (internal quotation marks omitted)). Thus, while an evaluation of the merits to determine the strength of plaintiffs’ case is not part of a Rule 23 analysis, the factors spelled out in Rule 23 must be addressed through findings, even if they overlap with issues on the merits. Eisen’s prohibition against assessing plaintiffs’ likelihood of success on the merits as part of a Rule 23 certification does not mean that consideration of facts necessary to a Rule 23 determination is foreclosed merely because they are required to be proved as part of the merits. The analysis under Rule 23 must focus on the require- ments of the rule, and if findings made in connection with those requirements overlap findings that will have to be made on the merits, such overlap is only coincidental. 368 F.3d at 365-66. The majority’s citation to Thorn v. Jefferson-Pilot Life Ins. Co., 445 F.3d 311 (4th Cir. 2006), is also puzzling, as this post-Gariety decision acknowledged the type of inquiry undertaken by the district court in the case at bar: At the class certification phase, the district court must take a "close look" at the facts relevant to the certification question and, if necessary, make spe- cific findings on the propriety of certification. Gari- ety, 368 F.3d at 365 (internal quotations omitted). Such findings can be necessary even if the issues tend to overlap into the merits of the underlying case. Falcon, 457 U.S. at 160. 26 BROWN v. NUCOR CORP. Thorn, 445 F.3d at 319. Thus, the district court’s examination and evaluation of the plaintiffs’ evidence to support the class certification claim was not only appropriate, but required under Falcon and Gariety. That the majority would have reached a different conclusion, albeit lacking all the advantages of the district court’s first-hand knowledge of the case, establishes nothing because the standard of review goes to what a reasonable jurist could have found, which may very well lead to differing views on the same set of facts. See Morris, 448 F.3d at 277 ("We do not ask whether we would have come to the same conclusion as the district court if we were examining the mat- ter de novo.") (citing Fellheimer, Eichen & Braverman, P.C. v. Charter Techs., 57 F.3d 1215, 1223 (3rd Cir. 1995)). In my view, the majority opinion steps beyond the standard of review and improperly reweighs the evidence so as to achieve a certification that was properly within the district court’s discretion to deny. The majority appears to sanction a rule of mandated class certification even where a reasonable jurist could properly find the named plaintiffs’ evidentiary basis was insufficient. Such a result is not consonant with the appellate standard of review. III. The majority opinion initially rests on the conclusion that the appellants’ direct evidence alone was sufficient to require class certification of the claims for disparate treatment and disparate impact. Supra at 6-7. I agree with the majority opin- ion that direct evidence could be sufficient, in a proper case, to support a certification decision without corollary statistical proof, although that is not a common practice in at least the disparate impact context. Nonetheless, the majority’s conclu- sion — that the direct evidence alone was sufficient to merit certification — seems an odd decisional basis because that argument was not made by the appellants in the district court. BROWN v. NUCOR CORP. 27 Neither the appellants’ Statement of Issues, briefs or oral argument raised, as a point of error, the failure of the district court to grant certification on the basis of direct evidence alone. The appellants’ argument below was that the direct evi- dence "bolstered" their main evidentiary claim, which was their statistical analysis. See, e.g., J.A. 965 ("The plaintiffs have also bolstered the statistical evidence of a pattern or practice of subjective bias with both direct and anecdotal evi- dence of actual racial bias by the promotion decision-makers . . . .") (emphasis added); J.A. 8473 ("The commonality and typicality of plaintiffs’ pattern-or-practice claim is further bol- stered by the subjective nature of the interview and selection stage of Nucor’s promotion process.") (emphasis added); Br. of Appellant at 37 ("Plaintiffs did not rely on the mere exis- tence of a subjective promotion procedure, but presented a combination of statistical, anecdotal and direct evidence showing that such procedure was racially discriminatory throughout the plant."). Nonetheless, as noted below, when the appellants’ limited direct evidence is examined, a reason- able jurist could find that evidence failed to meet the appel- lants’ burden of proof as to either commonality or typicality for the certification of a class of all African-American plant employees. Similarly, the majority opinion appears to reweigh the evi- dence in order to sustain the appellants’ statistical model. As noted below, the district court was within its proper exercise of discretion to accord the statistical evidence little or no weight in the class certification decision. A reasonable jurist, on this record, could find the appellants’ statistical evidence too speculative and lacking a proper foundation so as to be without evidentiary value in meeting the appellants’ burden of proof. A. The Direct Evidence The appellants did not argue to the district court that their "direct evidence" was alone sufficient to support a finding of 28 BROWN v. NUCOR CORP. commonality or typicality for class certification under Rule 23. They presented direct evidence to "bolster" their statistical evidence, not as standalone proof for Rule 23(a) purposes. However, the appellants’ choice of methodology is not the dispositive issue on appeal. What the standard of review should examine in this case is whether the district court, based on the direct evidence alone, could have reasonably found that evidence insufficient to meet the appellants’ burden of proof on typicality and commonality for the proposed class. That a reasonable jurist could have also found the same evidence sufficient for certification is irrelevant to the abuse of discre- tion standard where the evidence also could support the oppo- site finding. It appears plain, on this record, that a reasonable jurist could have found, as the district court did, that the appellants’ direct evidence (though considered as only supple- mental evidence) was simply inadequate to meet their burden of proof. Under Rule 23(a)(2) and (3), the appellants were required to prove there were "questions of law or fact common to the class" and "claims . . . of the representative parties are typical . . . of the class." FED. R. CIV. P. 23. The purported class, all former and current African-American employees of the Nucor plant during the relevant time period, worked in all the plant’s departments, which include at least the beam mill, hot mill, melt shop, cold mill and shipping departments. Thus, the appellants’ claims must have common questions among the employees in all these departments and be typical of those claims. The appellants’ direct evidence, however, could rea- sonably be determined as failing to meet the burden of proof for either commonality or typicality for the purported class. The statements of the three affiants cited by the majority do claim race-based denials of promotions, the employment prac- tice at issue in their disparate treatment and disparate impact claims. However, all three were employees only in the beam mill. While some of the appellants’ declarations allege employment discrimination in other departments as to them as BROWN v. NUCOR CORP. 29 beam mill employees, they fail to do so as to non-beam mill employees. The district court noted that of all the appellants’ direct evi- dence, only two non-beam mill employees allege discrimina- tion in another department. Even the statement alleging bias of a supervisor was a statement by the department manager in charge of the beam mill about employees in the beam mill. [T]he existence of a valid individual claim does not necessarily warrant the conclusion that the individual plaintiff may successfully maintain a class action. It is equally clear that a class plaintiff’s attempt to prove the existence of a companywide policy, or even a consistent practice within a given department, may fail even though discrimination against one or two individuals has been proved. Cooper v. Fed. Reserve Bank of Richmond, 467 U.S. 867, 877–78 (1984) (characterizing the holding in Falcon); see also Int’l Bd. of Teamsters v. United States, 431 U.S. 324, 336 (1977) (to prove a pattern or practice violation a plaintiff must "prove more than the mere occurrence of isolated or ‘acciden- tal’ or sporadic discriminatory acts" but must "establish by a preponderance of the evidence that racial discrimination was the company’s standard operating procedure—the regular rather than the unusual practice"). On this record, a reasonable jurist could conclude the appellants failed to meet their burden of proof to show com- mon issues and typical claims between themselves, as employees in the beam mill seeking promotions, and the workers in all the other departments who sought promotions. The appellants simply failed to produce that evidence. The effect of this failure of proof, for certification pur- poses, is that the appellants failed to prove commonality or typicality for the non-beam mill employees they seek to repre- 30 BROWN v. NUCOR CORP. sent. In that circumstance, a reasonable jurist, like the learned district court judge in this case, could find the appellants failed to show commonality or typicality for certification of a class consisting of all African-American plant employees. Accordingly, the district court acted within its discretion to deny class certification on the basis of the appellants’ direct evidence.3 B. Statistical Evidence Nucor provided the appellants with actual promotion- related job posting and applicant data from January 2001 to February 2006.4 According to the appellants’ own expert, sta- tistical analysis of this data resulted in a standard deviation of -.84, a result indicating no statistically significant disparity in the actual versus expected percentage of African-American employees successfully bidding on jobs. J.A. 5872-73. The appellants, however, objected to use of the 2001 to 2006 time period because it included job posting and bidding informa- tion occurring after suit was filed at the end of 2003.5 The dis- 3 If the purported class was only beam mill employees, it may well have been an abuse of discretion not to certify the promotion claims (ignoring the numerosity factor). However, that was not the proposed class. The pur- ported class in this case is all the plant’s African-American employees and, for the reasons stated, the direct evidence is simply insufficient to find an abuse of discretion by the district court. 4 The appellants’ case is solely concerned with Nucor’s actions on pro- motions within the South Carolina plant’s internal workforce. Hiring or termination of employment is not at issue. Moreover, it is uncontested that promotions at the Nucor plant are only from within the existing workforce and the promotions at issue in this case are those subject to bid by employ- ees. 5 The district court’s opinion states that suit was originally filed on August 25, 2004. However, court records indicate that suit was originally filed against Nucor Steel on a company-wide basis in the Western District of Arkansas on December 8, 2003. That portion of the suit pertaining to discrimination claims at the Nucor-Berkeley plant at issue here was trans- ferred to the District of South Carolina on August 25, 2004. J.A. 9007. The appellants then filed a Third Amended Complaint on January 28, 2005. J.A. 43. BROWN v. NUCOR CORP. 31 trict court agreed and held "that the most reliable statistics are those gathered from [actual] promotion data occurring before this action was filed", and determined that "post-suit promo- tion data is entitled to minimal weight." J.A. 8986 n.4. With the post-lawsuit data favorable to Nucor excluded from consideration, the appellants’ experts calculated the actual versus expected standard deviation based on the actual job posting and bidding data from 2001 to 2003 as -1.48.6 In other words, the appellants were unable to produce statisti- cally significant evidence of race discrimination in promo- tions based on actual job posting data.7 See Castaneda v. Partida, 430 U.S. 482, 496 n.17 (1977) (recognizing that dif- ferences between the expected value and the observed number greater than two or three standard deviations are significant); Warren v. Halstead Indus., Inc., 802 F.2d 746, 759 n.17 (4th Cir. 1986) (recognizing that a standard deviation greater than two or three excludes "chance" as the cause of under- representation). Since the appellants could not prevail based on the actual promotions data from 2001 through 2003, they sought to create additional data that would dilute the actual data analy- 6 Our case law makes clear that standard deviation analysis is the proper method of statistical analysis. See EEOC v. Federal Reserve Bank of Rich- mond, 698 F.2d 633, 647 (4th Cir. 1983), rev’d on other grounds, sub nom. Cooper v. Fed. Reserve Bank of Richmond, 467 U.S. 867 (1984) ("[W]e have adopted the rule that the proper method for determining ‘legal significance’ on the basis of statistical evidence is through the use of the standard deviation analysis. . . .") (citing Moultrie v. Martin, 690 F.2d 1078 (4th Cir. 1982)). The -1.48 standard deviation resulted from data that controlled for two factors, the employees’ training and discipline. When those two factors were not controlled, the standard deviation was calcu- lated at -1.53, which the appellants’ expert testified was not a statistically significant difference. J.A. 5857. 7 The actual data from 2001 to 2003 came from "the posting notices of the jobs, the bids that were submitted by the individual employees, the prevailing bidder, and the personnel files and everybody that bid on those jobs." J.A. 9024. 32 BROWN v. NUCOR CORP. sis and produce the needed -2.00 or greater standard devia- tion. The appellants contended they could do so by comparing "the estimated percentage of blacks who sought promotions between December 1999 and December 2003 with the esti- mated percentage of blacks who received promotions during that period." Supra at 8. However, their proposed data substi- tute (the "2000 data") was only estimated for 2000 because the bulk of the time period (from 2001 to 2003), reflected actual job postings for which African-American employees were known to have applied.8 9 The appellants contended to the district court that if their version of the promotion statis- tics for the 2000 through 2003 period (four years) was consid- ered instead of three years (2001 to 2003), they could show the required standard deviation to make a prima facie case of discrimination. It is the appellants’ means of creating the 2000 data that validates the district court’s exercise of discretion in discounting that analysis because the 2000 data was without a valid foundation, purely speculative, and thus not entitled to probative weight. Even the appellants’ own experts agreed the 2000 data suffered from significant infirmities. Dr. Edwin Bradley, one of the appellants’ statistical experts, stated that "[a]ll statistics involve a comparison between a benchmark of what was expected to occur and what actually happened." J.A. 404 (emphasis added). Thus, there were two variables that the appellants’ experts, Dr. Bradley 8 The majority opinion recites the additional time period for statistical information as "1999-2000." The actual period was December 9, 1999 to January 4, 2001, not quite 13 months. J.A. 1162 n.16-17. I refer to that December 9, 1999 to January 4, 2001 time period as simply "2000" to more accurately reflect the single year for which the appellants sought to create a statistical data substitute. 9 As explained in more detail, infra, the "2000 data" consisted of 27 "change-of-status" forms for employment actions taken by Nucor from December 1999 to January 2001 and the appellants’ resulting projection of the percentages of African-American employees who applied for the employment positions represented by the change-of-status forms during this timeframe and were not promoted. BROWN v. NUCOR CORP. 33 and Dr. Liesl Fox, needed to formulate as to what was expected to occur in 2000 and what actually happened: (1) the number of posted job promotions available for bidding by the existing Nucor employees during 2000, and (2) whether those job promotions were "similarly situated" as defined by the district court (i.e., job promotions for which at least one African-American employee applied). The assumptions required for both variables undermine the validity of the 2000 data. In order to construct the first variable, the number of pro- motion positions available in 2000 open for employee bid, the appellants relied on twenty-seven "change-of-status" forms culled from the personnel files produced by Nucor. It is not clear from the record whether the twenty-seven change-of- status forms represent all, or only selected, jobs open for pro- motion in 2000 because these forms are not identified. Nucor argues that the change-of-status forms "are simply a company record which documents any employee’s change of status, whether the employee was promoted, demoted, received a standard pay increase, or was transferred." Appellee’s Br. at 36. Indeed, the change-of-status forms found in the record for 2000 fail to bolster the appellants’ claim that those forms reflect only promotion positions open to bid. For example, nine change-of-status forms appear in the record dated between December 1999 and January 2001 (including three in January 2001). J.A. 8397-99, 8406-07, 8416, 8673-74, 8701. Of these forms, one (J.A. 8399) simply reflects an increase in pay for an existing employee. Two (J.A. 8397, 8406) reflect completion of probation for an existing employee. Another (J.A. 8407), reflects a new hire and not a promotion. The remaining five reflect promotions of some type, though only one (J.A. 8673) identifies a promotion acquired through a bid process. On this record, it is difficult, if not impossible, to dis- cern whether the 2000 data based on the nebulous change-of- status forms proves those positions were promotion positions 34 BROWN v. NUCOR CORP. available for employee bidding and thus relevant to the for- mulation of statistical evidence for the appellants’ claims. Perhaps that is one reason the district court found not that the 2000 data represented twenty-seven positions reflecting pro- motion openings available for bidding but that "the plaintiffs searched through personnel files provided by Nucor and located twenty-seven positions filled between December 1999 and January 2001." J.A. 8984 (emphasis added). But even if one assumes the twenty-seven change-of-status forms represented actual promotion openings in 2000, Nucor’s expert, Dr. Finis Welch, testified "it is not known whether these were posted positions" (i.e. available for bid by employees). J.A. 5911 n.1. Thus, based on the record, or lack of it, to conclude that appellants established the first variable necessary for their statistical analysis to verify the 2000 data is highly dubious. For the second variable, whether the twenty-seven positions represent positions "similarly situated," Drs. Bradley and Fox assumed "that the racial composition of the bidding pool [applicants] for those jobs was the same as the weighted aver- age of the racial composition of the bidding pools (applicants) for the" 2001-2003 period. Supra at 9; J.A. 8984. However, through its discovery orders the district court had defined "similarly situated" jobs to mean those on which an African- American employee actually bid. Drs. Bradley and Fox could only confirm that one of the twenty-seven "promotions" in the 2000 data involved an African-American employee actually bidding on the job.10 J.A. 5853. Dr. Fox testified in her depo- sition as follows: Counsel: Okay. Now, what determine – what did 10 In addition to Nucor’s assertion that the twenty-seven change-of-status forms may not have represented a "promotion," Nucor also asserts that even if each change-of-status form represented a promotion, there is no way to know if the job was posted for bidding. BROWN v. NUCOR CORP. 35 you do to determine that these [27] job selections met the Court’s definition of the same or similar? Dr. Fox: Job title. Counsel: Okay. How many of these job selections involved – these 27 job selections involved an African-American actually bidding on the job? Dr. Fox: At least one. Counsel: Okay. Other than the one, how many involved? Dr. Fox: I don’t know. Counsel: You don’t know. You have no way of knowing that? Dr. Fox: No. Counsel: Okay. Is not the definition of similarly sit- uated, does that not also include a job on which an African-American actually bid? Dr. Fox: That was the Court’s order, yes. Counsel: Okay. So have you made any attempt in your analysis of these 27 to bring them within the Court’s definition of same or similar? Dr. Fox: Those were destroyed. It’s not possible to do that.11 11 The appellants proffered no evidence, from themselves or any other Nucor employee (past or present) that any African-American employee bid on any of the positions ostensibly represented by the 2000 data. 36 BROWN v. NUCOR CORP. J.A. 5852-54. Instead, Dr. Bradley and Dr. Fox determined the jobs were "similarly situated" based on the similarity of the job title alone. J.A. 5854. Dr. Welch criticized these assumptions because "it is not known whether . . . any African-American employees applied" for the twenty-seven job selections.12 J.A. 5911 n.1. Dr. Welch also believed that "Drs. Bradley and Fox may have missed two selections of African-Americans into a job that is included in the posting data," and that there were "African Americans, including a named plaintiff, who moved into a job prior to 2001 that is a job held by employees at the time they won a position in the posting data." J.A. 5911-12 n.2,3. Drs. Bradley and Fox also recognized that Bidding records [from 2001 to 2006] were provided only for "similarly situated" jobs for which at least one African-American bid. That means that any postings which were for "similarly situated" jobs but had no African-American bidders, and therefore all white bidders, were not included in our calculation. Thus, the African-American representation among bidders for "similarly situated" jobs provided by Nucor-Berkeley is necessarily inflated. J.A. 583. According to Dr. Welch, this necessarily means that by applying the 2001 to 2003 weighted average of the bid pool applicants to the 2000 data, "they are overstating the expected number of African-American selections." J.A. 5912. In Dr. Welch’s view, "there is no reason to assume that the applicant pools for each of [the twenty-seven] positions nec- essarily included at least one African American." J.A. 5912 12 Accordingly, Dr. Welch limited his analysis to "all of the job postings containing applicant information" because "[p]rior to 2001 we know nei- ther the applicant mix regarding race nor do we know anything regarding qualifications." J.A. 5893. BROWN v. NUCOR CORP. 37 n.4. Thus, the appellants’ support in the record for the second variable, that the twenty-seven positions used to create the 2000 data were positions for which it could be legitimately assumed at least one African-American applied, is as dubious as the first variable. The appellants certainly were entitled to proffer valid statis- tical evidence for 2000 upon which to expand the standard deviation analysis. That being said, however, no precedent or principle of law requires inherently unreliable evidence to be given evidentiary weight. The appellants chose a method of statistical proof, but based it on evidence a trier of fact could determine was inherently faulty.13 Consequently, a reasonable jurist could have determined the appellants failed to meet their burden of proof for the Rule 23(a) factors through the proffered statistical evidence. Accordingly, the district court did not abuse its discretion by failing to give the appellants’ 2000 data, and the resulting standard deviation analysis, evi- dentiary weight and denying certification.14 IV. Even if we assume that the 2000 data (and its resulting standard deviation analysis) was entitled to some evidentiary weight, the precedent cited by the majority opinion does not support the conclusion that the district court abused its discre- tion in refusing certification as to the disparate treatment and disparate impact claims. Our decision in United States v. 13 The district court’s observations that the appellants’ "assumptions may be reasonable and the statistics based thereon may be relevant" means no more than it says. J.A. 8985. The district court, in context, was merely acknowledging the theoretical possibility that the appellants could present relevant evidence — a proposition disproved by the evidence they did present. 14 Insomuch as neither the appellants’ direct evidence or statistical evi- dence met the Rule 23 burden of proof, it was not an abuse of discretion to determine the two deficient forms of evidence could not, taken together, meet the appellants’ burden of proof. 38 BROWN v. NUCOR CORP. County of Fairfax, 629 F.2d 932 (4th Cir. 1980), does not mandate that a district court must give evidentiary weight to every datum a prospective class representative presents just to increase the strength of their statistical proof. Moreover, there are significant distinctions between the case at bar and County of Fairfax. A distinction of particular importance is that the "missing" data in County of Fairfax involved three out of the four years from which the statistical analysis was to be made. There was no credible basis in that case by which to use the one year of actual data to foster a valid analytical sample. One year’s data was simply insuffi- cient upon which to make a statistical analysis. Accord EEOC v. Am. Nat’l Bank, 652 F.2d 1176, 1195 (4th Cir. 1981). In contrast, the case at bar is the exact opposite, with three years of actual data. There is no precedent that requires a court in all cases to add an additional period of data, particularly data lacking in evidentiary credibility.15 Moreover, Fairfax County, as a municipality, was under a statutory duty to maintain employment data in the 1970s, a duty which it failed to fulfill. No such duty is applicable to Nucor. See County of Fairfax, 629 F.2d at 937 n.4 ("In violation of the record keeping regu- lations of the Revenue Sharing Act and the Crime Control Act, defendants had destroyed pre-1978 applications for employment."). Another distinction is that County of Fairfax involved an appeal from the district court’s judgment on the merits of the case, not a denial of class certification under Rule 23. Thus, the case at bar involves a different legal analysis and is sub- ject to an abuse of discretion standard of review on appeal. 15 The majority opinion seems to imply some bad purpose on the part of Nucor because the 2000 data on actual bid promotions was unavailable. Appellants have never made a spoliation of evidence claim or pointed to any legal duty on the part of Nucor to have maintained the precise records found absent. While appellants were certainly entitled to seek to create substitute data for 2000, nothing entitled them to a data substitute based on insufficient evidence and invalid assumptions. BROWN v. NUCOR CORP. 39 Further, although County of Fairfax involved claims of dis- crimination "in recruitment, hiring, assignments, and promo- tions," 629 F.2d at 936, the sufficiency of the applicant flow data in that case involved only the hiring data, not promotions data. Unlike hirings, which came from the surrounding com- munity in County of Fairfax, Nucor only promotes from within its existing workforce. The data universe for hiring in County of Fairfax was thus much larger and without restrict- ing variables that made projections on estimated data prob- lematic. However, in the case at bar, the data universe contained restrictive variables limiting extrapolations because only current Nucor employees who were qualified and who in fact bid on a promotion could comprise that group. County of Fairfax does not stand for the proposition that class certification claimants are entitled to a waiver of the rules of evidence for purposes of meeting their burden of proof. The district court’s judgment was not based on errone- ous legal principles and was thus not an abuse of discretion. V. When an appellate court reviews a trial court’s determina- tion upon an abuse of discretion standard, it must accord that court’s factual conclusions supported by the record proper deference. See United States v. Pittman, 209 F.3d 314, 316 (4th Cir. 2000) (The abuse of discretion "standard of review mandates a significant measure of appellate deference to the judgment calls of trial courts."); see also United States v. Mason, 52 F.3d 1286, 1289 (4th Cir. 1995) ("Under the abuse of discretion standard, this Court may not substitute its judg- ment for that of the district court; rather, we must determine whether the court’s exercise of discretion, considering the law and the facts, was arbitrary or capricious."). This Court has recently stated that "[a]t its immovable core, the abuse of discretion standard requires a reviewing court to show enough deference to a primary decision-maker’s judg- 40 BROWN v. NUCOR CORP. ment that the court does not reverse merely because it would have come to a different result in the first instance." Evans v. Eaton Corp. Long Term Disability Plan, 514 F.3d 315, 322 (4th Cir. 2008). Based on the record in this case, a reasonable jurist could plainly find the appellants’ direct and statistical evidence failed to show commonality and typicality for the purported class sufficient to meet the certification threshold. As the district court’s judgment was neither based on a "clearly erroneous factual finding" nor "erroneous legal prin- ciples," it should be accorded the deference due under the abuse of discretion standard of review. In my view, the majority opinion fails to accord the district court proper deference in this case and, instead, reaches fac- tual conclusions de novo that are not within the proper scope of review by an appellate court. The majority’s opinion could reasonably be interpreted to require class certification so long as future plaintiffs seeking class certification can produce any data set exceeding two standard deviations without regard to the unreliability of that data or the process by which it was derived. Equally troubling is the prospect that certification is required no matter how attenuated and insufficient a plain- tiff’s direct evidence is shown to be. Such an extension of our case law ignores the prudent judgment and proper discretion of a district court, which has seen and heard the prolific evi- dence first-hand, and constrains the district court to a merely ministerial act devoid of evidentiary grounding. That result is incongruent with an appellate court’s adherence to the stan- dard of review for an abuse of discretion. I would thus affirm the district court’s judgment denying class certification on the appellants’ disparate impact and disparate treatment claims for the reasons set forth above, and respectfully dissent from the majority opinion in that regard.