UNITED STATES COURT OF APPEALS FOR THE EIGHTH CIRCUIT ______________________________ Nos. 05-4183, 05-4311 ______________________________ EQUAL EMPLOYMENT OPPORTUNITY COMMISSION, Plaintiff-Appellee/Cross-Appellant, v. THE DIAL CORPORATION, Defendant-Appellant/Cross-Appellee. _______________________________________________________ On Appeal from the United States District Court for the Southern District of Iowa, Eastern Division Civil Action No. 3-02-CV-10109 The Hon. Ronald C. Longstaff, U.S.D.J., Presiding _______________________________________________________ U.S. EQUAL EMPLOYMENT OPPORTUNITY COMMISSION'S ANSWER BRIEF AS APPELLEE and OPENING BRIEF AS CROSS-APPELLANT _______________________________________________________ JAMES L. LEE SUSAN R. OXFORD Deputy General Counsel Attorney U.S. EQUAL EMPLOYMENT LORRAINE C. DAVIS OPPORTUNITY COMMISSION Acting Associate General Counsel 1801 L Street, N.W., Room 7010 Washington, D.C. 20507 Tel. (202) 663-4791 Fax (202) 663-7090 SUMMARY OF THE CASE AND REQUEST FOR ORAL ARGUMENT In this lawsuit, EEOC alleged, and the district court found, that Dial's use of a physical ability test to screen job applicants for a meat processing plant violated Title VII because it had a disparate impact against women and Dial failed to justify it as a "business necessity." EEOC further alleged, and the jury concurred, that Dial's continued use of this test, even after realizing it was screening out women, constituted intentional discrimination in violation of Title VII. The district court denied Dial's motion for judgment as a matter of law, and Dial has appealed both rulings. Dial alleges that because it instituted this preemployment test out of an "honest belief" that it would reduce workplace injuries, it is entitled to prevail on both claims even though the evidence shows no reduction in injuries that can be traced to the test, and even though Dial's managers never attempted to ascertain whether any decrease in injuries was due to the test. Dial also appeals several rulings on monetary relief. The EEOC cross-appeals denial of backpay to one woman based on her criminal record. The record relevant to Dial's appeal includes extensive documentary evidence and witness testimony, including three expert witnesses who testified at length. In addition, EEOC's cross-appeal raises legal questions about the proper application of the "after-acquired evidence" doctrine in these novel circumstances. EEOC believes oral argument would assist this Court in resolving this matter, and respectfully requests 20 minutes per party to address these important issues. TABLE OF CONTENTS page SUMMARY OF THE CASE AND REQUEST FOR ORAL ARGUMENT TABLE OF CONTENTS. . . . . . . . . . . . . . . . . . . . . . . .i TABLE OF AUTHORITIES . . . . . . . . . . . . . . . . . . . . . iv STATEMENT OF JURISDICTION. . . . . . . . . . . . . . . . . . . .1 STATEMENT OF THE ISSUES AND APPOSITE CASES . . . . . . . . . . .1 STATEMENT OF THE CASE. . . . . . . . . . . . . . . . . . . . . .2 A. Course of Proceedings. . . . . . . . . . . . . . . . .2 B. Statement of Facts . . . . . . . . . . . . . . . . . .3 1. Background . . . . . . . . . . . . . . . . . . . .3 2. Evidence at Trial. . . . . . . . . . . . . . . . .7 C. District Court Decisions . . . . . . . . . . . . . . 19 1. Denial of Dial's Motion for Judgment as a Matter of Law19 2. Dial's Liability for Disparate Impact Discrimination20 3. Remedies . . . . . . . . . . . . . . . . . . . . 21 SUMMARY OF ARGUMENT. . . . . . . . . . . . . . . . . . . . . . 24 ARGUMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 I. THE DISTRICT COURT'S FINDING THAT DIAL DID NOT PROVE A BUSINESS NECESSITY FOR THE WTS IS NOT CLEARLY ERRONEOUS. . . . . . . . . . . . . . . . . . . . . . . . 25 A. Standard of Review . . . . . . . . . . . . . . . . . 25 B. The Parties's Evidentiary Burdens in a Disparate Impact Claim . . . . . . . . . . . . . . . . . . . . 26 C. The Evidence Supports the District Court's Determination27 1. The District Court's Finding that Dial Failed to Establish Content Validity Is Not Clearly Erroneous . . . 31 2. The District Court's Finding that Dial Failed to Establish Criterion-Related Validity Is Not Clearly Erroneous . . . . . . . . . . . . . . . 35 3. The District Court Did Not Improperly Shift the Burdens of Proof. . . . . . . . . . . . . . . . 42 II. THE DISTRICT COURT PROPERLY DENIED DIAL'S MOTION FOR JMOL BECAUSE THE EVIDENCE SUFFICED FOR THE JURY TO FIND DIAL ENGAGED IN A PATTERN OR PRACTICE OF INTENTIONAL DISCRIMINATION AGAINST WOMEN BEGINNING APRIL 2001.. . . . . . . . . . . . . . . . . . 44 A. Standard of Review . . . . . . . . . . . . . . . . . 44 B. The Record Contains Sufficient Evidence to Support the Jury's Verdict.. . . . . . . . . . . . . . . . . . . 45 III. THE DISTRICT COURT DID NOT ABUSE ITS DISCRETION IN AWARDING BACKPAY AND BENEFITS TO CLAIMANTS REGARDLESS OF WHETHER THEY COMPLETED THE WTS, WITHOUT REGARD TO AVERAGE LENGTH OF TENURE, AND WITHOUT REQUIRING PROOF OF ACTUAL MEDICAL EXPENSES.. . . . . . . . . . . . . . . . . . . . . . . . 52 A. Standard of Review . . . . . . . . . . . . . . . . . 52 B. The District Court Did Not Abuse its Discretion in Awarding Backpay and Benefits to 53 Claimants.. . . . . . . . 52 CROSS APPEAL . . . . . . . . . . . . . . . . . . . . . . . . . 57 SUMMARY OF ARGUMENT . . . . . . . . . . . . . . . . . . . . . 57 ARGUMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 THE DISTRICT COURT ERRED AS A MATTER OF LAW IN DENYING BACKPAY TO ONE CLASS MEMBER BASED ON AFTER-ACQUIRED EVIDENCE OF HER CRIMINAL RECORD.. . . . . 58 A. Standard of Review . . . . . . . . . . . . . . . . . 58 B. The District Court Erroneously Denied Backpay Based on the Claimant's Criminal Record. . . . . . . . . . 58 1. The District Court's Decision Conflicts with McKennon.. . . . . . . . . . . . . . . . . 58 2. Dial Failed to Establish the Requisite Factual Predicate.. . . . . . . . 63 CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . 65 CERTIFICATE OF COMPLIANCE. . . . . . . . . . . . . . . . . . . 66 ADDENDUM A. District Court's September 14, 2005, Order Denying Backpay for HeatherWright-Bradley B. EEOC 1970 Guidelines on Employee Selection Procedures 35 Fed.Reg. 12333-12336 C. EEOC Exhibits 14, 30, 36 CERTIFICATE OF SERVICE TABLE OF AUTHORITIES Cases page Albemarle Paper Co. v. Moody, 422 U.S. 405 (1975). .1, 29, 53, 56 Allen v. Entergy Corp., 181 F.3d 902 (8th Cir. 1999) . . . . . 29 Anderson v. Bessemer City, 470 U.S. 564 (1985) . . . . . . 35, 38 Ass'n of Mex.-Amer. Educators v. State of California, 231 F.3d 572 (9th Cir. 2000). . . . . . . . . . . . . . . 25 Bernard v. Gulf Oil, 890 F.2d 735 (5th Cir.1989) . . . . . . . 25 Bradley v. Pizzaco of Neb., 7 F.3d 795 (8th Cir. 1993) 26, 42, 44 Chambers v. Omaha Girls Club, 834 F.2d 697 (8th Cir. 1987). . 26 Craik v. Minn. State Univ., 731 F.2d 465 (8th Cir. 1984) . 47, 49 Davey v. City of Omaha, 107 F.3d 587 (8th Cir. 1997) . . . . . 25 Donnell v. General Motors, 576 F.2d 1292 (8th Cir. 1978) . . . 42 Dothard v. Rawlinson, 433 U.S. 321 (1977). . . . . . . . . 27, 39 E.E.O.C. v. Delight Wholesale, 973 F.2d 664 (8th Cir. 1992)52, 53 EEOC v. Farmer Bros., 31 F.3d 891 (9th Cir. 1994). . . . . . . 63 EEOC v. Joe's Stone Crab, 220 F.3d 1263 (11th Cir. 2000)25, 46, 48 EEOC v. McDonnell Douglas Corp., 191 F.3d 948 (8th Cir.1999) . . . . . . . . . . . . . . . 46 EEOC v. Rath Packing, 787 F.2d 318 (8th Cir. 1986) .1, 43, 53, 56 Cases (cont'd) page Fiedler v. Indianhead Truck Line, Inc., 670 F.2d 806 (8th Cir. 1982). . . . . . . . . . . . . . . 55 Firefighter's Inst. for Racial Equality v. City of St. Louis, 220 F.3d 898 (8th Cir. 2000). . . . . . . . . 26, 34, 35, 41 Firefighter's Inst. for Racial Equality v. City of St. Louis, 616 F.2d 350 (8th Cir. 1980). . . . . . . . . . . . . . . 40 Firefighter's Inst. for Racial Equality v. City of St. Louis, 549 F.2d 506 (8th Cir. 1977). . . . . . . . . . . . . . . 29 Franks v. Bowman Transp., 424 U.S. 747 (1976). . . . . . . . . 52 Griggs v. Duke Power Co., 401 U.S. 424 (1971). . . . . 26, 28, 29 Hamer v. City of Atlanta, 872 F.2d 1521 (11th Cir.1989). . . . 25 Hartley v. Dillard's, Inc., 310 F.3d 1054 (8th Cir. 2002). . . 55 Hartman Bros. Heating & Air Conditioning v. NLRB, 280 F.3d 1110 (7th Cir. 2002) . . . . . . . . . . . . 59, 60 Hawkins v. Anheuser-Busch, 697 F.2d 810 (8th Cir.1983) . . 26, 43 Hazelwood Sch. Dist. v. United States, 433 U.S. 299 (1977) . . 47 Int'l Bhd. of Teamsters v. United States, 431 U.S. 324 (1977)1, 46, 47, 49, 51 Int'l Bhd. of Elec. Workers v. Miss. Power & Light, 442 F.3d 313 (5th Cir. 2006). . . . . . . . . . . . . . . 43 Isabel v. City of Memphis, 404 F.3d 404 (6th Cir. 2005). . . . 34 Johnson v. AT&T, 422 F.3d 756 (8th Cir. 2005). . . . . . . . . 51 King v. Staley, 849 F.2d 1143 (8th Cir.1988) . . . . . . . . . 52 Cases (cont'd) page Kirby v. Colony Furniture, 613 F.2d 696 (8th Cir. 1980). . 40, 43 Kratzer v. Rockwell Collins, 398 F.3d 1040 (8th Cir. 2005) . . 51 Kraul v. Iowa Methodist Med. Ctr., 95 F.3d 674 (8th Cir. 1996) 26 Lanning v. Southeastern Penn. Transp. Auth., 181 F.3d 478 (3d Cir. 1999) . . . . . . . . . .1, 26, 33, 37 Mardell v. Harleysville Life Ins., 65 F.3d 1072 (3d Cir. 1995)59, 61 McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973). . . . . 50 McKennon v. Nashville Banner Pub., 513 U.S. 352 (1995)2, 57, 58, 60, 62, 63 Melendez v. Ill. Bell Tel. Co., 79 F.3d 661 (7th Cir.1996) . . 25 Mems v. City of St. Paul, 327 F.3d 771 (8th Cir. 2003) . . . . 58 Morgan v. United Parcel Serv. of Amer., 380 F.3d 459 (8th Cir. 2004)46 Newark Branch, N.A.A.C.P. v. City of Bayonne, 134 F.3d 113 (3d Cir. 1998) . . . . . . . . . . . . . . . 28 Nolting v. Yellow Freight System, 799 F.2d 1192 (8th Cir. 1986)44 O'Neal v. City of New Albany, 203 F.3d 998 (7th Cir. 2002) . . 60 Phillips v. Collings, 256 F.3d 843 (8th Cir. 2001) . . . . . . 44 Red Deer v. Cherokee County, 183 F.R.D. 642 (N.D. Iowa 1999) . 59 Rivera v. NIBCO, 364 F.3d 1057 (9th Cir. 2004) . . . . . . . . 59 Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147 (2d Cir. 2001) . . . . . . . . . . . . . 46, 51 Cases page Russell v. Microdyne,, 65 F.3d 1229 (4th Cir. 1995). . 61, 63, 65 Sellers v. Mineta, 358 F.3d 1058 (8th Cir. 2004) . . . . . . . 59 Shattuck v. Kinetic Concepts, 49 F.3d 1106 (5th Cir. 1995) . . 61 Taylor v. Teletype Corp., 648 F.2d 1129 (8th Cir. 1981). . . . 46 Wallace v. Dunn Const., 62 F.3d 374 (11th Cir. 1995) . 59, 61, 62 Wehr v. Ryan's Family Steak Houses, 49 F.3d 1150 (6th Cir. 1995). . . . . . . . . . . . . 59, 61 Welch v. Liberty Mach. Works, 23 F.3d 1403 (8th Cir. 1994)2, 63, 64 Woods v. Perry, 375 F.3d 671 (8th Cir. 2004) . . . . . . . . . 41 Statutes 42 U.S.C. §§ 2000e et seq. . . . . . . . . . . . . . . . . . . .2 42 U.S.C. § 2000e(m) . . . . . . . . . . . . . . . . . . . . . 26 42 U.S.C. § 2000e-2(h) . . . . . . . . . . . . . . . . . . . . 28 42 U.S.C. § 2000e-2(k)(1)(A)(i). . . . . . . . . . . . . . 26, 27 42 U.S.C. § 2000e-2(k)(1)(A)(ii) . . . . . . . . . . . . . 27, 42 42 U.S.C. § 2000e-2(k)(1)(C) . . . . . . . . . . . . . 27, 42, 43 Regulations page 29 C.F.R. § 1607.1 (1978). . . . . . . . . . . . . . . . . . . 28 29 C.F.R. § 1607.1(C) (1978) . . . . . . . . . . . . . . . . . 29 29 C.F.R. § 1607.2 (1978). . . . . . . . . . . . . . . . . . . 28 29 C.F.R. §§ 1607.2(C) (1978). . . . . . . . . . . . . . . . . 30 29 C.F.R. § 1607.3 (1970) . . . . . . . . . . . . . . . . . . 29 29 C.F.R. § 1607.3(A) (1978) . . . . . . . . . . . . . . . . . 29 29 C.F.R. § 1607.3(B) (1978) . . . . . . . . . . . . . . . . . 44 29 C.F.R. § 1607.4(c) (1970) . . . . . . . . . . . . . . . . . 29 29 C.F.R. § 1607.5(A)-(C) (1978) . . . . . 29, 30, 31, 32, 35, 36 29 C.F.R. § 1607.5(H) (1978) . . . . . . . . . . . . . . . . . 32 29 C.F.R. § 1607.14(C)(4) (1978) . . . . . . . . . . . . . 31, 32 29 C.F.R. § 1607.16(Q) (1978). . . . . . . . . . . . . . . . . 30 Court Rules Fed.R.Civ.P. 50(a)(1). . . . . . . . . . . . . . . . . . . . . 44 Fed.R.Civ.P. 52(a) . . . . . . . . . . . . . . . . . . . . . . 25 Fed.R.Civ.P. 59. . . . . . . . . . . . . . . . . . . . . . . . 19 Miscellaneous 35 Fed. Reg. 12333 (August 1, 1970). . . . . . . . . . . . . . 29 STATEMENT OF JURISDICTION Appellee/Cross-Appellant Equal Employment Opportunity Commission concurs with Appellant Dial Corporations jurisdictional statement. STATEMENT OF THE ISSUES AND APPOSITE CASES 1. Whether the district court's finding that Dial failed to meet its burden to prove a business necessity for the Work Tolerance Screen is clearly erroneous. Apposite cases: Albemarle Paper Co. v. Moody, 422 U.S. 405 (1975) Lanning v. Southeastern Penn. Transp. Auth., 181 F.3d 478, 489 (3d Cir. 1999) 2. Whether the district court properly denied Dial's motion for judgment as a matter of law because the evidence was legally sufficient to support the jury's verdict that Dial intentionally discriminated against women beginning April 2001. Apposite cases: Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 335 (1977) 3. Whether the district court properly awarded backpay and benefits to claimants irrespective of whether they completed the WTS, without regard to average tenure, and without requiring proof that claimants incurred actual medical expenses. Apposite cases: Albemarle Paper Co. v. Moody, 422 U.S. 405 (1975) EEOC v. Rath Packing, 787 F.2d 318 (8th Cir. 1986) STATEMENT OF ISSUE IN EEOC'S CROSS-APPEAL 1. Whether the district court erred in denying backpay to one claimant because of her criminal record. Apposite cases: McKennon v. Nashville Banner Pub., 513 U.S. 352 (1995) Welch v. Liberty Mach. Works, 23 F.3d 1403 (8th Cir. 1994) STATEMENT OF THE CASE A. Course of Proceedings EEOC filed suit against Dial under Title VII of the Civil Rights Act of 1964 ("Title VII"), 42 U.S.C. §§ 2000e et seq., alleging Dial's use of the "Work Tolerance Screen" ("WTS"), a strength test Dial instituted in January 2000 to screen job applicants for a meat processing plant, had an unlawful disparate impact on women in violation of Title VII. EEOC also alleged that Dial was engaged in a pattern or practice of intentional discrimination against women. R.1 (Complaint ¶ 7).<1> After a six-day, bifurcated trial, the jury found intentional discrimination on August 20, 2004, and awarded compensatory damages on August 25, 2004, to several claimants. R.82, 84 (Verdict Forms) (Joint Appendix ("JA") 71, 73). On February 3, 2005, the district court denied Dial's motion for judgment as a matter of law ("JMOL"), R.106 (JA79), and ruled that the WTS had a disparate impact on women in violation of Title VII. R.107 (JA84). Dial ceased using the test, and the district court ordered Dial to offer jobs to the affected women and to remit backpay to 53 of the 54 women rejected under the test. R.118, 134, 137 (JA112, 120, 128). The district court entered final judgment on September 29, 2005. R.138, 139 (JA147, 155). On December 21, 2005, the district court stayed enforcement of the judgment pending appeal. R.157 (JA156). On appeal, Dial challenges three rulings of the district court: the denial of JMOL on the jury's finding of intentional discrimination; the determination that the WTS had an unlawful disparate impact; and three specific rulings on damages. EEOC's cross-appeal challenges the denial of backpay to one claimant. B. Statement of the Facts 1. Background Dial is an international company that produces canned meats at a plant in Fort Madison, Iowa. Tr.I 73.<2> Dial alleges that entry-level employees generally begin working in the "smokehouse" on the sausage-making team, loading or unloading 35 pound rods of sausage links. Employees carry the rods between a moving conveyor and a processing table and place the rods into notches from 33 to approximately 60 inches above floor level. See EEOC Exhibit ("Ex.") 11 (sausage- making team video). Prior to January 2000, Dial's hiring process included a written test, group problem-solving, and two interviews. Successful candidates were offered a job contingent on passing a medical exam and a drug test. Tr.I 74. In January 2000, ostensibly to reduce smokehouse injuries, Dial added the Work Tolerance Screen ("WTS"), a seven minute physical ability test administered after a contingent job offer. Id. Applicants were required to lift a bar with 35 pounds of weights off a table, carry it ten feet and place it on a wooden frame at a height of 35 inches, lift and carry it back to the table, immediately lift and carry it back to the frame and place it at a height of 67 inches, then lift and carry it back to the table again. Applicants were instructed to repeat this cycle "at their own pace" continuously for seven minutes. An occupational therapist ("OT") recorded the number of cycles completed and comments on the applicant's body mechanics and signs of fatigue. The number of women hired dropped immediately and substantially after Dial instituted the WTS. Virtually all male applicants (97% of the 162 men tested through September 2003) passed the WTS and were offered jobs. Most female applicants (81%) were able to complete the seven minute test, Joint Exhibit ("J") 1, but Dial nevertheless deemed many of these women to have "failed." As a result, the proportion of new hires who were women dropped from 46% during the three years before the WTS (1997-1999) to 14.2% during the three-and-a-half years after the WTS was instituted. Tr.I 74, Campion Tr.I 17-18. Paula Liles applied to Dial in January 2000 and was among the first applicants to take the WTS. Tr.I 79, 83. The observing OT noted Liles was able to lift to 65 inches, adding, "Lifting up over her head was difficult due to her height." See J-1, WTS result form #17 (3/9/00). Although Liles completed the seven minutes and the OT recorded her as having "passed," id., Dial notified Liles she would not be hired. Tr.I 84-88. Twenty-two additional women took the WTS in 2000, and Dial deemed thirteen (57%) to have passed. J-1 at 23. As with Liles, Dial determined several failed even though the OT marked them as having "passed." Tr.III 377. Dial thereafter eliminated the "pass/fail" notation from the WTS result form, substituting a rating of body mechanics ("good," "fair" or "poor") and indication whether instruction on body mechanics was required and "safety demonstrated with lifting." Tr.III 379; see generally J-1. Liles filed a discrimination charge with EEOC and the Iowa Civil Rights Commission in August 2000. J-8. Dial did not administer the WTS while the charge was pending. See J-1. In January 2001, Dial received the Iowa Commission's administrative closure of the charge, citing Liles' failure to submit a completed questionnaire. Tr.III 396-97, 410-12; Dial Ex. O (JA45-46). Dial next administered the WTS in February/March 2001. Fourteen women took the test, but only three (21%) passed. J-1 at 23. Shortly thereafter, EEOC notified Dial that it did not accept the Iowa Commission's dismissal and was conducting its own investigation. Tr.III 412-13; see Tr.III 454 (HR manager learned this in April 2001). Dial did not administer the WTS during EEOC's fifteen-month investigation.<3> On July 19, 2002, EEOC notified Dial that EEOC found cause to believe Dial had discriminated against Liles and other female applicants based on their sex in violation of Title VII. EEOC Ex.13. The notice also stated that EEOC believed Dial's use of the WTS had an adverse impact on women. Id. After failing to resolve the matter informally, EEOC filed this lawsuit on September 24, 2002. R.1. Dial next conducted WTS screenings in October/November 2002. Thirteen women took the WTS. At least eight completed the seven minutes, but Dial determined that only one woman, Stacy Meyer, actually passed.<4> Thus, in contrast to the 97% pass rate for men who took the test between January 2000 and the end of 2002, the pass rate for women dropped from 57% in 2000, to 21% in 2001, to 8% for the women who took the test in October/November 2002. Dial hired David Jones, Ph.D., to review the validity of the WTS, and he submitted his first report in December 2002. Jones/Campion Tr. 49, 65. The next year, sixteen women took the WTS and Dial determined eight (50%) passed; 98% of the men who took the WTS in 2003 passed. Dial Brf. at 18. 2. Evidence at Trial The case was tried in August 2004. During the five-day trial on liability, the jury heard testimony from ten women rejected based on the WTS. The jury also heard extensive testimony from EEOC's expert, Michael Campion, Ph.D., explaining that the WTS was appreciably more difficult than the actual job and that no reduction in smokehouse injuries could be attributed to the WTS. Dr. Campion has a doctorate in industrial and organizational psychology and has consulted and written extensively on preemployment testing. CampionTr.I 2-4; CampionTr.II 5-8 (describing physical ability tests he helped develop or validate), 52-53 (has studied jobs in "70 different companies, 250 projects, 240 articles on the topic"). Dr. Campion observed both the job and the WTS. He also reviewed information on the smokehouse job requirements, the incidence of injuries, various injury-reduction programs instituted by Dial, and relevant research literature. CampionTr.I 9-11. Dr. Campion explained that in the three years before Dial implemented the WTS (1997-1999), 45% of the workers Dial hired for the smokehouse were women, compared to 15% in the following three-and-a-half years. Id. at 16-17; EEOC Exhibit ("Ex.") 23. The WTS was the only change in Dial's hiring practices between these two periods. Tr.I 74. Dr. Campion noted the wide disparity in WTS "pass" rates for men and women, Campion Tr. 18 (citing EEOC Ex.24), and testified that this disparity was statistically significant: the likelihood that this was a mere random occurrence was one in a million, or "nearly ten standard deviations." Id. at 21-22. Dr. Campion explained that determining whether a preemployment test is a valid predictor of successful job performance generally involves two kinds of validity: content validity, "where the test itself looks like the job," and criterion- related validity, where you can show statistically that "people who do better on the test do better on the job . . . or, in the case here have fewer injuries." Id. at 7-8. Dr. Campion testified that after examining the WTS, he had significant concerns with both forms of validity. Id. at 22-57. The WTS lacked "content validity," according to Dr. Campion, because it was more difficult than the actual job. In the plant, where the sausage-making machinery dictates the pace, workers carry the rod in only one direction, followed by a long pause before the next lift-and-carry – approximately 1.25 lifts per minute. Id. at 23. Applicants taking the WTS, on the other hand, were required to carry the rod back and forth continuously, without pause, resulting in an average of six lifts per minute for women who "passed" the WTS. Id. As Dr. Campion explained, the fact that test- takers were told to go at their own pace was a "major concern" of the WTS, because when people know they are being tested they work faster, perhaps faster than they should, and then show fatigue and may fail the test for that reason, whereas they may have been able to work without fatigue at the actual pace of the job. CampionTr.II 28. Dr. Campion testified that because the WTS involved far more lifts per minute, the test "might be six times more difficult than the job." CampionTr.I 23. See also CampionTr.II 25 (structural differences between test apparatus and job mean test-taker must hold weighted bar up longer to secure it onto supporting lip; actual job is easier because a cup on each end holds the bar).<5> Dr. Campion testified that inconsistencies in the test's subjective scoring also seriously undermined its content-validity. Campion Tr.I 26-35; see id. at 25 ("[I]f you can't measure things in a reliable way, it can't possibly predict other things . . . ."). The WTS and the forms used to record applicants' WTS performance were designed by Molly Sichterman, an OT at the local hospital. Tr.III 326-30. The first year, an OT scored each applicant's WTS performance and submitted individual "result forms" to Martha Lutenegger, the Fort Madison plant nurse, Tr.III 298, who then decided who was hired. Tr.II 218, Tr.III 342-43. Starting in 2002, Lutenegger also observed and scored each applicant, but remained solely responsible for deciding whether an applicant who completed the WTS would be hired. Tr.III 366-67; Dial Brf. at 13. Dr. Campion reviewed the WTS result forms and concluded the results were very subjective. CampionTr.I 35. As an example, Dr. Campion compared the WTS result forms for nine men and six women and testified that the ratings and narrative comments on the forms showed that where male and female candidates appeared to have similar difficulties in performing the test, male candidates virtually always passed and female candidates were often deemed to have failed. Id. at 27-36. Dr. Campion noted that Dial hired several men in 2000 who were rated only "fair" or "fair to good" in terms of body mechanics and had to be instructed in safe lifting techniques. CampionTr.I 27-29; see J-1 at 11-12 (Massey, Harmon and Dietrich hired in July 2000). In contrast, five women tested in 2000 and 2001 completed the seven minute exercise, but Dial subsequently determined they had failed even though the comments on their result forms – relating to body mechanics and the need for instruction – were very similar to those for the above-noted men who passed. CampionTr.I 31-35; see J-1 at 10-11, 14-15. Indeed, the OT marked "pass" on the result forms for three of the women and noted "overall she did fairly well" for a fourth who took the test after the "pass/fail" notation was eliminated. CampionTr.I 32-34; J-1, result forms ## 14, 17, 20, 88. Dr. Campion testified that Dial revised the WTS score sheets several times, but the inconsistencies between the assessments of men and women remained. Six men tested in 2002 and 2003 all passed the WTS even though comments indicated difficulty either completing the test or maintaining good body mechanics. CampionTr.I 29-31 (Ackerson experienced "some increased strain and slight arching of back on higher level;" Grafton "became slightly fatigued;" Land "tended to slow down as test progressed with heavier breathing;" Smith "went up on tiptoes" but was able to self-correct; Chaknone "arched back but self-corrected" and "perspired last couple of minutes" and Ferguson "arched back but self corrected and stated his arms felt like rubber). Kimberly Walker, in contrast, was told she failed the WTS even though she, like several of these men, completed the seven minutes with some arching of her back and, like at least one of the men, remarked that she was becoming fatigued. Id. at 35.<6> Lastly, Dr. Campion rebutted Dail's claim that the WTS caused a decrease in smokehouse injuries. Dr. Campion examined reported injuries from before and after institution of the WTS, id. at 39-57, and testified that "[f]rom a research point of view . . . the evidence is not sufficient to support the conclusion that the work tolerance test has reduced injuries." Id. at 57. To gauge the impact of the WTS, Dr. Campion grouped injuries by the year the employee was hired (rather than the year the injury occurred) and then compared the number of injuries experienced during employees' first three years of employment. Comparing total recorded injuries, Dr. Campion noted that the 87 employees hired in 1997 experienced far more injuries in their first three years of employment than employees hired after that, both before and after the WTS. Id. at 45-47; EEOC Ex.30. Since the decrease in injuries began with the 1998 hirees, before implementation of the WTS, Dr. Campion testified that the decrease could not be attributed to the WTS. CampionTr.I 47. Dr. Campion further explained that Dial's data included injuries that were not very serious or were not related to employee strength. Since deciding whether to report non-serious injuries could be very subjective, a more reliable inquiry was to examine three categories of more serious injuries: strength-related injuries, those that resulted in workers' compensation, and those required to be reported by the U.S. Occupational and Safety Hazard Agency ("OSHA"). Id. at 42-43. Dr. Campion compared employee injuries, grouped by year of hire, for these three categories of injuries and found the data supported the same conclusion: the decrease in injuries started with employees hired in 1998 and was, therefore, unrelated to the WTS. Id. at 48-51 (citing EEOC Exs.31-35). Indeed, when 1997 employees were excluded, not only were injury rates relatively consistent before and after institution of the WTS, "but the absolute number of injuries is not very large." CampionTr.I 49. Finally, recognizing that workers spend differing amounts of time in the smokehouse following hire, Dr. Campion controlled for time actually spent in the smokehouse ("exposure rate") and concluded that apart from employees hired in 1997, whose injury rates were still significantly higher, the rate of injury per year of worker exposure to the smokehouse "is not any lower now than it was before [the WTS]." Id. at 51-56; EEOC Ex.36. Indeed, excluding workers hired in 1997, the highest injury rate per year of smokehouse exposure was among employees hired in 2001, after the WTS was put into effect. CampionTr.II 64-65; EEOC Ex.36. Dr. Campion agreed the smokehouse job is physically demanding and that where a job is physically demanding it is reasonable, although "not that common," to use a physical ability test. CampionTr.II 11, 13, 16. The issue, according to Dr. Campion, was whether a particular selection procedure is valid. Id. at 16-13 to 15. Dr. Campion noted that Dial previously required newly-hired employees to undergo a probationary period, which is another way of assessing whether new employees are capable of performing the job. Id. at 29. Dr. Campion testified that under this practice, "[w]omen successfully did the job . . . with no difficulty, and now they don't get a chance to even try." Id. Dr. Campion stated Dial had provided "1,000 pages of reports" of other steps Dial had taken to reduce injuries at the plant, including "ergonomic changes to the jobs, training programs, studies, guidelines, analyses, . . . [and] job rotation" which, Dr. Campion explained, is "commonly thought to be a helpful way of reducing the physical demands of jobs." Id. at 54.<7> Dr. Campion explained that "those things for sure had some effect, and it may be that they caused most of the [reduction in injury rates] . . . ." Id. at 54-55, 59-60. Dr. Campion never said the WTS was generally necessary, as Dial asserts. See Dial Brf. at 42, 45. To the contrary, Dr. Campion testified that in his opinion Dial had achieved a reduction in serious injuries before the WTS was implemented, through the various other measures Dial undertook beginning in 1996. Dial's two expert witnesses, Dr. Jones and Andrew Jackson, P.E.D., presented contrary testimony on several points.<8> Based on analyses prepared after EEOC filed suit, see Jones/Campion Tr. 65 (Dr. Jones' first report was December 2002), they stated, first, that the WTS closely resembled the smokehouse jobs and, therefore, served as a valid screen for those jobs. Tr.IV 522-26, 562-63; Jones/Campion Tr. 10- 12, 25-29. Dr. Campion had previously explained that job analyses "conducted on the eve of trial" are inherently problematic because the company's defensive posture makes these analyses "less objective." CampionTr.II 29, 31-32 (noting his publications on this point). Relying on essentially the same injury data as Dr. Campion, Dr. Jones also testified that smokehouse injuries dropped significantly after the WTS was instituted. Jones/CampionTr. 11, 31-45 (citing Dial Ex. U(11), U(12), U(18) (JA54-56)). Like Dr. Campion, Dr. Jones looked at worker's compensation injuries separately, explaining they "are normally considered more serious" than some of the other injuries on the list. Id. at 43. Unlike Dr. Campion, however, for most of his testimony, Dr. Jones combined all injuries occurring in 1997 through 1999 (the three years before Dial began using the WTS) and compared them to all injuries occurring after Dial adopted the WTS in January 2000. See, e.g., id. at 39-18 to 21 ("Workers' comp injury rates fell about 80 percent after implementation of the WTS. That's taking into account the new hires from 1997 through 1999 as the pretest group and 2000 through 2003 as the post-tested group."); see Dial Exs. U(11), (12) & (18) (JA54-56). Responding to Dr. Campion's more probing data analysis, Dr. Jones concluded with the bald assertion that the post-WTS injury decrease was significant even when 1997 hirees were excluded. Jones/CampionTr. 44-45. Dr. Jones did not support this conclusion with any data, see id., nor is this data laid out in any of Dial's charts. See JA47-56. On rebuttal, Dr. Campion disputed Dr. Jones' bald assertion of a 65% reduction of injuries post-WTS even excluding the accident-prone workers hired in 1997, stating, unequivocally, "That cannot possibly be true." Jones/CampionTr. 101. Dr. Campion pointed out that he, in contrast to Dr. Jones, had "shown the data broken down by year." Id.; see generally id. at 89-102; EEOC Ex. 36. Asked, on cross- examination, whether it was possible to draw any conclusions given the small number of strength-related, OSHA-recordable and workers' comp injuries, Dr. Campion responded that if Dial, comparing 160 employees hired during the three years before the WTS with the 160 employees hired in the following three-and-a-half years, "cannot demonstrate a drop in meaningful injuries[,] . . . [i]f in so many years of data they've been unable to show an effect, then maybe there's not an effect there." Id. at 99-100, 102 (emphasis added). Before beginning deliberation, the court instructed the jurors that they were the "sole judges of the facts" (instruction #1) and that expert testimony should be considered "just like any other testimony" (instruction #8). After hearing Dr. Campion testify that he had studied the data "in great detail" and that the data he summarized was "very close to what is reported in [Dial's] expert's report," CampionTr.II 40, the jurors were instructed they could accept or reject each expert's testimony and give it "as much weight as you think it deserves" considering, among other things, "the reasons given for the opinions and all the other evidence in the case." Id. Among the other evidence in the case, Lutenegger testified that she was aware, before implementing the WTS, that women generally have less upper body strength than men. Tr.III 391-2 to 8. She testified that after implementing the test, she realized it was screening out women at a higher rate than men, Tr.III 392-93, as did Farrington, the human resource official, and Hartlage, the plant manager. Tr.III 457, 465-67, 486-87. Lutenegger said she was not troubled by this drop in the number of women hired. Tr.III 367-68, 372-73. She testified that no one at Dial ever asked her to consider the test's fairness toward women, and she never did so on her own. Tr.III 391-93. Lutenegger told the jury she was the sole person to determine whether someone was hired or not based on the WTS, Tr.II 218; Tr.III 342-43, and she admitted that she failed a number of women whom the OT had marked "pass." Tr.III 377; see also Tr.II 270-71 (Lutenegger's comments on WTS result form conflicted with OT's). The jury heard testimony from ten women who completed the WTS but nevertheless "failed." Most had job experiences with comparable heavy lifting. See Tr.I 78; Tr.II 153-54, 173-74, 194-96, 200-01, 225-28, 251; Tr.III 281-82. Several testified that after they took the WTS, they were told they had passed, sent for a physical examination and given a written job offer with a start date (later rescinded). Many of these women disputed the accuracy of the negative comments and notations on their WTS results forms. See Tr.II 135-38, 157-58, 168-70, 175-77, 186-87, 200-03, 206-09, 231-32, 252, 257-58, 266-70; Tr.III 284-85. The jury had the opportunity to weigh Lutenegger's credibility when she testified that she implemented and continued to use the WTS solely to reduce injuries in the smokehouse, Tr.III 326, 361-62, in light of conflicting evidence concerning how the WTS was conceived and implemented. For instance, Lutenegger stated under oath that Joe Linares, the plant Safety and Security Manager, participated in the decision to implement the WTS. Tr.II 217. The jury heard Mr. Linares' sworn statement that he was not consulted or involved in the decision to implement the WTS. Although he was responsible for plant safety, Mr. Linares stated in his deposition (read into the record) that no one at Dial ever asked him to analyze how workers performed the job in connection with the design and implementation of the WTS, and he was never asked to look at injury rates for new hires after the WTS was implemented, to see if injuries decreased as a result. Tr.II 221-23. Dial did not call Mr. Linares as a witness to explain this discrepancy. After hearing this evidence, the jury concluded that Dial's continued use of the WTS intentionally discriminated against women beginning April 2001, R.82 (JA71), the month when Dial's managers would have realized that Lutenegger's hiring decisions had caused the WTS pass rate for women to drop from 57% in 2000 to only 21% in February/March of 2001 (compared to a consistent pass rate of approximately 95% for men who took the WTS in 2000 and 2001). Cf. Tr.III 466-67, 486-87 (HR official and plant manager recognized disparate impact on women in early 2001 but did nothing in response). The jury awarded compensatory damages of $5,000 each to six women who completed the WTS in 2002 and 2003 but were rejected by Dial. R.84 (JA73). The jury did not award any punitive damages. R.82 (JA72).<9> C. The District Court's Decisions 1. Denial of Dial's Motion for Judgment as a Matter of Law Dial moved for JMOL, arguing there was no legally sufficient basis for the jury to find for EEOC on the disparate treatment claim. R.95 (JA77-78).<10> The district court denied the motion. R.106 (JA79-81). Along with the Commission's statistical evidence, the district court observed that "EEOC produced ample additional evidence of overt acts on the part of Dial employees from which the jury could infer intentional discrimination" including the differences in the interpretation of the score sheets for men versus women and Dial's affirmative decision to continue using the WTS even after the disparate impact on women became apparent. Id. at 3 (JA81). 2. Dial's Liability for Disparate Impact Discrimination The district court also ruled Dial was liable under Title VII for the WTS's disparate impact on women for the entire period it was used. Order on Liability, R.107 (JA84-99). Noting that Dial had asserted a "business necessity" defense under alternative theories of content validity or criterion-related validity, the district court found that Dial failed to meet its burden under either theory. The district court found Dial failed to demonstrate the WTS was "content-valid" because the job itself is actually easier, in significant respects, than the WTS. Id. at 10-11. The district court also concluded that Dial failed to demonstrate the WTS is "criterion-valid" – i.e., significantly correlated with injury reduction – because "Dial could not show that the reduction in Smokehouse injuries following implementation of the WTS was due solely, or even primarily, to the test." Id. at 14-15. The district court further ruled that the relatively small number of non-trivial injuries occurring before implementation of the WTS could not "justify a test with the enormous [disparate] impact [on women] of WTS." Id. at 15. Dial thereafter stopped using the WTS. 3. Remedies The district court granted EEOC's request for job offers for the 54 women denied employment due to the WTS since March 9, 2000 (the date the first woman was denied employment under the test), and ordered backpay for 53 of them. The district court denied backpay to one claimant because of her criminal record. As a condition of employment, Dial required the claimants who accepted job offers to authorize Dial to conduct criminal background investigations, claiming it was part of Dial's long-standing hiring practices. Heather Wright-Bradley signed a criminal background authorization as requested and began working in the smokehouse on July 18, 2005. Dial suspended her on August 4th based on a criminal background report dated August 3, 2005. Dial thereafter disputed Wright-Bradley's entitlement to backpay, and the district court heard argument telephonically on September 6, 2005. See Transcript of Telephone Conference, 9/6/05 ("Tel.Con. Tr."). EEOC argued there was no evidence Dial routinely checked the criminal background of new employees when Wright-Bradley applied in May 2000. Tel.Con.Tr. 12; see, e.g., EEOC Ex.12 (Paula Liles employment application). During discovery, Dial had never indicated that its multi-step selection process included a criminal background check, and EEOC informed the district court that Dial never responded to our repeated requests for sample criminal background checks for other employees hired in May 2000. Id. at 4-5.<11> EEOC also advised the court that Dial told EEOC it had never previously rejected an applicant because of a criminal background check. Id. at 5, 11-12. Dial did not dispute this. Id. at 11. Instead, Dial alleged that going back to 1987 it had identified five Dial employees who were discharged when Dial learned they had been convicted of crimes they committed after they started working at Dial. Id. at 8-9, 11. Although Dial's counsel represented that the company has had a process of preemployment criminal background checks "in place for years," id. at 13, counsel also appeared to contradict this representation. For instance, after summarizing Wright-Bradley's criminal record, counsel stated that if Dial had initially hired Wright-Bradley back in 2000, the company "believes very strongly that it would have, within no time, gotten this information" (Wright-Bradley's criminal record) "and disqualified her from continued employment." Id. at 9 (emphasis added).<12> The policy is unwritten, and no Dial officials testified as to any specifics – including what kinds of criminal convictions would warrant non-hire or termination. Dial's counsel stated only: [I]t is clear that the company, when it gets information regarding these kinds of convictions, acts on them, and does not maintain in its workforce, for safety purposes, these employees who have these types of criminal records. Id. at 14 (emphasis added). Dial never explained whether a criminal record would bar an individual from employment permanently or just for a period of time. At one point during the hearing, however, Dial's counsel referred to Wright-Bradley's December 2004 conviction for theft and then stated: "For the year following that the company would not have hired her." Id. at 8 (emphasis added). Despite no actual evidence concerning either the existence or scope of Dial's purported unwritten policy, the district court accepted Dial's representation that if the company had hired Wright-Bradley when she applied in May 2000, it would have uncovered her criminal history at some point and would have terminated her. Order, 9/14/05, at 3 (JA130). Reasoning that the court's duty extended only to restoring a claimant to the position she would have held absent the discrimination, and that awarding Wright-Bradley back pay would place her in a better position, the district court denied EEOC's backpay claim for Wright-Bradley. Id. at 2-3 (JA129-30). SUMMARY OF ARGUMENT The district court, having reviewed the conflicting evidence and having made express credibility determinations, found that Dial failed to establish a "business necessity" for the WTS. The district court found Dial failed to prove content validity because the WTS was substantially more difficult than the actual job. The court found the WTS lacked criterion-related validity because Dial could not demonstrate that injuries decreased because of the test. This Court should affirm because the district court's findings are amply supported by the evidence and are not "clearly erroneous." The district court also properly denied Dial's motion for JMOL, because the jury's finding that Dial's continued use of the WTS after April 2001 constituted a "pattern or practice" of intentional discrimination is amply supported by the record. The jury was entitled to infer discriminatory intent from the collective evidence including the profound statistical disparity between the WTS pass rates of men and women, individual examples of differential treatment of men and women who took the WTS, and evidence that Dial's managers continued to use the test, after realizing it was screening out women, without even attempting to ascertain whether the perceived reduction in injuries was actually due to the WTS. The district court's award of damages fall well within the district court's broad discretion to award full relief to victims of employment discrimination consistent with Title VII's dual purposes of compensating claimants and deterring discrimination. I. THE DISTRICT COURT'S FINDING THAT DIAL DID NOT PROVE A BUSINESS NECESSITY FOR THE WTS IS NOT CLEARLY ERRONEOUS. A. Standard of Review EEOC tried its disparate impact claim to the district court. This Court reviews a district court's factual findings for clear error. Fed.R.Civ.P. 52(a) ("In all actions tried upon the facts without a jury . . . . [f]indings of fact . . . shall not be set aside unless clearly erroneous . . . ."). Where the district court's findings are based largely on conflicting witness testimony, they are accorded even greater deference, given the court's opportunity to judge the credibility of the witnesses. Id.; EEOC v. Joe's Stone Crab, 220 F.3d 1263, 1273 (11th Cir. 2000). "Business necessity" is a question of fact subject to review under the "clearly erroneous" standard. Davey v. City of Omaha, 107 F.3d 587, 593 & n.8 (8th Cir. 1997); see Ass'n of Mex.-Amer. Educators v. State of California, 231 F.3d 572, 585 (9th Cir. 2000) ("whether a test has been validated properly is primarily a factual question, which depends on underlying factual determinations regarding the content and reliability of the validation studies that a defendant utilized"); Melendez v. Ill. Bell Tel. Co., 79 F.3d 661, 669 (7th Cir.1996) (applying "clearly erroneous" standard); Bernard v. Gulf Oil, 890 F.2d 735, 743 (5th Cir.1989) (same); Hamer v. City of Atlanta, 872 F.2d 1521, 1526 (11th Cir.1989) (same). Any purely legal rulings of the district court are reviewed de novo. Davey, 107 F.3d at 591. B. The Parties's Evidentiary Burdens in a Disparate Impact Claim Title VII prohibits employment practices that are "‘fair in form' or facially neutral" but "‘discriminatory in operation.'" Griggs v. Duke Power Co., 401 U.S. 424, 431 (1971); Kraul v. Iowa Methodist Med. Ctr., 95 F.3d 674, 681 (8th Cir. 1996). If a plaintiff demonstrates that a facially neutral employment practice has a disparate impact – that it "falls more harshly on one protected class than another" – the burden then shifts to the employer to justify the challenged practice by proving it is "job related for the position in question and consistent with business necessity." 42 U.S.C. §§ 2000e-2(k)(1)(A)(i); 2000e(m) ("‘demonstrates' means meets the burdens of production and persuasion"); see Firefighter's Inst. for Racial Equality ("FIRE") v. City of St. Louis, 220 F.3d 898, 904 (8th Cir. 2000) ("FIRE III") (defendant must show employment practice is "related to safe and efficient job performance" and "consistent with business necessity"). An employer's burden to show a "substantial business justification" "is a heavy one." Bradley v. Pizzaco of Neb., 7 F.3d 795, 798 (8th Cir. 1993) (citing Hawkins v. Anheuser-Busch, 697 F.2d 810, 815 (8th Cir.1983); see Chambers v. Omaha Girls Club, 834 F.2d 697, 701 (8th Cir. 1987). It requires more than just showing the employment practice bears some relation to the job. Lanning v. Southeastern Penn. Transp. Auth., 181 F.3d 478, 489 (3d Cir. 1999) ("a standard focusing solely on whether the qualities measured by an entry level exam bear some relationship to the job in question would impermissibly write out the business necessity prong of the Act's chosen standard"). Where, as here, the challenged practice is an entry-level physical test, a defendant must demonstrate that the passing standard "reflects the minimum qualifications necessary to perform successfully the job in question." Id.; see Dothard v. Rawlinson, 433 U.S. 321, 332 (1977) ("a discriminatory employment practice must be shown to be necessary to safe and efficient job performance to survive a Title VII challenge.") (emphasis added). If a defendant fails to make this showing, the employment practice constitutes "unlawful . . . disparate impact." 42 U.S.C. § 2000e-2(k)(1)(A)(i). If a defendant meets this burden, the practice is still unlawful if the plaintiff shows the availability of an alternative practice with a less discriminatory impact that the respondent refuses to adopt. 42 U.S.C. §§ 2000e- 2(k)(1)(A)(ii), (C). C. The Evidence Supports the District Court's Determination Dial does not dispute the district court's well-supported conclusion that the WTS had a disparate impact on women. See Dial Brf. at 44. The district court found that in the three years before Dial instituted the WTS (1997-1999), Dial hired 162 persons to work at the Fort Madison plant, of whom 75 (46%) were women. See Order on Liability (R.107) at 4 (finding 18) (JA87). Between January 2000 and September 30, 2003, Dial tested 230 applicants (162 men and 68 women). Most of the men (157, or 97%) passed and were hired. In contrast, only 26 women (38% of the female applicants) passed and were hired, id. (finding 19), a disparity of "nearly ten standard deviations" between the hiring rates for men and women for the four-year period following implementation of the WTS. Id. (finding 20). These statistics are undisputed. See Dial Brf. at 16. The district court's conclusion that this amounts to a discriminatory effect under Title VII, Order on Liability at 8-9 (JA91-92), is not clearly erroneous, and Dial does not challenge it on appeal. Dial challenges the court's ruling that Dial failed to prove a business necessity for the WTS. Dial Brf. at 45-55. This challenge is without merit. The district court's factual findings on this point are amply supported by record evidence and, therefore, are not "clearly erroneous." See Newark Branch, N.A.A.C.P. v. City of Bayonne, 134 F.3d 113, 120 (3d Cir. 1998). Under Title VII, an employer may lawfully "give and . . . act upon the results of any professionally developed ability test provided that such test . . . is not . . . used to discriminate because of . . . sex . . . ." 42 U.S.C. § 2000e-2(h) (emphasis added). In Griggs, the Court explained that this provision requires that any tests used "measure the person for the job and not the person in the abstract." 401 U.S. at 436. In determining whether a particular selection procedure meets this standard, courts look to the Uniform Guidelines on Employee Selection Procedures ["Uniform Guidelines" or "UGESP"], jointly issued by EEOC, the Civil Service Commission, and the Departments of Labor and Justice. See 29 C.F.R. §§ 1607.1, 1607.2 (1978). "The message of these Guidelines is the same as that of Griggs – that discriminatory tests are impermissible unless shown, by professionally acceptable methods, to be ‘predictive of or significantly correlated with important elements of work behavior which comprise or are relevant to the job or jobs for which candidates are being evaluated.'" Albemarle Paper Co. v. Moody, 422 U.S. 405, 431 (1975).<13> Although the Uniform Guidelines do not have the force of law, this Court accords them "great deference." Allen v. Entergy Corp., 181 F.3d 902, 905 (8th Cir. 1999); cf. Albemarle, 422 U.S. at 430-31 (predecessor 1970 guidelines "entitled to great deference"); Griggs, 401 U.S. at 433-34 n.9 & 436 (predecessor 1966 and 1970 employment testing guidelines "comport with congressional intent"). Indeed, this Court has long looked to EEOC guidelines when assessing the validity of selection procedures that have a disparate impact on a protected group. See, e.g., FIRE v. St. Louis, 549 F.2d 506, 513 (8th Cir. 1977) ("FIRE I") (directing City to replace discriminatory promotional test with a new test to be validated in accordance with EEOC's published guidelines). The Uniform Guidelines apply to physical ability tests as well as paper and pencil tests. See 29 C.F.R. §§ 1607.2(C) (guidelines apply to "selection procedures"), 1607.16(Q) ("selection procedure" includes "full range of assessment techniques" including "performance tests" and "physical requirements"). Thus, there is no merit to Dial's unsupported contention that the Uniform Guidelines "have limited utility in the context of physical tests." Dial Brf. at 47. The district court was properly guided by the UGESP in its analysis of the proofs presented below. See Order on Liability at 4 (finding 21) & 9 n.4 (JA87, 92). The Uniform Guidelines describe several means for an employer to demonstrate a test's validity, two of which Dial attempted to demonstrate below: criterion-related validity and content validity. 29 C.F.R. § 1607.5(A).<14> The district court found it "undisputed" that "Dial did not attempt to establish the test's validity under the Uniform Guidelines prior to implementing the WTS in January 2000." Id. at 10 n.5 (JA93) (emphasis in original); id. at 4 (finding 23) (JA87). That, however, was not why the district court found Dial's validity studies insufficient, as Dial suggests. Dial Brf. at 49. Rather, the district court carefully considered the testimony of Drs. Jones and Jackson, explaining why they concluded the WTS had both content and criterion-related validity, in light of Dr. Campion's contrary testimony, and found Dial had failed to meet its burden of proof on either. Id. at 9-10 (JA92-93).<15> The district court's findings are not clearly erroneous. 1. The District Court's Finding that Dial Failed to Establish Content Validity Is Not Clearly Erroneous. Content validity exists when "the content of the selection procedure is representative of important aspects of performance on the job . . . ." 29 C.F.R. § 1607.5(B); Order on Liability at 4 (finding 22) (JA87). To be content-valid, a selection procedure that measures a skill or ability should "closely approximate an observable work behavior." 29 C.F.R. § 1607.14(C)(4). In other words, "a test that is significantly more difficult than the job is not content valid." Finding 24 (JA87). The district court acknowledged the opinions of Dial's experts that the WTS closely approximated the smokehouse jobs, but was persuaded by Dr. Campion's testimony that a crucial aspect of the WTS (the frequency of lifting) is significantly more difficult than the actual job. Id. at 10 (JA93). The district court found that the WTS "requires constant carrying and lifting," amounting to "at least 6 lifts per minute" for women who completed the test. Id. at 5 (findings 26, 28) (JA88). The loader job, in contrast, requires only 1.25 lifts per minute with breaks between each lift. Id. at 4-5 (findings 25, 27) (JA87-88). This difference was exacerbated by the fact that Dial instructed applicants to perform the WTS at their own pace, which likely encouraged applicants to proceed faster than they could sustain for a full seven minutes. Id. at 11 n.7.<16> The district court did not reject Dial's content validity argument because the WTS failed to "absolutely replicate" the jobs in question, as Dial suggests. Dial Brf. at 50-51. A selection procedure that is "representative of important aspects of performance on the job," 29 C.F.R. § 1607.5(B), need not test all or nearly all the skills required. The test cannot, however, be more difficult than the job itself. E.g. 29 C.F.R. §§ 1607.14(C)(4) ("If a test purports to sample a work behavior . . . the manner . . . of the selection procedure . . . should closely approximate the work situation."), 1607.5(H) (testing cutoff scores should be "consistent with normal expectations of acceptable proficiency within the work force"). In this instance, Dial required applicants to engage in more rigorous physical activity (continuous lifting and carrying) for a "representative" period of time (seven minutes) on the theory that this would permit Dial to project whether an applicant could perform the actual, slower-paced job for eight hours without injury. Dial's only evidence to justify its selection of seven minutes for this continuous carrying and lifting exercise was the bald, unsupported conclusions of its experts. See Dial Brf. at 22-23. Martha Lutenegger did not know why the test was seven minutes long, testifying that Molly Sichterman set up the protocols. Tr.II 218; Tr.III 334. Dial never presented Sichterman as a witness to explain why she chose seven minutes. Neither Dr. Jackson nor Dr. Jones explained why they believed seven minutes was the appropriate duration for the test. See Tr. 552-63; see Tr. 563 (Dr. Jackson asserted seven minutes gives "a good accurate snapshot of the capacity to do the work," but cited no basis for this opinion); Jones/CampionTr. 28 (Dr. Jones asserted only, "Seven minutes I think is a good sample."). The district court correctly noted that even if an individual might be able to perform a carrying/lifting cycle faster for a shorter period of time, that did not change the fact that the pace of the WTS was significantly faster than the actual job (thereby undermining the requisite similarity on which content validity is based). Order on Liability at 11 (JA94). Where a test creates such a drastic disparate impact on women, Dial's choice of seven minutes must be justified by more than the mere unsupported assertions of its experts. Cf. Lanning, 181 F.3d at 491-93 (conclusion of expert, unsupported by validating data, insufficient to sustain aerobic capacity standard of 1.5 mile run); Isabel v. City of Memphis, 404 F.3d 404, 413-14 (6th Cir. 2005) (test not content-valid where cutoff score never validated to ensure it measured minimal qualifications). Dial bore the burden of justifying the WTS. Given the evidence, the district court was free to, and properly did, reject Dial's unsupported claim that carrying and lifting 35 pounds constantly for seven minutes is comparable to lifting 35 pounds 1.25 times per minute for five to seven hours. JA94. The district court was also free to, and did, credit the testimony of Dial employee Michael Kruse, explaining how the job is easier than the WTS. Order on Liability at 10-11 & n.6 (JA93-94).<17> The district court further noted inconsistencies in WTS evaluations that worked to the detriment of female applicants, id. at 5 (findings 29-32) (JA88); see generally J-1 (underscoring the absence of a reliable and uniform basis for evaluating WTS performance). The district court's factual findings on all these points are amply supported by the record. See discussion at pp 8-11 & nn 5&6, supra. Thus, the testimony of Dial's experts was not "undisputed," and the district court did not rebut Dial's expert evidence sua sponte. Dial Brf. at 21, 48-49. This case is very different from FIRE III (Dial Brf. at 48), where this Court affirmed summary judgment for the City after the district court struck the plaintiffs' expert witnesses as a discovery sanction, leaving plaintiffs without any evidence to refute the City's expert testimony on content validity. 220 F.3d at 902, 904-05. Here, EEOC presented extensive expert testimony which the district court credited, and Dial is left simply to argue the district court should have credited its experts instead. It is the responsibility of the district court, as the factfinder on the disparate impact claim, to make credibility determinations between competing fact and opinion witnesses. Given the district court's express finding that EEOC's expert was more credible, and given that Dr. Campion's testimony was coherent, plausible and internally consistent, the district court's findings of fact on this point are not clearly erroneous. See Anderson v. Bessemer City, 470 U.S. 564, 575 (1985) (when trial judge's finding is based on decision to credit testimony of one of two or more witnesses, each of whom told a coherent and facially plausible story not contradicted by extrinsic evidence, "that finding, if not internally inconsistent, can virtually never be clear error"). The district court's finding that Dial failed to establish the content validity of the WTS is not clearly erroneous. 2. The District Court's Finding that Dial Failed to Establish Criterion- Related Validity Is Not Clearly Erroneous. Criterion-related validity is established through "empirical data demonstrating that the selection procedure is predictive of or significantly correlated with important elements of job performance." 29 C.F.R. § 1607.5(B); Order on Liability at 5-6 (finding 35) (JA88-89). Dial attempted to establish this by demonstrating that using the WTS to screen job applicants would correlate with a significant drop in smokehouse injuries. The district court found, however, that although the overall number of smokehouse injuries dropped after Dial implemented the WTS, id. at 6 (finding 37) (JA89), Dial could not show that this decrease was due solely or even primarily to the WTS. JA97-98. This finding is not clearly erroneous. The court found, first, that individuals hired in 1997 were far more likely to be injured than persons hired in any subsequent year. Id. (findings 41-44) (JA89-90). The court further found that excluding persons hired in 1997, the number of smokehouse injuries was virtually the same for persons hired before and after the WTS. Id. at 7 (comparing total injuries, strength-related and compensation-cost injuries, and controlling for exposure time (findings 45-49)) (JA90).<18> Thus, considering injuries by employee year of hire, the district court found there was a dramatic 80% reduction of strength-related workers' compensation injuries between persons hired in 1997 and those hired in 1998 (well before Dial began using the WTS). Id. (finding 50). Comparing employees hired each year from 1998 through 2003, on the other hand, the court found virtually no change in injury rates before and after institution of the WTS.<19> Furthermore, given the high rate at which the WTS screened out female job applicants and the fact that men and women were hired at approximately equal rates before the WTS, if the WTS was effective at identifying injury-prone employees, one would expect that before the WTS was implemented, women must have had a higher rate of smokehouse injuries than men. The district court found, however, no appreciable difference between the injury rates of men and women before the WTS. Indeed, taking into account exposure time, the court found that women incurred injuries at a lower rate than men for two of the three years before Dial began using the WTS. See id. at 8 (findings 51-52), 14 (JA91, 97). The district court acknowledged that Dial's expert analyzed the injury data differently and reached contrary conclusions. See id. at 12-13 (JA95-96). Validation studies done in anticipation of litigation, as Dial's were here, "must be examined with great care due to the danger of lack of objectivity." See Lanning, 181 F.3d at 492 n.21. EEOC's expert rebutted Dial's evidence by analyzing injuries according to the year in which the employee was hired, an approach that makes eminent sense when considering the effects of a new hiring standard. The district court was more convinced by this analysis. Order on Liability at 14 (JA97). Where experts present differing opinions at trial, the factfinder is free to exercise its discretion in deciding whose view to credit. Anderson, 470 U.S. at 575. The district court also found, and Dial admits, that the company took various other steps to reduce injury rates in the smokehouse, beginning in late 1996. Finding 40 (JA89). Dial implemented an ergonomic job rotation program in 1996, added a team approach to increase job rotation in April 2000, lowered the height of the bar on the Frank-O-Matic machines, and conducted periodic ergonomic and safety audits, including an annual "back injury prevention program" beginning in 2000. Id.; see EEOC Ex.14 (Dial's answer to interrogatory 11). The district court found there was no way to factor out the beneficial effects of these additional measures. Order on Liability at 14-15 (JA97-98). It was Dial's burden to demonstrate a correlation between the WTS and a reduction in injuries. Dial did not meet that burden here, having conceded below that "all of these measures have contributed to the reduction of work related injuries at the plant." EEOC Ex.14 at 8. Thus, based on all of these findings, the district court properly concluded that "no matter how the numbers are ‘massaged,' Dial could not show that the [overall] reduction in Smokehouse injuries following implementation of the WTS was due solely, or even primarily, to the [WTS]." Order on Liability at 14-15 (JA97-98). The court's decision comports fully with the Supreme Court's analysis in Dothard, 433 U.S. 321 (1977). Contrary to Dial's assertions, Dial Brf. at 43, Dial committed the same mistake in implementing the WTS as did the Alabama Corrections Board in adopting minimum height and weight requirements purportedly to ensure prison guards were strong enough to supervise correctional inmates. 433 U.S. at 323-24, 331. The Court invalidated the height and weight requirements because the Corrections Board could not show they were actually correlated with the necessary strength for good job performance, and suggested the Board could achieve its asserted purpose by adopting a test that measured strength directly. Id. at 331-32. Although the WTS tests upper-body strength, Dial adopted it purportedly to reduce injuries, not to ascertain if applicants could adequately perform the job.<20> Dial, however, never established that smokehouse injuries were due to lack of upper-body strength (as opposed to, say, carelessness or lack of coordination). Dial also never assessed whether persons hired after January 2000 incurred fewer injuries than persons hired before the WTS was implemented. See, e.g., Tr.III 485 (plant manager did not recall having those types of discussions, but agreed it was "possibly" relevant); Tr.III 491 (plant manager did not know whether persons hired in 2002 were more or less likely to be injured than those hired in 1998); Order on Liability at 11 (JA94). Dr. Campion's testimony (which the district court credited) showed that persons hired with the WTS did not have a lower injury rate, and the district court's factual findings on that point are not clearly erroneous. Dial, therefore, committed the same error as the Alabama Correctional Board in Dothard: Dial adopted a hiring procedure (the WTS) that measures one attribute (upper-body strength) as a proxy for something else (injury avoidance), without ever ascertaining whether the two are causally related. Dial argues that it was motivated by safety concerns and that the district court's ruling essentially "punishes" it for "taking too many steps to protect employee safety" and will prevent employers like Dial from implementing more stringent requirements to make the workplace safer. Dial Brf. at 42, 50, 53. Of course, Dial's motivation is not at issue in a disparate impact claim. Kirby v. Colony Furniture, 613 F.2d 696, 703 (8th Cir. 1980); cf. FIRE v. St. Louis, 616 F.2d 350, 357 n.10 (8th Cir. 1980) (defendant's good intentions not determinative of whether exam is content-valid). Employers like Dial are free to adopt more stringent requirements to make the workplace safer, as long as the measures do not disproportionately exclude members of a protected group or are properly justified by "business necessity" under Title VII. Here, the district court found no business justification because when all the evidence is considered, Dial failed to establish any correlation, let alone a significant correlation, between the WTS and injury reduction. This finding – based on the record as a whole and, in particular, on EEOC's unrefuted expert analysis of Dial's injury statistics and the district court's express determination that EEOC's expert was more credible than Dial's – is not clearly erroneous. The district court properly found that Dial failed to demonstrate "criterion-related validity." Dial argues that it is entitled to meet its burden by demonstrating, in the alternative, that the WTS is "related to safe and efficient job performance" and "consistent with business necessity." Dial Brf. at 54-55. This, however, just restates the business necessity standard. See FIRE III, 220 F.3d at 904. Moreover, although protecting employees from workplace hazards is undeniably a legitimate business goal, here, given the WTS's disparate impact on women, Dial must demonstrate, not just allege, that the WTS actually serves this important purpose. The Snook data, based on nationwide assessments of men and women in the workforce, Dial Brf. at 55, do not satisfy Dial's burden. This generalized data does not necessarily reflect the capabilities of the individual women who applied to work at Dial, many of whom had years of work experience in jobs with similar strength requirements. See p.20, supra. In any event, Dial did not argue this below and, therefore, it is waived on appeal. Woods v. Perry, 375 F.3d 671, 674 n.2 (8th Cir. 2004). 3. The District Court Did Not Improperly Shift the Burdens of Proof. In its conclusion, the district court set forth an additional, wholly alternative ground for finding Dial did not establish business necessity: Dial failed to show a "compelling" need for the WTS and the absence of less discriminatory mechanisms. JA98 (citing Bradley, 7 F.3d at 797).<21> Dial argues, from this, that the district court improperly required it to disprove less discriminatory alternatives. Dial Brf. at 56. This argument is unavailing, for several reasons. First, the district court found that Dial failed to demonstrate "business necessity" based on the two grounds Dial asserted below – criterion-related and content validity. Since this finding is not clearly erroneous, there was no need for the district court to reach, and no occasion for this Court to consider, the question of less discriminatory alternatives under 42 U.S.C. § 2000e-2(k)(1)(A)(ii), (C). The statutory burden never shifted to EEOC, so the district court's statement that Dial failed to establish the absence of alternatives is moot. See Donnell v. General Motors, 576 F.2d 1292, 1299 n.15 (8th Cir. 1978) (declining to reach question of alternatives where defendants did not prove business necessity). Second, even if this Court were to reach the question of less discriminatory alternatives, the record contains ample evidence of alternative injury-prevention measures that Dial identified and put into place either before or around the same time as the WTS. See discussion at pp 13-14 & n.7, supra. These measures have no disparate impact on women and (as Dial recognized, see EEOC Ex. 14 at 7-8) they collectively reduced smokehouse injuries. Where, as here, sufficient evidence exists in the record to support the district court's finding, the question of who has the burden of proof is immaterial.<22> In any event, the district court's statement accurately reflects circuit precedent. E.g., Hawkins, 697 F.2d at 815; Kirby, 613 F.2d at 705 n.6; see 42 U.S.C. § 2000e- 2(k)(1)(C) (expressly endorsing law "as it existed on June 4, 1989, with respect to the concept of ‘alternative employment practice.'"). Moreover, it is logical for this Court to require an employer, as part of establishing "business necessity" to show it truly needed the practice in question, and an obvious element of "need" is the absence of readily apparent, less discriminatory ways of accomplishing the same goal. See EEOC v. Rath Packing, 787 F.2d 318, 328 (8th Cir. 1986) (employment practice "must not only foster safety and efficiency, but must be essential to that goal"; practice is not justified on basis of business necessity "if there exists a ‘nondiscriminatory alternative means of determining qualification'") (citation omitted; emphasis in original); cf. Nolting v. Yellow Freight System, 799 F.2d 1192, 1198-99 (8th Cir. 1986) (applying similar standard under ADEA); 29 C.F.R. § 1607.3(B) ("whenever a validity study is called for by these guidelines, the user should include . . . an investigation of suitable alternative selection procedures . . . which have as little adverse impact as possible . . . ."). Bradley thus reflects the law of this Circuit, and the district court's reference to it, although not necessary to the outcome, was proper. II. THE DISTRICT COURT PROPERLY DENIED DIAL'S MOTION FOR JMOL BECAUSE THE EVIDENCE SUFFICED FOR THE JURY TO FIND DIAL ENGAGED IN A PATTERN OR PRACTICE OF INTENTIONAL DISCRIMINATION AGAINST WOMEN BEGINNING APRIL 2001. A. Standard of Review This Court reviews a district court's denial of JMOL de novo, using the same standards as the trial court. Phillips v. Collings, 256 F.3d 843, 847 (8th Cir. 2001). Judgment as a matter of law may be granted only when "there is no legally sufficient evidentiary basis for a reasonable jury to find for [the non-moving] party . . .." Id. (citing Fed.R.Civ.P. 50(a)(1)). The court must draw all reasonable inferences in favor of the nonmoving party without making credibility assessments or weighing the evidence. Id. B. The Record Contains Sufficient Evidence to Support the Jury's Verdict. The trial evidence is legally sufficient to support the jury's verdict that Dial's continued use of the Work Tolerance Screen constituted a pattern or practice of intentional discrimination against female job applicants beginning April 2001. The jury was properly instructed that to rule in EEOC's favor the jury had to find that Dial implemented and/or continued the WTS "with the intent to discriminate against women on the basis of sex;" that EEOC could prove its "pattern or practice" claim through "statistical evidence of unfavorable treatment of women together with individual instances of discrimination against women;" and that EEOC did not need to prove Dial's officials harbored some "special animus or malice towards women," but only that Dial acted (through its officials) "consciously and intentionally based on sexual stereotypes." Jury Instruction Nos. 12, 14 (emphasis added). These instructions properly state the law, and Dial does not challenge them on appeal. Dial argues that the jury's finding of intentional discrimination is unsupportable because Dial's managers were motivated to reduce injuries, not to keep women out of the smokehouse, when they implemented the WTS. See generally Dial Brf. at 32-39. Based on the record, however, the jury could reasonably infer intentional discrimination from the profound statistical disparity between the treatment of men and women under the WTS and evidence that Dial's managers acted based on a stereotypical belief that women are less able to perform this strenuous job injury-free. JMOL was properly denied. In a pattern-or-practice case, the plaintiff must prove "the defendant ‘regularly and purposefully' treated members of the protected group less favorably and that unlawful discrimination was the employer's ‘regular procedure or policy.'" Morgan v. United Parcel Serv. of Amer., 380 F.3d 459, 463 (8th Cir. 2004) (quoting EEOC v. McDonnell Douglas Corp., 191 F.3d 948, 951 (8th Cir.1999), and Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 335, 360 (1977)). Statistical evidence, accompanied by anecdotal evidence of discrimination, is often used to establish a pattern or practice of intentional discrimination. Teamsters, 431 U.S. at 335 n.15; Morgan, 380 F.3d at 463-64; Joe's Stone Crab, 220 F.3d at 1287; see Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147, 168 (2d Cir. 2001) (liability phase focused on class-wide statistical evidence; specific instances of alleged discrimination provide "texture" to the statistics). Once the plaintiff establishes a discriminatory pattern or practice, "the burden shifts to the employer to prove that [] individual employment decision[s were] free from that discrimination." Taylor v. Teletype Corp., 648 F.2d 1129, 1136 (8th Cir. 1981). Here, EEOC presented strong statistical evidence of a profound disparity in Dial's hiring of women following adoption of the WTS. Before the WTS, 46% of Dial's new hires were women; after the WTS, that proportion dropped to 14%. The differences in WTS pass rates for men and women were stark the first year (95% for men versus 57% for women) and widened further by the end of the third year. See pp 5-6, supra (pass rate for women dropped to 21% in 2001 and 8% in 2002, compared to steady 95% for men). Statistical disparities are significant, for purposes of proving a pattern or practice of intentional discrimination, "if the difference between the expected value and the observed number is greater than two or three standard deviations." Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 308 n.14 (1977); Craik v. Minn. State Univ., 731 F.2d 465, 475 n.13 (8th Cir. 1984). The disparity between the hiring rate of men versus women under the WTS from 2000 through 2003 "was nearly ten standard deviations." See Finding 20 (JA87). Significantly, 81% of the women who took the WTS completed the seven minute exercise, but Lutenegger decided only 39% passed. J-1. EEOC also presented examples where Dial's managers gave the benefit of the doubt to men struggling to complete the WTS, but not to women who showed similar signs of fatigue or poor body mechanics. See discussion at 11-14 & n.8, supra. As in Teamsters, these individual experiences "brought the cold numbers convincingly to life." See 431 U.S. at 339. Discriminatory intent in a pattern or practice case can sometimes be inferred from the mere fact of differences in treatment. Teamsters, 431 U.S. at 335 n.15. Moreover, a Title VII plaintiff "need not prove that a defendant harbored some special ‘animus' or ‘malice' towards the protected group to which she belongs" because "Title VII prohibits ‘the entire spectrum of disparate treatment of men and women resulting from sex stereotypes,' . . . even where the stereotypes are benign or not grounded in group animus." Joe's Stone Crab, 220 F.3d at 1283-84. It is immaterial that the WTS was conceived of and implemented by women and that the WTS results forms do not contain any expressly discriminatory statements, as Dial repeatedly notes. See, e.g., Dial Brf. at 40. Rather, on this record, the jury could reasonably conclude that Dial's managers were motivated by unlawful stereotypes concerning the abilities of women in the workplace when Dial continued using the WTS after April 2001. For instance, EEOC presented the admissions of various Dial managers that they were aware of, but unconcerned by, the marked reduction in the hiring of women after the WTS began. The jury could reasonably infer from this that Dial's managers assumed the WTS would inevitably screen out women and, therefore, never bothered to take rudimentary steps to see if the WTS, as opposed to something else, was actually preventing injuries. See Tr.II 218-20 (Lutenegger never checked to see if WTS actually reduced injuries; she was "comfortable" continuing the test despite its impact on women); Tr.III 466-67, 486-87 (HR official and plant manager recognized disparate impact on women in early 2001 but did nothing in response). Among other things, Dial's managers never took the logical step of ascertaining whether persons hired with the WTS incurred fewer injuries than those hired before the WTS was instituted. See, e.g., Tr.II 221-23 (Dial never asked Safety Manager to look at injury rates to see if they decreased after WTS). Likewise, even after Dial's managers realized the WTS was screening women out of the smokehouse job, they never inquired whether women had been injured more frequently than men before the WTS – a logical means of verifying whether the WTS was a proper hiring tool. See, e.g., Jones/CampionTr. 81 (Dial never asked Dr. Jones to compare injury rates of men and women before WTS). The jury was entitled to infer, from the absence of any of these logical inquiries, that Dial was acting based on a stereotypical misperception that women could not perform the smokehouse jobs without injury. This evidence, viewed collectively, is sufficient to sustain the jury's verdict of intentional discrimination beginning April 2001, see Craik, 731 F.2d at 471, and Dial's arguments, see Dial Brf. at 30-41, are unavailing. EEOC was entitled to rely on the same statistical evidence used to prove its disparate impact claim (Dial Brf. at 32-34). See Craik, 731 F.2d at 472 ("evidence relating to one claim may be relevant to and persuasive with regard to others"); see also Teamsters, 431 U.S. at 339 n.20 (statistical imbalance probative in Title VII pattern or practice case "because such imbalance is often a telltale sign of purposeful discrimination" and in many cases is "the only available avenue of proof . . . to uncover clandestine and covert discrimination by [an] employer"); id. at 335 n.15 ("Either theory [disparate impact or disparate treatment] may, of course, be applied to a particular set of facts."). Moreover, although the EEOC did not rely, in its "pattern or practice" claim, on comparative evidence of "similarly situated" persons – an analytical approach more commonly used in individual cases under McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973) – men and women are "similarly-situated" for purposes of examining whether Dial's use of the WTS was discriminatory. Male and female job applicants all received a conditional offer of employment and, therefore, were deemed by Dial to be eligible to take the WTS. The issue here was whether the implementation of the WTS was discriminatory. Any purported "physiological differences between the sexes," Dial Brf. at 35, are irrelevant to this inquiry. Indeed, in making this argument Dial persists in the very same stereotypic approach being challenged in this lawsuit – that it intentionally continued to use a screening device which requires more upper body strength than does the actual job. Notably, Dial never asserted it adopted the WTS because it believed female production workers had a greater history of, or propensity for, injury than male workers. Dr. Jones baldly asserted that prior to the WTS, women had higher injury rates than men, but he did not support this statement with any underlying data and said Dial had never asked him about it. Jones/CampionTr. 62-63. Dr. Campion, on the other hand, testified this was not true, backing this conclusion with specific data. Jones/CampionTr. 91-94. The jury certainly could have credited Dr. Campion's testimony on this point, as the district court did in its disparate impact order. JA97. EEOC did not rely on Dial's "mere knowledge" of the disparity in pass rates, as Dial wrongly suggests. Dial Brf. at 35-36. EEOC also relied on the fact that men who apparently struggled to complete the WTS (based on comments on the result forms) generally passed while women with similar comments on their WTS result forms failed. Dial argues that these numerous examples are merely "isolated remarks" that do not prove a pattern or practice of discrimination. Dial Brf. at 37. The pattern, however, is shown by the statistics, and the specific examples simply provide "texture" and "[bring] the cold numbers convincingly to life." See Teamsters, 431 U.S. at 339; Robinson, 267 F.3d at 168. That the result forms for most of the men had no comments does not mean the jury was not entitled to consider those that did. The jury had all of the forms during its deliberations, see J-1, and it was entitled to weigh the differences in treatment between men and women who took the WTS.<23> Finally, Dial argues that its managers' "honest belief" that injuries were decreasing as a result of the WTS (even though they admittedly did nothing to verify that assumption) is sufficient to rebut EEOC's prima facie evidence of a pattern or practice of discrimination. Dial Brf. at 39-41. The cases Dial cites for this argument do not address pattern or practice claims. See Kratzer v. Rockwell Collins, 398 F.3d 1040 (8th Cir. 2005); Johnson v. AT&T, 422 F.3d 756 (8th Cir. 2005). They are, therefore, inapposite. In any event, the jury apparently agreed that Dial's managers acted out of such an "honest belief" when they first implemented the WTS, as the jury did not find intentional discrimination as of January 2000. Given the mounting evidence that women were being excluded from a job they previously performed without problem, the jury apparently concluded that as of April 2001 Dial's managers could no longer reasonably believe the WTS was an unbiased measure of an applicant's ability to perform the smokehouse job without injury. The evidence is sufficient to support such a jury conclusion, and JMOL was properly denied. III. THE DISTRICT COURT DID NOT ABUSE ITS DISCRETION IN AWARDING BACKPAY AND BENEFITS TO CLAIMANTS REGARDLESS OF WHETHER THEY COMPLETED THE WTS, WITHOUT REGARD TO THE AVERAGE LENGTH OF TENURE, AND WITHOUT REQUIRING PROOF OF ACTUAL MEDICAL EXPENSES. A. Standard of Review "[A] district court has broad equitable discretion to fashion back pay awards in order to make the Title VII victim whole." E.E.O.C. v. Delight Wholesale, 973 F.2d 664, 669-70 (8th Cir. 1992) (citing Franks v. Bowman Transp., 424 U.S. 747, 763-64 (1976); King v. Staley, 849 F.2d 1143, 1144 (8th Cir.1988)). This Court, therefore, reviews such awards for abuse of discretion. B. The District Court Did Not Abuse Its Discretion In Awarding Backpay and Benefits to 53 Claimants. The district court's determination of remedies must "be measured against the purposes which inform Title VII." Albemarle, 422 U.S. at 417. Backpay, in addition to making victims of discrimination whole, furthers the Act's purposes of preventing workplace discrimination by making an employer's failure to eliminate discrimination a costly choice. See id. at 417-18 (citations omitted). Consequently, the district court is obligated to exercise its sound discretion to grant a prevailing Title VII claimant "the most complete relief possible." Rath Packing, 787 F.2d at 329. In this case, once the district court found that EEOC had proven its disparate impact claim against Dial, the district court was presumptively obligated to award backpay and other make-whole relief to each of the women denied employment because of the WTS. See Delight Wholesale, 973 F.3d at 669-70 (citations omitted). With the exception of one claimant who was denied backpay based on her criminal record, see infra, the district court's backpay award is well within the court's discretion and should be affirmed. First, the district court properly awarded backpay irrespective of whether a claimant actually completed the WTS. Backpay was awarded as relief to EEOC's disparate impact claim. See JA112-14. EEOC argued, and the district court found, that the unlawful disparate impact arose from Dial's use of the WTS itself, not just the manner in which the results were interpreted. JA113-15. Thus, the women who were adversely affected by this unlawful employment practice are those who were required to take the WTS and then denied employment, irrespective of whether they completed the full seven minutes. The district court properly awarded these women backpay to compensate for this discrimination. JA114-15. Dial's contrary argument (Dial Brf. at 57-59) would have merit only if the district court had found that the WTS itself was valid and the disparate impact was caused solely by the subjective scoring system, or if Dial had proven at some point that only persons who can perform the WTS for seven minutes are actually able to perform the job. Neither is the case, however. The subjective scoring system played a role in the district court's disparate impact decision, JA114, but was not the sole or even primary basis for the court's ruling. See JA115 ("Eliminating the subjective component of the test certainly improves the test's viability, but does not eliminate all of the problems cited in this Court's Order."); see also JA92-94 (noting the statistical disparity for women between completing and passing the WTS, but highlighting the various ways in which the WTS was more difficult than the actual job). Dial never attempted to establish that only persons who could complete the WTS for seven minutes were able to perform the job. Cf. JA94 (even in its "post-hoc attempts to validate the test," Dial never tried to determine whether employees hired after passing the test performed better than those who never took the test). Thus, all of the women who took the WTS and were not hired on that basis were "actual victims of discrimination," Dial Brf. at 58, and properly received backpay. Likewise, the district court acted well within its discretion in awarding backpay without regard to tenure data. Without even explaining what "empirical tenure data" Dial is referencing, Dial wrongly argues that "[c]ourts typically use such data to determine what a claimant would have earned over time had he or she actually been hired." Dial Brf. at 60. To the contrary, as the district court properly noted, the "general rule" for back pay is what the claimant would have earned absent the discrimination, minus whatever wages the claimant earned during the relevant period of time. See JA124 (citing Hartley v. Dillard's, Inc., 310 F.3d 1054, 1062 (8th Cir. 2002)). In a case like this, the appropriate time period runs from the date of the discrimination to an unconditional job offer. See JA124 (citing Fiedler v. Indianhead Truck Line, Inc., 670 F.2d 806, 808 (8th Cir. 1982)). There is no principled basis for deviating from this general rule here. The cases cited by Dial, Dial Brf. at 60, require no different result, as the district court explained. See JA124-125 & n.3. Moreover, although Dial may have experienced an unusually high turnover rate after implementating the WTS, see Campion Tr.I 36-38, this did not account for the majority of new employees. According to Dial, 60% of workers hired between January 1, 2000, and October 18, 2004, were still employed at Dial in June 2005. See JA125. There is no proper ground for assuming the claimants in this lawsuit would not have been among these workers. Thus, there is no proper legal or factual basis for Dial's argument that the district court should have deprived claimants of full backpay relief, contrary to the intent of Title VII. See Albemarle, 422 U.S. at 418-21. This Court should not disturb the district court's sound exercise of discretion to make these victims of discrimination whole by ordering backpay through March 31, 2005, the agreed-upon backpay cut-off date. See JA124-26, 151. Finally, the district court also acted well within its discretion when it ordered Dial to compensate claimants for lost medical benefits based on the amount Dial would have paid on their behalf as premiums, without regard to whether a particular claimant incurred out-of-pocket medical expenses. JA121-24. Requiring claimants to prove they incurred medical expenses, as Dial argues, Dial Brf. at 60-61, would undermine the purposes of Title VII's remedy provisions by essentially giving Dial a windfall for its discriminatory actions for any claimants who lack such proof. As this Court has recognized, "[w]hen EEOC acts, albeit [often] at the behest of and for the benefit of specific individuals, it acts also to vindicate the public interest in preventing employment discrimination." Rath Packing, 787 F.2d at 325. Allowing Dial to avoid paying any damages for lost medical benefits where a claimant did not actually incur any out-of-pocket expenses would permit Dial to benefit from its own wrongdoing. Such a result does not serve to "prevent discrimination," because to the extent discrimination is cost-effective for an employer, the prophylactic purpose of the statute is thwarted. The district court's well-reasoned explanation of why it exercised its discretion in this fashion, JA121-24, amply demonstrates that this award properly furthers the purposes of Title VII. CROSS-APPEAL SUMMARY OF ARGUMENT The district court wrongly denied backpay to one claimant based on her criminal record. After the district court ordered Dial to offer jobs to the 54 claimants, Dial asked those who accepted the offer (but none of the other claimants who only sought backpay) to authorize a criminal background check. Heather Wright-Bradley accepted a job, authorized a criminal background check, and began working in the smokehouse. When Dial received her criminal record two weeks later, it terminated her. The district court's denial of backpay on this ground is legal error (and, a fortiori, likewise an abuse of discretion) because it conflicts with the after-acquired evidence doctrine set forth by the Supreme Court in McKennon v. Nashville Banner Pub., 513 U.S. 352 (1995). Alternatively, even if a court might properly deny a Title VII discrimination victim backpay based on a bona fide employer policy, Dial offered no proof here that the policy actually existed when Wright-Bradley applied to Dial in May 2000, or that the terms of the policy were such that Wright-Bradley would have been terminated based on her criminal record at that time. There is no record evidence that in May 2000 Dial required job applicants to submit to a criminal background check as part of the hiring process and, even if it did, there is no evidence of the nature of this unwritten alleged policy. ARGUMENT THE DISTRICT COURT ERRED AS A MATTER OF LAW IN DENYING BACKPAY TO ONE CLASS MEMBER BASED ON AFTER-ACQUIRED EVIDENCE OF HER CRIMINAL RECORD. A. Standard of Review The district court denied backpay to Wright-Bradley based on an erroneous legal ruling. This Court reviews legal rulings de novo. See Mems v. City of St. Paul, 327 F.3d 771, 784 (8th Cir. 2003). B. The District Court Erroneously Denied Backpay Based on the Claimant's Criminal Record. 1. The District Court's Decision Conflicts with McKennon. In McKennon, 513 U.S. 352, the Supreme Court held that an employer found liable for age discrimination cannot completely avoid payment of monetary damages to a victim of discrimination based on after-acquired evidence of the employee's on- the-job misconduct, even if the misconduct would have resulted in discharge had the employer known of it at the time. Id. at 356. The Court reasoned that the imposition of backpay, notwithstanding later-discovered employee misconduct, ensures "proper recognition [of] the fact that [a statutory violation] has occurred which must be deterred and compensated . . . ." Id. at 362. As the Court explained: "An absolute rule barring any recovery of backpay . . . would undermine the [statute's] objective of forcing employers to consider and examine their motivations, and of penalizing them for employment decisions that spring from age discrimination." Id. In calculating such an award, the Court indicated "[t]he beginning point in the trial court's formulation of a remedy should be calculation of backpay from the date of the unlawful discharge to the date the new information was discovered." Id. The rule applies equally to claims under Title VII, including hiring claims. See Sellers v. Mineta, 358 F.3d 1058, 1061-62 (8th Cir. 2004) (applying McKennon to a Title VII case); Rivera v. NIBCO, 364 F.3d 1057, 1071 n.15 (9th Cir. 2004); Wallace v. Dunn Const., 62 F.3d 374, 378-79 (11th Cir. 1995) (same) (en banc); Wehr v. Ryan's Family Steak Houses, 49 F.3d 1150, 1153 (6th Cir. 1995) (same); Red Deer v. Cherokee County, 183 F.R.D. 642,647-48 (N.D. Iowa 1999) (applying McKennon to a Title VII failure-to-hire claim). As the Third Circuit noted in Mardell v. Harleysville Life Ins., 65 F.3d 1072, 1074 (3d Cir. 1995), "[o]ne purpose of Title VII is ‘to make persons whole for injuries suffered on account of unlawful employment discrimination.'" This purpose pertains, whether an individual is injured because of discriminatory discharge or discriminatory failure to hire, and Title VII dictates the individual should ordinarily be compensated. Id. The district court's denial of backpay is wrong for the same reasons that the Seventh Circuit rejected the employer's attempt to avoid backpay in Hartman Bros. Heating & Air Conditioning v. NLRB, 280 F.3d 1110 (7th Cir. 2002). The National Labor Relations Board ("NLRB") had ruled Hartman committed an unfair labor practice and had awarded backpay to an employee who was terminated immediately upon starting work, after revealing he was a union organizer. Hartman appealed, arguing it was not liable for backpay because within hours of sending the organizer home, it discovered, through its normal hiring process, that the organizer had a blemished (and disqualifying) driving record which he had lied about on his job application in violation of state law. Id. at 1114-15. The Seventh Circuit upheld the backpay award nevertheless, characterizing Hartman's arguments as "really just the old ‘unclean hands' defense of equity" that the Supreme Court expressly rejected in McKennon for statutory claims "where Congress authorizes broad equitable relief to serve important national policies." Id. at 1115-16. See also O'Neal v. City of New Albany, 203 F.3d 998, 1003-04 (7th Cir. 2002) ("An employer may still be liable for race discrimination under Title VII even though it later discovers information that would have otherwise disqualified the plaintiff from employment.") (age limitation for police officers). As with the driving record requirement in Hartman and the age limit in O'Neal, even assuming Dial demonstrated that its alleged "no criminal record" policy is an objective hiring qualification, it would not justify denying backpay to Wright-Bradley. McKennon suggests only one limitation on an award of backpay – "extraordinary equitable considerations." See 513 U.S. at 362 (before awarding backpay, court "can consider taking into further account extraordinary equitable circumstances that affect the legitimate interests of either party"). The Court did not explain what this means, and subsequent decisions in the courts of appeals provide no clarification. See, e.g., Mardell, 65 F.3d at 1074 n.4 ("We make no effort at this juncture to adumbrate the contours of the ‘extraordinary equitable circumstances' doctrine."). Since McKennon involved an employee who had engaged in on-the-job misconduct, however, misconduct, alone, cannot possibly constitute "extraordinary equitable circumstances" for denying backpay to a victim of discrimination. Whatever it may mean, equitable considerations here would appear to weigh more in Wright-Bradley's favor than in the usual case of employee misconduct. The most common "after-acquired evidence" of pre-hiring misconduct is where an employee lies to the defendant during the application process, by either embellishing a resume with false information or withholding from the job application form relevant or damaging information such as the absence of a required educational degree or the existence of a criminal record. See Mardell, 65 F.3d at 1073 n.1 (plaintiff lied about educational background on resume and job application); Russell v. Microdyne, 65 F.3d 1229, 1232 (4th Cir. 1995) (plaintiff misrepresented work history on resume and job application); Shattuck v. Kinetic Concepts, 49 F.3d 1106 (5th Cir. 1995) (plaintiff lied about educational background on resume); Wehr, 49 F.3d 1150 (plaintiff lied about employment background and medical history on resume); Wallace, 62 F.3d at 377 (plaintiff lied about prior criminal conviction on job application) (en banc). Wright-Bradley, on the other hand, did not withhold any information from Dial. Dial's job application does not ask for criminal background information, see, e.g., EEOC Ex.12, and Dial did not request it during the interview process. When Dial asked Wright-Bradley to authorize a criminal background check, she did. Thus, Wright-Bradley never misled Dial concerning this subject, and Dial never argued so below. Significantly, if Wright-Bradley had declined the job offer, she would have received full backpay without question (subject to the outcome of the present appeal), because Dial has not asked any of the claimants who declined job offers to undergo a criminal background search as a prerequisite to receiving court-ordered backpay. In sum, if plaintiffs who lied to their employers to hide a criminal record that would have disqualified them from employment are, nevertheless, entitled to backpay once they prove discrimination, see, e.g., Wallace, 62 F.3d at 378-79, then the mere existence of Wright-Bradley's criminal record cannot constitute "extraordinary equitable circumstances" that would warrant deviating from the "beginning point" articulated by the Court in McKennon: backpay should be awarded from the date of the unlawful discrimination to the date the after-acquired evidence was discovered. See 513 U.S. at 362. Even assuming Dial could meet its evidentiary burden (see discussion infra), all that would be barred is front pay and instatement (neither of which is at issue in this appeal). Id. at 361-62. Wright-Bradley is entitled to receive backpay from May 2000 (the month she took the WTS) to March 31, 2005 (the agreed-upon cut-off date for all backpay). The district court erred as a matter of law in ruling otherwise. 2. Dial Failed to Establish the Requisite Factual Predicate. The district court erred in denying backpay to Wright-Bradley for a wholly separate reason: Dial failed to make the requisite factual showing. To assert the partial defense of "after-acquired evidence" at the remedy stage, the employer must "first establish that the [employee's] wrongdoing was of such severity that the employee in fact would have been terminated on those grounds alone if the employer had known of it at the time of the discharge." McKennon, 513 U.S. at 362-63; see Sellers, 358 F.3d at 1069 ("after-acquired evidence doctrine ‘requires factual as well as legal development'") (Bye, J., dissenting) (quoting EEOC v. Farmer Bros., 31 F.3d 891, 901 (9th Cir. 1994)). The evidentiary standard the employer must satisfy on this point is stringent. See Russell, 65 F.3d at 1241. It is not met by mere representations of counsel, but requires evidentiary proof that the policy actually exists, is regularly followed, and would have actually required the plaintiff to be fired upon learning of the misconduct. Welch v. Liberty Mach. Works, 23 F.3d 1403, 1405-06 (8th Cir. 1994) ("employer bears a substantial burden of establishing that the policy pre-dated the hiring and firing of the employee in question and . . . constitutes more than mere contract or employment application boilerplate") (emphasis added), abrogated on other grounds by McKennon, 513 U.S. at 362.<24> Thus, an employer seeking to invoke this limitation on remedies bears the burden of establishing that the allegedly disqualifying information would have been uncovered and was, in fact, serious enough to cause the employer to fire her upon learning of it. Dial offered nothing more than vague and inconsistent oral representations of its counsel during the September 6, 2005, telephone hearing on this issue. This, alone, is wholly insufficient to satisfy Dial's burden of establishing that it would have conducted a criminal background check of Wright-Bradley in May 2000 and would not have employed her based on her convictions as of that date. In Welch, for example, this Court rejected as inadequate a greater quantum of evidence than Dial offered in this case, stating: As the movant for summary judgment, Liberty bore the significant burden of establishing that it had a settled policy of never hiring individuals similarly situated to Welch. Because of our concerns about creating perverse incentives for employers, we find that Maier's affidavit alone is insufficient to establish this material fact. 23 F.3d at 1405-06 (emphasis added). Because of the importance of the issue, this Court rejected Liberty's effort to rely solely on the contents of a company manager's "self-serving" affidavit to demonstrate its purported policy. Id. Here, Dial presented absolutely no evidence, relying solely on the arguments and representations of its counsel. The doctrine plainly requires more. In sum, this Court should reverse the district court's denial of backpay to Wright-Bradley because Dial presented no evidence that it actually conducted criminal background checks as part of the hiring process, would have learned of Wright- Bradley's criminal record in May 2000, and would not have hired her because of it. Given this absence of any evidence, backpay should be granted because Dial did not "prove" the elements for asserting this defense. Russell, 65 F.3d at 1240-41 (defendant did not meet its burden under McKennon to demonstrate it would have fired plaintiff for alleged misrepresentations in her resume and job application). CONCLUSION EEOC respectfully urges this Court to affirm the district court's finding of disparate impact, denial of JMOL, and award of backpay and other monetary benefits to 53 of the claimants, and to reverse the denial of backpay to Wright-Bradley. Respectfully submitted, JAMES L. LEE Deputy General Counsel LORRAINE C. DAVIS Acting Associate General Counsel DATED: April 24, 2006 SUSAN R. OXFORD, Attorney U.S. Equal Employment Opportunity Comm. 1801 L Street, N.W. Washington, D.C. 20507 (202) 663-4791 CERTIFICATE OF COMPLIANCE I certify that this brief complies with the type-volume limitation set forth in Fed. R. App. P. 32(a)(7)(B). This brief contains 16,496 words. See Fed. R. App. P. 32(a)(7)(B)(i). The brief was prepared using the WordPerfect 9 word processing system, in 14-point proportionally-spaced Times New Roman type for text and 14-point Times New Roman type for footnotes. See Fed. R. App. P. 32(a)(5). The accompanying diskette has been scanned for viruses and is virus-free. April 24, 2006 Susan R. Oxford CERTIFICATE OF SERVICE I, Susan R. Oxford, hereby certify that on this 24th day of April, 2006, I caused two copies of the attached brief and one diskette containing an electronic version of the brief in pdf format to be sent by Federal Express, postage prepaid, to counsel of record at the following addresses: Camille A. Olson, Esq. Anne E. Duprey, Esq. SEYFARTH SHAW LLP 55 East Monroe Street, Suite 4200 Chicago, Illinois 60603 (312) 346-8000 Attorneys for Appellant/Cross-Appellee Dial Corporation April 24, 2006 Susan R. Oxford ADDENDUM *********************************************************************** <> <1> “R.#” refers to the district court’s docket entry for the document in question. <2> The trial transcript is denoted “Tr.Vol#” except for the portions that were transcribed and paginated separately: Dr. Campion’s direct examination (August 16) is “CampionTr.I”; his cross-examination (August 17) is “CampionTr.II.”; and Dr. Jones’ testimony and Dr. Campion’s rebuttal testimony (both August 19) are “Jones/CampionTr.” <3> Dial asserts it “did not fail a single female applicant from March 6, 2001 until October 28, 2002.” Dial Brf. at 16 n.3. This is because Dial did not administer the WTS to anyone from mid-March 2001 until October 2002. <4> See J-1 at 16-18 & WTS result forms ## 120-23, 128-35, 198-99 (Walker, Sperry, Harriman, Vanderlip, Hawkins, Featheringill and Guthrie performed the WTS for at least seven minutes; Dial deemed all seven to have failed). <5> Michael Kruse, union president and a maintenance mechanic, testified that the smokehouse jobs had not changed since he started out as a smokehouse unloader in 1974, Tr.II 106-07, and that the videotape (EEOC Ex.11) accurately portrayed the smokehouse jobs. Tr.II 111-14. He testified that the WTS was more difficult than the job in several respects. Id. at 114-16 (in job, “cups” secure rod, making it easier to lift and place than in WTS; pace in job is more relaxed than WTS, which requires continuous movement), 123 (upper grooves are several inches higher on WTS; in actual job, chains “come down at an angle” to a height of about five feet), 126-29 (some distances walked are much shorter in job than in WTS). <6> Asked, on cross-examination, about similar comments on the forms of women who passed the WTS, Dr. Campion reiterated that scoring the WTS was “a very subjective process.” CampionTr.II 46, 58 (“unclear from the sheets themselves how the judgments were made”). Indeed, the WTS result forms reveal additional men who passed and were hired even though their score sheets suggest they had difficulty completing the test. Michael Howell was told to slow down because he was breathing heavily and perspiring, and took a short rest toward the end; Jerry Kester slowed his pace because of fatigue and became somewhat short of breath as the test progressed; Blain Martin was “very nervous [and] shaky” as well as “flushed” and “perspiring”; and Ken Martin perspired profusely. See J-1, WTS result forms ## 218-19, 270-71, 276-77, 279. Although these comments suggest these men struggled to complete the test, Lutenegger deemed each to have passed. See J-1 at 19-20. <7> Lutenegger also testified concerning various measures Dial took to reduce injuries, prior to the WTS. See Tr.III 313-20. <8> Dr. Jackson’s expertise is the measurement of human performance. Tr.IV 498-500. Dr. Jones testified that his expertise – industrial and organizational psychology – is “generally the same discipline” as Dr. Campion. Jones/CampionTr. 4. <9> The jury awarded compensatory damages of $1 each to two women who took the WTS in March 2001. The district court set these awards aside as inconsistent with the jury’s finding that intentional discrimination began in April 2001. JA82-83. <10> Dial never moved for a new trial under Fed.R.Civ.P. 59. See R.95 (JA77-78). <11> Dial claimed to have provided “examples of [applicant] background checks from 2004, 2003 and 2000.” TelCon.Tr. 6. In fact, Dial provided only two: Garrick Batley (August 2004) and Corey Anderson (December 2000). <12> Dial had informed EEOC, prior to the hearing, that in the five years preceding May 2000, Wright-Bradley was convicted of felony theft in 1995, two misdemeanors in 1999 and, although not relevant to what Dial might have decided in May 2000, felony theft in December 2004. Tel.Con.Tr. 4, 10. At the hearing, Dial claimed that her criminal record was actually more extensive, stating that in the seven years prior to 2000 Wright-Bradley “was convicted of 11 different criminal offenses, including felony retail theft, assault with intent to do serious bodily harm, two other misdemeanor assaults, a harassment misdemeanor conviction, a conviction for making false statements, [and] a conviction for obstructing justice . . . .” Id. at 7. <13> Albemarle was referring to EEOC’s 1970 Guidelines on Employee Selection Procedures, 35 Fed. Reg. 12333 (August 1, 1970) (Addendum). Like the present UGESP, the 1970 Guidelines provided that use of a test with an adverse impact on a protected group violated Title VII unless validated through “empirical data” demonstrating the test is “predictive of or significantly correlated with important elements” of the job. See 29 C.F.R. §§ 1607.3, 1607.4(c) (1970). Thus, Albemarle did not “introduce” the concept of test validation, Dial Brf. at 45, since it already existed in the 1970 Guidelines. The UGESP is based on and expressly supersedes the 1970 Guidelines, 29 C.F.R. § 1607.1(C), and reflects the same concepts. See 29 C.F.R. §§ 1607.3(A), 1607.5(A)-(C) (1978). <14> Dial argues it is not required to demonstrate “business necessity” in accordance with the UGESP, see Dial Brf. at 46 (EEOC’s Guidelines “do not trump Albemarle”), but that is exactly what its experts attempted to do at trial. See Dial Brf. at 47-54. In any event, Dial never explains how it thinks the requirements of UGESP and Albemarle might differ and, in fact, they do not. <15> The district court did not require Dial to prove both content validity and criterion-related validity, as Dial erroneously argues. Dial Brf. at 50. The district court recognized Dial was entitled to prove validity by any one of several means, and simply noted that Dial chose these two particular alternatives, either of which would have been adequate. JA92-93 (citing 29 C.F.R. § 1607.5 (“users may rely upon criterion-related validity studies, content validity studies or construct validity studies” to establish a test’s validity) (emphasis added)). <16> Dr. Campion made this point in his cross-examination, Campion Tr.II 28, contrary to the district court’s belief that this was not addressed. Id. at 11 n.7 (JA94). <17> The district court did not err as a matter of law in crediting this testimony, Dial Brf. at 52 n.13, which was based on Kruse’s personal observation and experience. See Tr.II 106-29. <18> Contrary to Dial’s assertion, Dial Brf. at 54, the district court’s distinction between more and less serious workplace injuries has support. UGESP directs test validity to be assessed against “generally accepted professional standards.” 29 C.F.R. § 1607.5(C). Dr. Campion explained that documentation of more serious injuries is likely to be more accurate and, therefore, provide a more reliable analytical basis. CampionTr.I 42-43. Dial’s expert agreed. Jones/CampionTr. 39-43; see Dial Exs. U(1), (12), (18) (JA47, 55-56). <19> The last sentence of the district court’s finding 50 (JA90) contains a typographical error. The average rate of injury for 1998-2003 was actually 0.08, not 0.8 (as the district court mistakenly wrote). See EEOC Ex.36 (total number of strength-related worker’s comp injuries from 1998-2003, 11, divided by 140.46, the total full-time exposure during this time, equals 0.078). The six-year average could not have been 0.8 because that would make the average higher than the rate in any of the composite years. See Finding 49 (reporting injury rates of 0.09 for 1998 hires, 0.14 for 1999, 0.05 for 2000, 0.16 for 2001, 0.06 for 2002, and 0.00 for 2003). <20> See Tr.III 466 (Farrington testified WTS “was not put in place regarding performance”); Tr.III 483 (Hartlage testified WTS was driven by “accident rates” and was not based on performance); Jones/Campion Tr. 57 (Jones testified WTS was not designed to predict general success on the job). <21> The district court stated: “Even assuming Dial had been able to prove the WTS was content- and/or criterion-valid, the Court finds Dial has failed to fulfill its burden to show it had a ‘compelling need’ for implementation of the WTS, and that other, non-discriminatory mechanisms – namely, many of the same safety programs actually implemented by Dial – could not produce the same results.” See Order on Liability at 15 (JA98) (emphasis added). <22> The district court’s well-supported rejection of Dial’s asserted “business necessity” and Dial’s concession that it instituted other injury-prevention measures which reduced injuries distinguish this case from Int’l Bhd. of Elec. Workers v. Miss. Power & Light, 442 F.3d 313 (5th Cir. 2006), where the question of alternatives was properly raised because the district court there (unlike here) found that the company had “most adequately . . . justified” the test being used. Id. at 319 & n.11. <23> Dial argues that 13 women who faltered during the WTS passed, and all but one of the men who received negative comments failed. Dial Brf. at 38 (citing Lutenegger’s testimony and the WTS result forms). The actual forms show otherwise, see, e.g. pp10-11 & n.6, supra, and the jury was entitled to reach its own conclusions based on the evidence. <24> In Welch, this Court ruled that after-acquired evidence can defeat a discrimination claim entirely. Although McKennon abrogated that ruling, Welch is nevertheless instructive with respect to the legitimate uses of after-acquired evidence – to establish the amount of damages for which the employer may be liable.