Searching over 5,500,000 cases.


searching
Buy This Entire Record For $7.95

Download the entire decision to receive the complete text, official citation,
docket number, dissents and concurrences, and footnotes for this case.

Learn more about what you receive with purchase of this case.

Equal Employment Opportunity Commission v. Freeman

United States District Court, D. Maryland

September 4, 2015

EQUAL EMPLOYMENT OPPORTUNITY COMMISSION, Plaintiff,
v.
FREEMAN, Defendant

Page 561

[Copyrighted Material Omitted]

Page 562

[Copyrighted Material Omitted]

Page 563

[Copyrighted Material Omitted]

Page 564

          For EEOC, Plaintiff: Debra Michele Lawrence, Ronald Lynn Phillips, LEAD ATTORNEY, Maria Kate Boehringer, U.S. Equal Employment Opportunity Commission, Baltimore, MD; Keyana Capri Laws, Melanie Marie Peterson, Philip Matthew Kovnat, U.S. Equal Employment Opportunity Commission, Philadelphia, PA.

         For Freeman, Defendant: W Randolph Teslik, LEAD ATTORNEY, Donald R Livingston, John T Koerner, Akin Gump Strauss Hauer and Feld LLP, Washington, DC.

         For U.S. Office of Personnel Management (" OPM" ), Intervenor: Alex S Gordon, LEAD ATTORNEY, Office of the United States Attorney, Baltimore, MD.

Page 565

         MEMORANDUM OPINION

         ROGER W. TITUS, UNITED STATES DISTRICT JUDGE.

         World-renowned poker expert Kenny Rogers once sagely advised, " You've got to to know when to hold 'em. Know when to fold 'em. Know when to walk away." [1] In the Title VII context, the plaintiff who wishes to avoid paying a defendant's attorneys' fees must fold 'em once its case becomes so groundless that continuing to litigate is unreasonable, i.e. once it is clear it cannot have a winning hand. In this case, once Defendant Freeman revealed the inexplicably shoddy work of the EEOC's expert witness in its motion to exclude that expert, it was obvious Freeman held a royal flush, while the EEOC held nothing. Yet, instead of folding, the EEOC went all in and defended its expert through extensive briefing in this Court and on appeal. Like the unwise gambler, it did so at its peril. Because the EEOC

Page 566

insisted on playing a hand it could not win, it is liable for Freeman's reasonable attorneys' fees.

         BACKGROUND

         I. Facts[2]

         Freeman, as a regular part of its hiring process, conducted criminal background checks on all applicants who were offered a position, and conducted credit background checks on applicants who were offered financially sensitive positions.[3] EEOC v. Freeman, 961 F.Supp.2d 783, 787 (D. Md. 2013). Importantly, applicants were not turned away for any negative information. Rather, Freeman limited in scope the type of negative information that would disqualify an applicant. For example, as to the criminal background check, Freeman generally did not consider arrests, but only convictions that had occurred within the past seven years.[4] Id. at 788. Furthermore, Freeman did not consider all convictions, but only those for certain crimes. Id. Similarly, with regard to credit checks, only certain negative items would exclude an applicant from being hired. Id. at 789. The Equal Employment Opportunity Commission (the " EEOC" ) alleged that Freeman's use of background checks was discriminatory. Id. at 790.

         II. Procedural History

         After being rejected by Freeman based on information in her credit report, Katrina Vaughn filed a charge of discrimination with the EEOC. Id. at 789. Based on this charge, the EEOC filed a complaint in this Court alleging that Freeman's use of background checks had a disparate impact on African-American, Hispanic, and male job applicants. Id. On April 27, 2010, the Court dismissed all claims relating to hiring decisions made before March 23, 2007, the date 300 days before the original charge of discrimination was filed on January 17, 2008. ECF No. 19. On January 31, 2011, the Court granted Freeman's partial motion for summary judgment and dismissed all claims relating to hiring decisions based on criminal background checks made prior to November 30, 2007, the date 300 days before September 25, 2008, when the EEOC first notified Freeman that it was expanding its investigation to include race discrimination based on the use of criminal background checks. ECF No. 43. On August 24, 2012, the EEOC voluntary dismissed with prejudice its claim that Freeman unlawfully discriminated against Hispanics.[5] ECF No. 95.

         On July 18, 2012, the EEOC served Freeman with the expert report of Dr. Kevin R. Murphy. ECF No. 186 at 9. After Freeman pointed out various errors in the report, the EEOC served a second expert report that purported to fix those errors on July 26, 2012.[6] ECF No. 186-17. On December 18, 2012, Freeman moved to exclude Dr. Murphy's report and

Page 567

the report of a second expert, Dr. Beth Huebner, arguing that a significant number of errors and omissions within the reports made them unreliable, and thus inadmissible.[7] ECF No. 108. On December 21, 2012, Freeman moved for summary judgment, arguing that, without reliable statistical evidence of a disparate impact, the EEOC could not make out a prima facie case of discrimination. ECF No. 114. The EEOC responded to each motion. ECF Nos. 121, 126. In its opposition to the motion to exclude the expert reports, the EEOC submitted a supplemental expert report from Murphy that purported to moot Freeman's criticisms. ECF No. 121 at 7. The EEOC also contended that Freeman was the source of any errors in the expert report. Id. at 13. The EEOC also proffered yet another report by Murphy at the motions hearing on June 19, 2013. Freeman, 961 F.Supp.2d at 797.

         On August 9, 2013, the Court excluded the EEOC's experts and granted Freeman's motion for summary judgment. ECF No. 150. Freeman subsequently moved for attorneys' fees. ECF No. 154. After filing its appeal, ECF No. 166, the EEOC moved to stay proceedings related to Freeman's motion for attorneys' fees. ECF No. 172. The Court granted that motion over Freeman's opposition. ECF No. 176. The EEOC also moved to amend the record for appeal with Murphy's report tendered at the motions hearing, ECF No. 177, and the Court granted that motion, again over Freeman's opposition. ECF No. 181.

         The Fourth Circuit affirmed the Court on February 20, 2015. EEOC v. Freeman, 778 F.3d 463 (4th Cir. 2015). Freeman submitted a renewed motion for attorneys' fees that included those fees incurred on appeal, ECF No. 186, and has further supplemented its request to account for fees incurred defending its fee request. ECF No. 197.

         ANALYSIS

         The EEOC has, understandably, taken a keen interest in employers' use of background checks to make hiring decisions.[8] " Because of the higher rate of incarceration of African-Americans than Caucasians, indiscriminate use of criminal history information might have the predictable result of excluding African-Americans at a higher rate than Caucasians." Freeman, 961 F.Supp.2d at 786. Similarly, because blacks typically fare worse in many indicators of economic health, such as income and poverty rate, than whites, credit checks may disproportionately exclude blacks. The EEOC is certainly entitled to attempt to police the use of background checks through litigation, and to attempt to use litigation to challenge whether an employer's use of background checks is " job-related for the position in question and consistent with business necessity." Id. at 791.

         However, before the EEOC can get to the question of business necessity, it must actually make out a prima facie case of disparate impact, and before it can make out a prima facie case of disparate impact, it must actually produce reliable statistical evidence showing that a particular employment practice has a disparate impact on a protected class. Id. at 791. The EEOC is not entitled to require an

Page 568

employer to demonstrate business necessity when it is consistently unable to produce any reliable evidence of a disparate impact. See EEOC v. Propak Logistics, Inc., 746 F.3d 145, 157 (4th Cir. 2014) (" The reference to statutory goals and missions, however, cannot be divorced from the manner in which those purposes are implemented." ) (Wilkinson, J., concurring). Continuing to litigate once the faults in the EEOC's case were laid bare was unreasonable, and Freeman is entitled to reasonable attorneys' fees from that point forward.

         I. The Availability of Attorneys' Fees Under Title VII

          Title VII provides that " the court, in its discretion, may allow the prevailing party...a reasonable attorney's fee (including expert fees) as part of the costs, and the Commission and the United States shall be liable for costs the same as a private person." 42 U.S.C. § 2000e-5(k). By its terms, this provision allows either a prevailing plaintiff or a prevailing defendant to recover attorneys' fees. However, the award of attorneys' fees to a prevailing plaintiff involves different considerations than an award to a prevailing defendant. The prevailing plaintiff is acting as a " private attorney general" in vindicating an important federal interest against a violator of federal law, and therefore " ordinarily is to be awarded attorney's fees in all but special circumstances." Christiansburg Garment Co. v. EEOC, 434 U.S. 412, 416-17, 98 S.Ct. 694, 54 L.Ed.2d 648 (1978). The opposite is true of a prevailing defendant. A prevailing defendant not only is not vindicating any important federal interest, but the award of attorneys' fees to prevailing defendants as a matter of course would undermine that interest by making it riskier for " private attorney generals" to bring claims.[9] Id. at 422. Accordingly, before a prevailing defendant may be awarded fees, it must demonstrate that a plaintiff's claim was " frivolous, unreasonable, or groundless, or that the plaintiff continued to litigate after it clearly became so." Id. Importantly, however, this standard does not require a plaintiff to have acted in bad faith. Id. at 421; see also Arnold v. Burger King Corp., 719 F.2d 63, 66 (4th Cir. 1983) (" A plaintiff acting in good faith may nevertheless be assessed fees if his claim is groundless or frivolous" ).

          In assessing whether to assess attorneys' fees against a plaintiff, it is important that a court not engage in minute, post hoc questioning of a plaintiff's claims and conduct in light of plaintiff's failure on the merits. EEOC v. Great Steaks, Inc., 667 F.3d 510, 517 (4th Cir. 2012). However, there is " neither a precise test to be used, nor a specific quantum of proof required, in determining whether a plaintiff's claim was unreasonable." Propak Logistics, 746 F.3d at 151 . A decision to award fees is committed to the discretion of the Court, which is in the best position to assess the considerations relevant to the conduct of litigation. Id.

         II. Freeman is Entitled to Reasonable Attorneys' Fees

         A. The EEOC Needed to Present Reliable Statistical Evidence that Freeman's Policies had a Disparate Impact to Make Out a Prima Facie Case of Discrimination

          It is black letter Title VII law that, in order to make out a prima facie case of

Page 569

disparate impact, a plaintiff must present " reliable and accurate statistical analysis performed by a qualified expert." Freeman, 961 F.Supp.2d at 786. This evidence is necessary because a disparate impact claim alleges that an otherwise neutral policy has a negative effect on the ability of a protected group to get hired. Accordingly, there must be a statistical disparity " between the number of protected class members in the qualified applicant group and those in the relevant segment of the work-force." Id. at 791.

         Simply reciting this black letter law meets several of the EEOC's arguments in its opposition to an award of attorneys' fees. First, the EEOC asserts that its case is not factually groundless. ECF No. 191 at 5. The basis for this argument appears to be that it may have been possible for the EEOC to show evidence of a disparate impact, and it just so happened to fail here. Id. at 5-7. But the possible existence of evidence somewhere to back up a claim does not constitute reasonable grounds for continuing to litigate that claim. Cases are litigated based on evidence actually produced, not hunches, however reasonable those hunches may be. Without reliable statistical evidence produced by an expert, which is entirely absent here, the EEOC simply could not make out its prima facie case. The factual lack of reliable expert testimony showing a disparate impact was fatal to the EEOC's disparate impact claim in this particular case, even if in some alternate universe (or simply by hiring a better expert), the EEOC could have made out a claim.

         The EEOC also could not reasonably rely on general population statistics to make out its prima facie case. It is true that, in some circumstances, general population statistics will suffice to show that a particular policy has a disparate impact. See Griggs v. Duke Power Co., 401 U.S. 424, 430-32, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). However, an examination of Griggs, the first Title VII disparate impact case, puts the folly to the EEOC's argument that it could have done so here. In Griggs, an employer had a facially neutral requirement that an applicant be a high school graduate. Id. at 427-28. The Supreme Court reversed the district court's dismissal of the complaint, reasoning that, despite the lack of discriminatory intent, the fact that African-Americans had lower high school graduation rates in the employer's state meant that the policy had a discriminatory impact, and that Title VII allowed such claims. Id. at 429-32. The difference between Griggs and this case is obvious. In Griggs, the challenged policy, a high school graduation requirement, and the general population statistic, high school graduation rates, were the exact same. Griggs simply does not apply to a case where the challenged policy is not congruent with the general population statistic being offered as proof of a disparate impact.

         In addition to Griggs, the EEOC cites several cases for the proposition that " [l]ower federal courts have also accepted external population proof in a variety of contexts despite arguments that they did not perfectly match the facts of the case." ECF No. 191 at 10. An examination of each of these cases reveals that none supports this argument, even by analogy.

         In Peightal v. Metropolitan Dade County, 26 F.3d 1545 (11th Cir. 1994), a white applicant challenged, on equal protection grounds, a fire department's affirmative action policies. The Eleventh Circuit noted that, for a government employer to show a compelling interest in combating discrimination that would justify using affirmative action, " statistical comparisons between the employer's work force and the composition of the relevant population are probative of a pattern of discrimination." Id. at 1553.

Page 570

The Eleventh Circuit did allow the defendant fire department to present statistical disparities between the general relevant population of Hispanics and the number of Hispanics who passed a qualifications exam, notwithstanding that those statistics did not account for supposed disparities in graduation rates for Hispanics, graduation from high school being a prerequisite to take the exam. Id. at 1556 n.17. It accepted this statistic, however, because (a) it determined that the relevant statistical refinement was not available, as there was no data on Hispanic high school graduation rates in the area, and (b) the plaintiff had not shown beyond speculation that it would matter. Id. Moreover, the Eleventh Circuit also noted the numerous refinements the fire department had made to ensure that the external population statistics reasonably matched the relevant applicant pool. Id. at 1553-57. In fact, the case had previously been remanded for those refinements precisely because of " the relative undesirability of general population figures." Id. at 1554.

         In Carpenter v. Boeing Co., 456 F.3d 1183, 1197 (10th Cir. 2006), the Tenth Circuit noted that " the population selected for statistical analysis need not perfectly match the pool of qualified persons." However, the EEOC could not have reasonably relied on that case. First, the plaintiff's expert in that case did attempt to set out a specific methodology for constructing a representative sample which, as explained below, the EEOC's expert here never did. Id. at 1194-95 (describing expert's process for defining similarly situated employees and measuring the appropriate statistics to determine disparate impact). Moreover, in that case, the Tenth Circuit still determined that the expert's conclusions were unreliable. Id. at 1198-99. It is curious that the EEOC would rely on a case that is directly contrary to its position to support the notion that its position was legally reasonable.

         The cases cited by the EEOC to support the reasonableness of its reliance on general population statistics are simply inapplicable. They support the proposition that general population statistics will be allowed where those statistics are completely congruent with an employment criteria, as in Griggs, or where a limited number of data points, in an otherwise refined sample, do not perfectly align with the general population data. The cases do not support the proposition that a plaintiff is entitled to rely on general population statistics divorced from any reference to the defendant's employment practices without any attempt to control for those practices.

         It has been clear since at least 1989, when the Supreme Court decided Wards Cove Packing Co., Inc. v. Atonio, 490 U.S. 642, 109 S.Ct. 2115, 104 L.Ed.2d 733 (1989), that refined, statistical comparisons are necessary to make out a disparate impact claim. Id. at 650-51. The EEOC's attempt to use general population statistics illustrates the rationale underlying Wards Cove, which is that an employer should not be held liable for disparities that are not caused by its employment practices. Id. at 651 (refusing to hold employer liable for disparities caused by " reasons that are not [the employer's] fault" ). The EEOC focused on general background statistics such as arrest and incarceration rates, and the racial disparity in those with " good" or " bad" credit histories. E.g. ECF No. 121 at 15-16; ECF No. 186-14 at 23. But Freeman was not generally concerned about arrests at all, and was not concerned about all convictions, or about whether those convictions led to incarceration. Instead, Freeman was concerned about only particular types of convictions. Similarly, Freeman was not concerned about whether an applicant had " good" or " bad" credit. Rather, Freeman was concerned about

Page 571

only certain indicators of bad credit. The EEOC failed to produce any general population statistics that approximated the criteria used by Freeman to make hiring decisions. The use of general population statistics also failed to take into account the subjective component of Freeman's policy, which gave employees who failed a background check an opportunity to explain the circumstances before Freeman denied them a position. ECF No. 108-4 at 15. And, indeed, the EEOC should have been aware that any reliance on general population statistics would be flawed, since population statistics demonstrated adverse outcomes for Hispanics generally, yet its expert failed to find a statistically significant disparate impact on Hispanics from Freeman's specific use of background checks. Freeman, 961 F.Supp.2d at 798-99.

         Even if general population statistics were appropriately tailored in this case, the EEOC overlooks a key issue, which is noted in the cases cited by the EEOC, yet ignored by it: in order to use general population statistics that do not match the particular employment criteria instead of statistics from the employer, the missing statistics must be impossible or extremely difficult to obtain. See Carpenter, 456 F.3d at 1199 (noting, in excluding statistics that were insufficiently tailored, that " Plaintiffs have not established that the data necessary to establish the impact on CBA-qualified workers were unavailable" ); Peightal, 26 F.3d at 1556 n.17 (noting that district court properly included undifferentiated statistics where " no differentiated population statistics exist based upon the level of education of minorities" ); see also Wards Cove, 490 U.S. at 651 (" Alternatively, in cases where such labor market statistics will be difficiult if not impossible to ascertain, we have recognized that certain other statistics...are equally probative for this purpose." ).[10]

         Here, it was not impossible for the EEOC to obtain the relevant data from Freeman that it needed to do the appropriate statistical analysis. The data necessary to determine the impact of Freeman's use of background checks was available, because it was provided by Freeman in discovery. Freeman, 961 F.Supp.2d at 793 (noting that Freeman " clearly demonstrated that it properly supplied the EEOC during discovery with complete background check logs for the entire period covered by the complaint." ); Freeman, 778 F.3d at 469 (Agee, J., concurring) (same). Accordingly, the EEOC could not have reasonably hoped to rely on general population statistics in the event that its flawed expert reports were excluded, because it could not have possibly demonstrated that the relevant data was unavailable.

         The upshot, then, is that the EEOC could not reasonably pursue its claims unless

Page 572

it could produce an arguably reliable expert report with statistics showing a disparate impact caused by Freeman's particular employment practices. As explained in the next section, the EEOC wholly failed to do this.

         B. It was Unreasonable for the EEOC to Continue to Litigate on the Basis of Obviously Flawed Expert Reports

         As made clear in the opinions of this Court and the Fourth Circuit, Murphy's report was severely flawed. His conclusions were drawn from an unrepresentative sample, even though data was available from which a representative sample could be drawn. Freeman, 961 F.Supp.2d at 793-94. He excluded data from much of the relevant time period, and excluded data from most of Freeman's branches. Id. at 794-96. He cherry-picked data. Id. at 795. Worse yet , his reports contained a " mind-boggling number of errors," most of which were introduced by Murphy. Id. at 796. The question is not whether Murphy's reports were flawed, because they clearly were. The question is if they were so flawed that it was unreasonable for the EEOC to litigate in reliance on them.

         1. The EEOC should have been aware of the unreliability of Murphy's analysis no later than after ...


Buy This Entire Record For $7.95

Download the entire decision to receive the complete text, official citation,
docket number, dissents and concurrences, and footnotes for this case.

Learn more about what you receive with purchase of this case.