Daniel Anderson, Jr. v. Douglas & Lomason Co., Inc., Douglas & Lomason Co., Inc.

REHEAR- PANEL FOR PETITION SUR RE- FOR SUGGESTION WITH ING BANC IN

HEARING 25, 1994

July by Appel- rehearing filed for petition judges who having submitted

lant been Court the decision

participated judges in circuit other available

to all service, judge who concurred no active rehearing, and for having asked

the decision the circuit judges of the circuit majority of having voted for service regular active banc, petition by the court rehearing rehearing DENIED. al., ANDERSON, Jr., et

Daniel

Plaintiffs-Appellants,

v. CO., LOMASON &

DOUGLAS Defendants, al., INC., et Inc., Co.,

Douglas Lomason

Defendant-Appellee. 92-7554.

No. Appeals, Court States

United

Fifth Circuit. 23, 1994.

June July

As Corrected Rehearing on Denial

As Amended En Bane Rehearing Suggestion 9, 1994. Sept.

1280 Vinick, Law- Seymour, Sharon

Richard T. Rights, Washing- for Civil yers’ Committee ton, DC, appellants. for Marcosson, E.E.O.C. Office Samuel A. DC, ami- Counsel, Washington, General cus E.E.O.C. McPherson, Godfrey, George K.

Robert GA, Hancock, Atlanta, Jr., Smith, Currie appellees. JOHNSON, SMITH, and EMILIO Before GARZA, Judges. Circuit M. GARZA, Judge: Circuit

EMILIO M. law- brought a class action The (“D Douglas & Lomason Co. against the suit 5, 1985, July Company”) on L” or “the & intentionally discrimi- D L alleging that & hiring, promotion, in its against blacks nated The district practices. termination trial, court, entered lengthy bench after a Title plaintiffs on their judgment against seq., and 42 U.S.C. § 2000e et VII, 42 U.S.C. court’s the district We affirm § claims. decision.1 I Cleveland, Boli- plant in operates LD & Mississippi, that manufactures County, var ap- employs plant parts. automobile 70% of whom people, proximately posi- entry-level seeking Applicants black. certain only meet L& need with D tions least being at as such requirements, minimal per- age, physically able years of eighteen write. work, to read and and able form hiring, pro- challenge certain motion, followed practices and termination plant after James Cleveland L at the D & court's decision L. We affirmed appealed earlier plaintiffs in this case 1. The v. injunction in Anderson deny preliminary appli- denying their the district court decision of (5th Co., Cir. F.2d 128 Douglas & Lomason injunctive preliminary relief for cation 1988). hiring practices D &of regulated the would have plant manager person became October ments.2 If a called or came to the period 1982. The October 1982 plant between while D L hiring & was but after the April being when Grizzard ceased man- applicant pool Haynes reached what consid- ager plant, at the Cleveland was referred to optimal size, ered to plant guards be the told years.” at trial as “the Grizzard the individual either that D L& was not hiring taking applications.3 or was not Dur-

A ing periods & L hiring, was not Haynes guards challenge or the first informed individuals practices adopted by Patty Haynes inquiring employment when she about either that the personnel manager Company became taking the Cleveland was not applications or was plant September Haynes’s hiring hiring. Thus, possible it was for individ- practices, D L throughout followed uals to be told that D & L taking was not years, substantially differed applications Company both when the was and predecessor, from those of her Harold Kee- hiring. was not ton. employees Keeton ap- allowed take D & L also hired in what can plications persons, people other allowed “spurts.” Grizzard, best be called after applications to submit twenty-four hours a meeting with department his heads and su- day, applications took whether or Dnot & L pervisors, would determine whether D L& hiring, kept applications in his needed to hire additional workers. If D L& year. “active” file generally for one He in- workers, did need additional Grizzard would applicants terviewed two to every three Haynes, so inform begin taking would position. making available hiring deci- applications. Haynes then would hire the sions, applicant’s Keeton looked at the expe- employees sought number of by Grizzard. D rience and attempted applicants to hire those L rarely & employees hired on a one-at-a- with the most experience. work relevant time or continuous basis and sometimes went Haynes, hand, on the other would not let anyone. months without hiring employees home, applications take required people who apply employ- wanted to for D & L application modified its hiring and ment to plant, come to D & accepted L’s procedures May using the Missis- applications only when D & L hiring, was sippi Employment (“MSES”) State Service to " kept applications and for six months. She provide pool of applicants from which it accepted only also applications or one two for could hire. When D & L needed to hire every available position. Haynes testified employees, notify it would the MSES. The decisions, that in making hiring she looked at MSES then would regular refer both and on- applicant’s experience work relevant and (“OJT”) the-job-training applicants to D & person would hire “the job.” for best Haynes L.4 would interview the referred During periods when D & L hiring, applicants and select employees using new Haynes permitted individuals who separate job called the regular orders for and OJT Company seeking possible information applicants about hiring hired. All from the MSES employment appointments to make at which referrals separate occurred on three spurts they could fill out application. Haynes May from September to 1985. D & L has accepted also applications from any production individuals not hired workers since that plant came to the appoint- without time. large 2. Because of employees number Haynes guards in- people instructed the to allow to deciding volved in D when & L needed to hire plant applications enter the to fill out employees, hiring news quickly that D & wasL hiring D gave & L was guards and also spread throughout Company’s general work names of appointments those individuals with to example, force. For one occasion individuals applications. fill out plant came applications to fill out even before Haynes Grizzard told that D & L would be 4.D & L received financial benefits from the employees. additional on-the-job training program state’s every for OJT applicant fence, hired. security 3. A gate, guardhouse prevent- and persons ed entering plant. unauthorized from such, D & positions. craft As to hold ments B jobs seeking employees those required LL D & next contend bidding process and main- participate of discrimi- practice or pattern in a engaged pro- supervisors selected department tenance blacks to lead- refusing to promote nation temporary upgrades to employees duction D L no positions.5 & had foreman er and primary duties positions. The maintenance promotions, guidelines for criteria or written employees included maintenance general concerning notices post did and Grizzard broken machin- welding repairing and pipe Instead, opportunities. promotion available however, Welding, primary ery. and selected leaders supervisors department employees. rack maintenance function of with or departments little for their foremen welding D L considered Consequently, & Although D & L did input from no Grizzard. important senior- more than experience to be promotions, criteria not have written temporary upgrades ity purposes of for the following consistently applied the supervisors to maintenance permanent promotions and selecting leaders subjective criteria when jobs. department record, attitude, relevant foreman: work abilities, willing- leadership experience, work D help employees, commitment other ness D L finally that & contend Thus, L, seniority. D & L’s D & Company’s discharged employees, using the predominately of a sub- criteria motion policy pretext, as a retaliation garnishment

jective nature. filing charges with the for the Opportunity Commission Equal Employment C (“EEOC”). provided shop D L’s rules & discharge any em- Company would allege that D & L en- also garnish- more ployee four or who received by failing discriminatory practices gaged in from, a two- within different creditors ments upgrades temporary with provide rule, only those year period. Under this promotions to maintenance permanent withdrawn, satisfied, or garnishments not employed L positions. & department L in- days D & thirty stayed within after required by its collec- job-bidding process, as of their existence employee formed the agreement with union bargaining tive limit. four-garnishment against the counted whereby employees, L’s representing to run on the period began two-year The bids for available employees had to submit garnishment L discussed the first date D & however, L, D & some- positions. “craft” date expired on the employee and with the temporary employees gave production times the period for thirty-day grace the jobs on a short-term basis.6 upgrades to craft Haynes, who expired. garnishment fourth employees to bid L D did allow & gar- administering the responsible for super- the positions, and temporary upgrade garnishments forgave all policy, nishment an em- departments into which visors she became L received before that D & temporarily upgraded deter- ployee would be manager. personnel em- The chosen upgrade. mined whom however, block the could supervisor, ployee’s II needed for employee was upgrade if the A work.

duction that the dis plaintiffs contend gener- employees D L considered hiring and by analyzing the erred trict court depart- maintenance and rack al maintenance promoting to the sala- in L discriminated production lead- promotes 5. & LD performance positions. supervisor and their ried so that it can evaluate ers D & L either potential foremen. become among hires the leaders or foremen from selects them temporary give employee a & L would company. Both leaders from outside the regular craft position a upgrade when to a craft paid hourly. L also em- D & are and foremen vacation, simply when or employee or on was ill supervisors, from ploys whom selects salaried employee be- craft an additional & LD needed the among hires from outside the foremen or load. work of an increased cause D & plaintiffs contend company. do not 1284

promotion class disparate claims under the employment, that can be shown to have a model treatment instead of dispa under the causal connection to a class based imbalance impact rate model.7 We examine discrimina in work Pouncy force.” v. Prudential using tion disparate impact claims model Co., 795, (5th Ins. Cir.1982); 668 F.2d 800 see employment practices “facially neu Cove, 657-58, also Wards 490 U.S. at 109 tral their treatment groups of different but (noting S.Ct. at 2125 plaintiff alleging that a ... fall harshly fact more group on one discriminatory impact “specifi- claim must justified than another be cannot busi cally challenged practice show[] each necessity.” ness International Bhd. of significantly disparate has a impact on em- States, Teamsters v. 324, United 335 ployment opportunities for whites and non- 15, 1843, 15, n. 97 S.Ct. n. 1854 52 L.Ed.2d whites”).8 (1977). Although 396 class action may challenge subjective The district court discretionary or refused to ana employment practices lyze plaintiffs’ disparate promotion under' the model, impact Watson using disparate v. Fort Worth claims impact Bank & model be Trust, 977, 990, U.S. 108 S.Ct. identify specific cause did not aspect 2786-87, (1988), 101 L.Ed.2d 827 subjective use of decision-making by D & L that subjective or discretionary decision-making was any shown to have causal connection to does not itself create an inference of discrim alleged class-based imbalance in D & L’s inatory 999, 108 conduct. Id. at S.Ct. at 2791 general supervisory or work force. The (plurality opinion). Instead, plaintiffs pointed only policy to D & L’s “the plaintiffs burden in establishing requiring applications individuals to fill out prima facie case goes [of discrimination] plant specific as a employment practice beyond the need to show that there are causing a class-based imbalance in the work disparities statistical employer’s This, however, force. justify does not ana work plaintiff force. The begin by must lyzing the case disparate under the impact identifying specific employment prac- model because the did not demon tice that challenged.... is Especially in disparity strate that the they complain of was an.employer cases subjec- where combines the result of the challenged policy. See tive criteria with the rigid use of more Cove, 657-58, Wards U.S. 109 S.Ct. at tests, standardized plaintiff rules or is Moreover, no identified responsible our view isolating specific policy allegedly caused a race- identifying specific employment prac- based imbalance in the persons number of allegedly tices that are responsible promotions. received Instead, plain observed disparities.” statistical merely tiffs a wide-ranging launched attack Atonio, Wards Packaging Cove on the Co. v. cumulative employ effects of & L’s 642, 656, practices. ment disparate impact model (1989) Watson, L.Ed.2d 733 (quoting appropriate not the vehicle from which to 994, 2788-89). U.S. at Thus, 108 S.Ct. at launch such an Pouncy, attack. 668 F.2d at disparate impact “[t]he applies only model Consequently, the district court did not employer when an has instituted a specific err refusing analyze to plaintiffs’ the procedure, usually a selection criterion using disparate claims the impact model.9 though plaintiffs alleged Even violations of employment practice.” fication for his Cove, Wards VII, § both 659, 1981 and Title the elements of both 490 U.S. at 109 S.Ct. at 2126. How- Flanagan ever, claims are identical. Henry v. Aaron E. proving "the ultimate burden of that dis- Cir., Community 1231, Health Servs. 876 against protected F.2d crimination group a has been (5th 1989); Lab., Inc., Payne 1233 Cir. v. Travenol specific employment a practice caused re- 798, (5th Cir.), denied, 673 F.2d 818 Watson, cert. plaintiff 459 mains with the at all times.” 1038, 451, (1982). U.S. 103 987, S.Ct. 74 L.Ed.2d 605 2790, 487 U.S. at quoted S.Ct. at 108 Therefore, employ only Cove, we analysis 549, one in evalu Wards 490 at 109 U.S. S.Ct. at 2126. ating plaintiffs' § the Title VII and 1981 claims. 9.The district court also held that if even the plaintiff 8. After prima the has established a present prima facie did a facie case under discrimination, case of employer "the disparate impact the theory, they carries not be would producing burden justi- a satisfy evidence of persuasion business able to their burden of because In clearly Id. de erroneous standard.” claims under plaintiffs’ review the thusWe clearly finding is erro termining whether a model. treatment disparate neous, give regard to the trial we must “due B credibility the witness judge to City, 470 U.S. Anderson v. Bessemer es.” action, class disparate treatment In a 1504, 1511, 564, 573, 105 84 L.Ed.2d 518 S.Ct. discriminatory in prove plaintiffs “must (1985). finding ‘clearly erroneous’ “A ‘the mere than demonstrate more tent and support although there is evidence when spo accidental or isolated or occurrence ” it, evidence reviewing court on the entire v. discriminatory acts.’ Carroll radic firm conviction (5th the definite is left with 188, 190 Co., Sears, 708 F.2d Roebuck & United a mistake has been made.” Teamsters, Cir.1983) 431 U.S. (quoting Co., 333 Gypsum v. States 1855). United States plaintiffs must The 97 S.Ct. at 525, 542, 364, 395, L.Ed. 746 S.Ct. the evi “by preponderance of establish (1948). [em that racial discrimination dence procedure operating ployer’s] standard —the practice.” than unusual regular rather Ill Id. plaintiffs argue the district court may pri establish D L did not discrimi- that & erred “by the disparate treatment

ma facie case of against by restricting access to nate disparity in the ‘gross’ if a of statistics use They argue that & L application forms. race is based on of workers treatment Haynes re- unlawfully restricted access not Id. “statistics shown.” applications to a give employment fused and, irrefutable; variety they infinite come in seeking D L of blacks while & was number evidence, may they be any other kind like employees. new hire short, depends their usefulness In rebutted. circum surrounding facts and all of the on prove prima attempting Teamsters, stances.” ease, plaintiffs primarily relied on facie If evidence is at 1856-57. S.Ct. statistical provided by class mem evidence anecdotal intent, discriminatory to establish insufficient A of individuals testified number bers. by intro may their case bolster applications from unable to obtain individual, historical, ducing or circumstantial found this anecdo L. The district court D & Corp., 841 v. Oil evidence. Bernard Gulf it was con unpersuasive because evidence tal Cir.1988). (5th employer F.2d poli general application with D & L’s sistent prima facie may plaintiffs’ then rebut finding is erroneous. This not cies. introducing proof plaintiffs’ sta “by case Additionally, the part VIII.A See infra. insignificant’ or are ‘inaccurate or tistics gate by away at the turned individuals ‘non-discriminatory explanation providing a give D L guards L never saw D & ” discriminatory result.’ apparently seeking employment, to whites applications omitted). (citations Id. derog directed employee D L no & ever at them.10 racially-oriented epitaphs atory or ultimate determination “[T]he L Moreover, that D & time around same is a discrimination of intentional existence applica- give these black individuals fact, appeal under did question of reviewed specific employ- against a result of a as justified and nated "adequately the method & L D practice. put ment hiring by its which it manner conducted totally legitimate to race reasons unrelated forth changes practice. Anderson, making individual, its alleged various that Poitier One 10. such, Company that plaintiffs failed to show was not ... have guard As him that the told changes hiring practices and there- district court proffered applications The giving to blacks. testimony em- be- legitimate ends of the credible Anderson's in did not serve did not find and filled plaintiffs claims filed two EEOC ployer.” Because hold that cause Anderson Id. we case, alleged questionnaire after prima we do not an EEOC present a facie out not did mentioning it. ever by prepon- made without was they statement whether demonstrated review erroneous. is not This were discrimi- that derance of the evidence tions, provided D applications many- & L with constituting blacks opera- 70.1% of all Finally, other blacks. the record contains no tors, laborers, and fabricators. indicating application

evidence that D & L’s procedures potential appli- affected A any differently cants potential than white Thus, applicants.11 the district court’s find-

ing D L against that & did not discriminate seeking applications

blacks who were is not plaintiffs The argue first that the district clearly erroneous. See Falcon v. General Co., court (5th crediting erred D Cir.1987) Tel. & L’s statistical 815 F.2d (finding that an employer’s application proce- analysis hiring practices of its during the dures were not discriminatory where no evi- years. plaintiffs The contend that dence was prove pro- introduced that the D analysis because & L’s statistical was fa- cedures affected differently non-whites flawed, tally the district court should have whites). than relied on seven sets pre- statistical data of by L, expert.12

sented their D surpris- & not ingly, plaintiffs’ finds fault with each seven IV standards contends that the district plaintiffs The contend that statistical and correctly upon court relied the standard anecdotal evidence established that D & L posed by D & expert. L’s against discriminated hiring blacks when production workers the Grizzard Where use statistical years. comparing In the number of blacks challenge evidence employer’s an hiring hired to the applied number of blacks who practices, evidence, probative to be production jobs, however, discriminatory intent, compare must the rele court found no evidence of discriminatory portion vant employer’s work force intent. qualified population with the in the relevant job production The worker —an un- labor market. EEOC v. Dairy Olson’s skilled, entry-level position required only — Queens, Inc., (5th 989 F.2d Cir. applicants old, be least be 1993). dispute between the write, bend, able to read and and be able to and D L& revolves properly around how to lift, stoop. During years, define the “relevant labor market.” D & L & L production hired 482 workers —46.5% of contends that the labor market should be those hired were black. Black using “applicant defined analysis— flow” constituted more 70% of D gener- than & L’s i.e., arguing that D L’s& available labor force, al work and more than 80% the force is best examining determined production force, work during the Grizzard applications years. In those County, persons actually Bolivar where & L’s plant located, sought employment Company. 1980 census data indicates with the black, overall plaintiffs, labor force was 51.8% hand, on the other contend that an argue also Haynes D & L discrimi- morning, arrived for work one against nated relying on word-of- Haynes asked application, Haynes for an told her persons mouth to job open- inform interested hiring. that D & L was however, morning, Later that ings having practice and that resulted in blacks Haynes Grizzard informed that D & L hiring less access than whites to informa- employees, would be new and Harris then officials, however, tion. D & L testified that it application. obtained an knowledge among employees, common including non-managerial positions, those in agree- The EEOC filed an curiae amicus brief Company fact, going hiring. the non-employees to be plaintiffs’ with the contention that the district occasion knew D & L was about to Haynes. hire before Grizzard analysis informed erred in its statistical of the evi- Harris, example, plaintiff For along note, however, Chris with dence. We that the EEOC office blacks, black, two other non-super- heard from Mississippi rejected in as meritless the discrimi- visoiy employee hiring. that &D L was Howev- charges by plaintiffs. nation filed the er, Harris, waiting when plant was at the

1287 available; imperfections, skewing allow for not be used analysis should applicant flow factors, error; margins of and then take applications sub- because some here figures worth. not the for what are Some- were during the Grizzard mitted much, Thus, plain- times this sometimes little.” Phil- the for examination.13 available Thus, applicant lips, 637 F.2d at 1025. of their seven differ- argue that one tiffs figures are flawed or otherwise measuring labor flow unavail- of the relevant methods ent able, plaintiffs to applicant we have allowed use other superior & L’s market was prima measures to establish a facie statistical analysis.14 flow See, e.g., (noting case of discrimination. id. applicant figures un- that actual flow figures are the applicant flow Actual employer the did not iden- available because by which to measure an preferred method race); tify applicants by v. Robinson Union practices performance. employer’s hiring (5th 652, Corp., 538 F.2d 657-58 Carbide 168; Queens, Phil Dairy 989 F.2d at Olson’s Cir.1976) (upholding the district court’s re- Comm., 637 F.2d lips Legislative v. Joint employ applicant analysis), flow fusal an denied, (5th Cir.1981), 1014, 456 1025 cert. part grounds, in on other 544 F.2d 2035, 960, L.Ed.2d 483 modified 72 102 S.Ct. U.S. (5th Cir.1977). 1258 Co., (1982); Ry. 497 F.2d Hester v. Southern (5th Cir.1974); 1374, Mister v. see also 1379 Co., 1427, expert, Dr. & L’s statistical Joan R.R. 832 F.2d

Illinois Cent. Gulf that, (7th Cir.1987) (“Statistical Haworth, according ap to an analysis of testified 1435 analysis, D L not discrimi advantage plicant flow & did applicants has the the actual hiring applicants. Haworth study how the nate the examines self-selection: 1985, April January found that from 1984 to actually people who wanted employer treated applicants. of all preferable as blacks constituted 46.7% job. Applicant studies are the compared figure per governs treat then Title VII the She a rule because centage by D L applicants hired & applicants.”). “Other statistical mea ment of differing years who were black —45.5%— necessarily imperfect are sures not the difference was varying degrees. The best the and determined ways and Thus, statistically significant.15 Dr. Haworth figures accept is to what court can do performance, whether suggest we need not address that we L’s 13. The also seem failing the against court erred in to follow D & L draw adverse inferences should plaintiffs to define Company's preserve seven standards offered the failure to because of the year period, pool to D & L. The district employment applications the labor available for a one being Employee rejected as flawed. each standard pursuant Guidelines on to the Uniform Procedures, 1607.4(A)-(B) § C.F.R. Selection 29 (1978). Guidelines are "Uniform applications into six 15.Dr. Haworth divided promul legally binding. They have not been event, "pools,” reflected each which gated regulations and do not have the force of as compared percentage of blacks and whites Clady County Angeles, F.2d v. Los law." pool. Only applicants each hired from the Cir.1985) (9th (citing General Elec. Co. v. 5, hiring pool pool covered one- one — 401, Gilbert, 125, 141-42, S.Ct. 410- period October 1984—reflected statis- month denied, (1976)), cert. 50 L.Ed.2d 343 tically significant Haworth standard deviation. (1986). 89 L.Ed.2d 915 basis, pools aggregate analyzed also Moreover, rough applications kept D & L comparing percentage hired who of blacks preceding period ly correspond month to the six mid-January through applied April 1984 from charges. plaintiffs' filing EEOC Al of their agree parties to be found what 1985. She normally though would have discarded D & L significant Ha- statistically standard deviation. months, Haynes April applications after six however, testified, pool largely ac- worth being kept applications after testified that she excluded deviation. When she counted for the Consequently, charges. EEOC notified of the analysis, found the standard pool she 5 from analysis applications available for number of statistically insignificant. The to be of deviation keeping appli policy with D & L’s consistent testimony that Haworth's district court credited We therefore de six months. cations for pool 5 was due result found in the aberrational plaintiffs' suggestion. to follow the cline appli- percentage white to the fact that experience employment with relevant cants result, pools. As a higher pool in other 5 than agree court that we with the district 14. Because pool 5 should be found that the district court applicant the best measure of D flow was actual 1288

concluded, engage applications, D L in discrimi- she nevertheless a & did had “reason- natory hiring practices. ably large applications amount” of the for the

purposes analysis. of her The district court dispute by finding D L resolved this that & challenge plaintiffs Dr. Ha- applications pertaining did retain all to the analysis grounds. They on several worth’s period, say relevant and we cannot that this keep that D L first assert & did not clear resolution was error. See Trevino v. applications April from the to October 16 (5th 896, Holly Sugar Corp., 811 F.2d analysis period, skewing thus the so that the Cir.1987) (finding that the district court did applicants period number of black from that accepting employer’s not err statistical was understated. As evidence of & L’s because, analysis flawed, while was the applications, plain selective retention of the plaintiffs’ analysis suffered from more seri- large point tiffs difference between the deficiencies); City ous Nash v. Houston applications by number submitted blacks Ctr., 491, (5th Cir.1986) F.2d Civic during April period to October and the (noting yet competing where there are January period.16 October 17 to The district evidence, permissible views of the we found, however, must explanation that one affirm finding). the district court’s large applications for the increase in submit by period ted latter Robinson, employ that Oliver D L black & argue next that even if ee, begun campaign encourage had applicant analysis flow was the best stan County apply blacks in Bolivar employ for by pool dard which to measure the labor ment at D & L and file discrimination L, available to & the district court erred charges against D & L with the EEOC.17 accepting analysis Dr. Haworth’s because it finding clearly

This is not erroneous. samples based too small to allow for a

meaningful analysis. statistical When statis sizes, As additional evidence that D L sample tics are based on small contend, applications, failed to retain presence or absence of applications significance contend that the number of re statistical is not a rehable indica by Haynes tained D L& indicates that ac tor of discrimination. Whether or not a sam cepted approximately eight ap ple yield statistically and one-half is too small to meaning plications every positions results, however, seven filled dur ful must be determined on Watson, pre-suit period, accepted but case-by-case about basis. applications every positions 3; ten seven 995 n. 108 S.Ct. at 2789 n. Rendon v. (5th post-suit period. plain filled Technologies, AT & T 883 F.2d Cir.1989). tiffs contend that the difference demon The district court found Dr. Ha- strates D L applica analysis that & did not retain all worth’s to be more credible and April period. tions from persuasive analyses to October more than the seven of however, Haynes, by plaintiffs’ testified that she expert. believed fered We find that kept applications April she all the from the the district court in crediting did'not err D & period, although ultimately October analysis, she was L’s statistical which demonstrated unsure ap engage whether she had discarded some that D L in discriminatory & did not plications. Moreover, hiring practices, analyses Haworth testified that over the offered plaintiffs.18 while she believed she had all the relevant analysis anomaly. charges they signed excluded from the as an This that Indeed, and filed with the is not erroneous. See Lewis v. person EEOC. one Grant- —Richard

NLRB, (5th Cir.1985) 750 F.2d 1276 n. 17 charge merely stated that he did not read the but (noting significant that one standard deviation signed presented it when it was to him. See note treatment). disparate did not establish 46 infra. Applications submitted blacks constituted April applications 18.Assuming arguendo pools to October 37.6% too January applications. of the October to permit meaningful analysis, 60.4% small to statistical analyzed pools Dr. Haworth also on an record, reviewing many 17. In we note that aggregate statistically signifi- basis and found no familiarity the witnesses had little or no with the B plaintiffs attempted to The buttress their argue that the dis- plaintiffs further The anecdotal statistical evidence with evidence anal- accepting Haworth’s erred trict court alleg- D L regarding class members whom & during the hiring practices ysis of D & L’s edly did not hire of their race. because Company exclusively re- period plaintiffs may which use While Title VII evidence regarding of discrimina- individual instances provide acceptable appli- upon to lied MSES claim, Bernard, tion to bolster their class analysis, relying Dr. Haworth’s cants. F.2d at the district court found that & job-order- found that her the district court refusing L did hot discriminate in to hire the analysis persuasive by-job-order was more individuals at issue. This is not closely reflected the actual it more because clearly part erroneous. See VIII.B infra. employ- through D & L hired process which Accordingly, plaintiffs’ evi- anecdotal analyses, plaintiffs’ tout au con- ees. The dence, combined with their statistical evi- traire, job orders grouped the six different dence, prima did not establish a facie case of set of statistics.19 into one disparate treatment. V placed job orders with MSES D & L statistical, plaintiffs introduced histori- years. On times three cal, attempt and anecdotal evidence in an occasion, types sought D L two of & each prove unlawfully D L discriminated that & requested regular refer order referral —one promoting employees to against blacks when requested OJT referrals. rals and one order positions. plaintiffs’ foreman leader and compared percentages Dr. Haworth regarding promotion statistical evidence by D L for each whites hired blacks and compared the number of blacks claim job that in one separate order and found positions and foreman dur- moted to leader

job the difference between order did years per- ing pre-suit Grizzard with the expected of blacks hired number centage present actual in & L’s work blacks force; sought compare also The dis two standard deviations.20 exceed promotions by percentage received found, light record as a court trict pre-suit whole, single disparity did that this statistical pre- respective percentages for the with the prima facie ease of discrimina not establish L, post-suit periods. D & on Grizzard and tory find that the district court’s intent. We hand, compar- introduced statistics the other analyses regarding the statistical conclusions holding of blacks leader the number clearly hiring practices are not D & L’s of blacks positions to the number foreman NLRB, 750 F.2d Lewis v. erroneous. See in Bolivar positions holding supervisory (5th Cir.1985) (noting that a 1276 n. 17 ultimately court deter- County. The district “by no statistically significant disparity single analysis was that D & L’s statistical mined of discrimina commands an inference it. means credited one and therefore the better Moreover, that the histori- tion”). the district found again district contend that the 20. The pool omitted. See 5 was cant deviation Thus, analysis, analy- supra. accepted note 15 Haworth's court should not have credited, on a was not based district court also sample sizes too was based on sis because it by significantly utilized sample smaller than that meaningful data. yield statistical small Therefore, expert. the district plaintiffs' meaning- yield sample is too small Whether a clearly refusing that D err in to find did not court by made is a determination ful results regard discriminatory possessed intent with & L Watson, case-by-case on a basis. hiring practices. to its Rendon, 13; 2789 n. 108 S.Ct. at at 995 n. case, the the facts of this at 397. Under 883 F.2d analysis plaintiffs' that D & L showed 19. The accepting D & err in court did not district applicants and of white 50.8% hired 84.4% analysis. L’s applicants the MSES. referred anecdotal, cal, rejected plaintiffs’ ap- district court and other circumstantial evi- The proach being as overbroad because not ev- submitted did not dence eryone production in D & L’s work force was engaged pattern that D & L or establish qualified promotion. The district court plaintiffs, practice of discrimination. D L further found that because & had hired course, challenge these determinations. individuals from outside & L’s work force supervisors, plaintiffs’ to be foremen and the A approach too was narrow because it excluded production individuals outside & L’s work may employer that an A court infer qualified force who were to be leaders and pro in when engaged racial discrimination agree plaintiffs foremen.21 that We the statistics, moting comparing if when workers assumption made the forbidden that all D & pro non-whites and whites the number of employees equally qualified L for moted, gross demonstrate a statistical dis Lewis, (reject- motion. See 750 F.2d at 1275 Lewis, parity. F.2d at 1271. 750 Statistical ing plaintiffs’ analysis the because evidence, however, meaningful in must be brought comparison “employees into their light surrounding circum the facts and promotion”); are not to be considered for Pouncy, “In stances. 668 F.2d at 1275. Pouncy, (refusing 668 F.2d at 803 to infer establishing an inference of discrimination employer against that discriminated evidence, ‘required employees awarding in promotions simply from statistical the com plaintiffs parison qualified pool employees because the demonstrated the [is] to a ” percentage employees promoted of black ;presumptively eligible promotion.’ percentage “far less” than the in black em- Lewis, (quoting Pouncy, 750 F.2d at 1275 668 force). ployees employer’s in the work Con- 803). F.2d at sequently, the district court did not err in plaintiffs’ the statistical evidence for 1 promotions unpersuasive. leader and foreman The introduced statistical comparing percentage evidence the of leader plaintiffs also introduced statis promotions given and foreman to blacks dur showing tical evidence that blacks constituted ing per the —23%—to persons promoted 64% of to leader and fore centage employees pro of black in D & L’s positions during pre-Grizzard man the According duction work force —80%. to the post-suit periods, only pro but 23% of those plaintiffs, pattern these statistics indicated a pre-suit period. moted promotion in of discrimination of D & L note, however, also positions. to leader and foreman pre-Grizzard period, apparently blacks comparison contend that this received regard favorable treatment with appropriate D L because & selected leaders promotions: average whites worked on organization. and foremen from promotion within the 71.8 months before while blacks (60.7%) 21. The district court also found that even if it with the blue-collar and manufactur- proper (58.2%) otherwise, to use the number of blacks in D & rates. Stated if blacks benchmark, production L's work force as a represented supervision at & were in [D L] in prima evidence did not establish a facie case of percentage the same as in other industries in discriminatory conduct: County percentage Bolivar relative to the workforce, production the benchmark comparing percentage [W]hen of blacks in 25%, percentage would be or seventeen leader, foreman, supervisor classifica- Thus, points less than what it has been. even company percentage tions at the to the perspective plaintiffs, from the workforce, defen- production blacks the ratio of performance production dant's relative to its percentage supervisory [of] cate- blackfs] performance (42%) workforce far gories exceeds of oth- percentage of blacks in (80%) County companies. er Bolivar production higher workforce is much Excerpts proportion uphold than the Record at 91. As we similar of the Bolivar County figure grounds, percent supervi- of 19 black in court's determination on other we need workers, production compared sors of not reach this issue. supervisors county-wide. employer An can- also average only 43 months. See worked on liability escape of not under Title VII dem- the bias in favor Because of note infra. that, line, onstrating at the bottom minorities pre-Grizzard period, during the blacks supervisory capacities compare represented unhelpful found district court Cove, non-discriminatory levels. See Wards period with statistics from that statistics Lewis, 655-57, 2124; 109 S.Ct. at Con- periods. post-Grizzard from the Cf. Teal, 440, 450, (noting an em necticut v. n. 14 F.2d at 1275 (1982). 2525, 2532, Thus, program could 73 L.Ed.2d 130 action ployer’s affirmative comparison is between the number analyses disparate treat relevant skew statistical eases). promoted during pre-suit Griz- D L was not re blacks Because & ment years and the number of preferential treatment zard continue the quired to statistically promoted should have been in during pre-Griz the given employees black any of discrimination.23 v. the absence Steelworkers Web period, zard see United 2721, 2728, 193, 205-06, er, 99 S.Ct. 443 U.S. plaintiffs have demonstrated (1979), court thus 480 the district 61 L.Ed.2d promotions of all that blacks received 23% disregarding plaintiffs’ the in did not err years. during pre-suit At the the regarding pre-Griz- the statistical evidence time, constituted 19.1% of the same blacks' period. zard supervisors production-related occupations in court, County.24 in The district be Bolivar 3 in D represented were & L’s cause blacks statistically higher in a ratio than of discrim work force no evidence figure to conduct, accepted predicted,25 D inflated the 19.1% inatory the district i.e., evidence, statistically should have re compared blacks L’s statistical & 25%— promotions to leader and D L ceived 25% of supervisors of at & percentage the black the number of positions. foreman Because supervisors of count percentage the black promotions is not statis blacks who received that the district court y-wide.22 We believe tically from the number of blacks of black different comparing the total number erred in promotions,26 to receive expected D & L to the number black supervisors at Anderson, 575, at D & L. 470 U.S. S.Ct. constituted & L demonstrated that blacks 22. D leaders, foremen, reasonably supervisors district court concluded and 1512. The its 42% evidence, sufficiently years. figure this the to the Based on the similar percentage supervisors, especially the of black occupation district court found that when of D L supervisor positions at D & L "was in evidence to failed to introduce range beyond within but contrary. not well compared expected with would be what which, case, market, is in this the relevant labor period, D & all times the relevant 25. At supervision pro- percentage in of blacks in black, and the was more than L’s workforce 70% occupations.” duction approximately production workforce was 80% that the civil- Census data showed black. 1980 by employer a demonstration Counly was force in Bolivar 51.9% ian labor “bottom-line,” that, overrepre- black, manufacturing labor force was 58.2% work force still is relevant sented in its black, 60.7% and blue-collar labor force employer en- ultimate determination whether black. gaged pattern practice of discrimination. in a or - Hicks, Mary's Honor Center v. U.S. See St. -, suggested Applying the formula mathematical 2742, 2751, -., 125 L.Ed.2d 113 S.Ct. States, 433 School v. United in District Hazelwood (1993). (1977), 299, 2736, 53 L.Ed.2d 768 97 S.Ct. ex- the number of blacks the difference between challenge use the district court’s 24. The promotions the number pected to receive encompassed only figure, arguing that it of this promotions not statis- actually did receive production occupa- supervisors "precision” (total promotions): tically significant of 22 tions, ap- working occupations, such as metal boilermakers, metal work- prentices, and sheet = O-NP Number of S/D examination, however, & L's ers. On direct (1-P) Haworth, expert, that leaders Dr. testified VNP = Deviations Standard would be included in foremen at D & L S/D O = who re- number of blacks Actual figure. district court did 19.1%-black promotion ceived resolving credibility dispute in favor err (5th ries, Inc., 798, failing Cir.1982); court did not err find an infer 673 F.2d Lewis, regard D disparate treatment with 750 F.2d at 1276. ence promotion practices. Terrell v. & L’s See distinguishable This case is from our Police, University Sys. 792 F.2d Tex. however, prior holdings, D L because the & (5th Cir.1986) (“When judgment 1362n. 3 supervisors promoted who evaluated and em correct, may court is district be ployees were both black and white. As the appeal affirmed reasons other than found, district court below.”), given those or relied on cert. de subjectivity [w]hile the of the criteria of nied, attitude and (1987). commitment could result L.Ed.2d 997 disparity potential

racial motion, it B is clear from the record evidence that the individuals recommended or con- contend that because promotion sidered for included black indi- promotion L& based its decisions on viduals recommended white foremen criteria, subjective D & L could have failed supervisors, white individuals recom- promote discriminatory blacks for reasons. mended supervisors, foremen and rejected The district court plaintiffs’ ar by and individuals recommended both. gument. promotional system A that is based subjective upon 90; criteria is not Excerpts Mary’s “discriminato Record at St. Honor cf. — Indus., Inc., ry per Page Hicks, U.S.-,-, se.” v. U.S. 726 Ctr. v. (5th 1038, 1046 Cir.1984); 2742, 2751, (1993) F.2d see also Wat (noting L.Ed.2d 407 son, (“[A]n at 108 S.Ct. at 2786 hiring tending that evidence a case to employer’s policy leaving promotion negate finding deci a of discrimination includes sions to the unchecked hiring discretion of lower fact that officer is the same supervisors minority level should itself no group plaintiff). raise infer as the The district conduct.”). discriminatory However, ence of finding supported by court’s the evidence.27 “promotion systems utilizing subjective Accordingly, eval the district court did not err by supervisors” uations all white refusing can be subjective evi to find that compo- disparate dence of discrimination in promotion policy treat nent of D & L’s constituted Payne ment cases. v. Travenol Laborato- discriminatory evidence of intent.28 See Ber- N = West, employee. Number of workers who re supervi- Charlie another black sor, promotion ceived a Haney, recommended James a white em- P = Probability being pro of a ployee, promotion. black Taylor, for L.C. a black shift popula moted from the Strickland, relevant supervisor, recommended Chris a tion | employee, promotion white while Howard - (22 25%) = x Watson, -0.25 Watson recommended Vera H a black em- Tolbert, ployee, position. for a leader E.W. a 2.03 supervisor, black testified that he was involved in Lopez See also v. Laborers Int'l Union Local No. Daniels, promoting Randy employee a white who (5th Cir.1993); 987 F.2d 1215 n. 14 promotion only a was offered after six black Lewis, 750 F.2d 1274 n. 12. Because the employees part had turned it down. See also V.C deviations, difference is less than 2 standard it is infra. statistically significant. not We also note that if leaders, we use number of D & L 42%—the 28. substantially We also note that D & L in- foremen, supervisors and who were black—in holding creased the number of blacks salaried 25%, place of the difference between the number supervisor positions, plaintiffs which the exclude promoted expected of blacks (9.24) and the number from their claim that D & L discriminated (1.84 statistically significant still is not against employees regard pro- black with to deviations). Consequently, plain- standard promoted motions. Grizzard E.W. Tolbert from pase prove prima tiffs failed to a facie of discrim- supervisor position production manag- a shift er, ination. plant. the second-in-command at the Charlie supervisors Grizzard recommended to position his West then took over Tolbert’s as a shift Brown, Grant, they supervisor. promoted consider Harold Richard also Richard Holmes, Marilyn head, supervisor and Vera department Watson—all black Williams from Williams, employees promotion. super- Richard a Vera Watson to foreman and then to shift —for visor, supervisor, Brown, Lofton, black testified that he offered one and Harold Charlie promotion it, employees, rejected Sylvester to four black who super- Parker from foremen to salaried LaPresto, offering before it to Robert a white visors. LaPresto, employee, pro- nard, (holding an em- Robert white at 547 841 F.2d subjective standards being employed by ployer’s unwritten moted to after & leader discriminatorily ap- promotion were not However, approximately L for four months. plied). position first offered to four black employees rejected promoted it. D L who C Haney, employee, James another white introduced evidence him, only leader nine after but months seniority black em- of white and comparative employees after two black first declined the support their promotion to ployees prior to employees promotion. offer of black re- Six discriminating L argument that D & was jected promotion given employee to white by promoting less senior against blacks Daniels, Randy promoted whom D & L rejected this evi- The district whites.29 after leader less than two months his em- ignored that more senior dence because ployment began.31 We also note that down employees turned several black post-suit which blacks re- period, given to less senior white em- later motions promotions ceived 64% the to leader ployees.30 foreman and which the do about average length time be complain, average 154.9 blacks worked may proving promotions be relevant tween promotion months while before whites Pouncy, employer. See discrimination comparison worked 51 months. This agree we with 668 F.2d regarding demonstrates statistics court that the submitted statistics seniority by D promoted relative of workers fatally because flawed support disparate L & cannot overlook the Pouncy, F.2d at treatment. Cf. promotions later offered to down turned discrepancy (finding that between the a mean Trevino, at employees. F.2d white See 811 employees salaries of and white could black (finding plaintiffs’ that the statistics dem by “any explained number of nondiscrimi- be disparity onstrating “significant between factors,” natory including skill different lev- non-Hispanies the total Hispanics and for els, previous training, experience). Con- pro hire” to elapsed time from the date of court, sequently, we to like refuse persuasive compellingly “not evi motion was discrimination). pro- that D & L’s example, infer from this evidence For dence” recently- employees employees According plaintiffs, the senior black in favor of the white to Moreover, September employees.... plain- for D 1978 and 1982 worked hired black between promotion; average figures attempt 71.8 months which & L on before tiffs have chosen com- employees average periods pare pre- post-Grizzard on 43.2 for worked the black presuit promotion. During ignoring promotions, conveniently before the fact months that years, plaintiffs that whites given contend pre-Grizzard period blacks were pro- average before on 15.1 months system worked was not more treatment —the favorable average 84.8 and blacks worked average motion were on less senior neutral. Blacks note, however, using plaintiffs' that We months. twenty- counterparts by some than their white 144, apparently on aver- whites worked exhibit eight months.” promotion. While we age before 51 months Excerpts at Record 101-02. every recomputed that not number have brief, point plaintiffs we out this cite in their that D & L knew that 31.The contend problems discrepancy to demonstrate the turn down the employees- black would analysis. plaintiffs' with the statistical exist however, court, The district cred- motion offers. testimony & L's E.W. Tolbert—D ited the plain- district court also stated that 30. The manager employees production black —that tiffs’ evidence promotion were offered offers who turned down they multiple promotions because above- “paint[ed] picture attaches a distorted supervisors average employees felt and their responsibility to Mr. Grizzard offer, they might accept an offer of a new employee that a black before his arrival by job, someone else. or an offer made different giving being promoted instead of him cred- province to court was within its The district employees predeces- promoting it black his Anderson, credibility disputes, 470 U.S. at resolve perspective, the From this overlooked. sors 1512, 575, and the district court’s only way S.Ct. Grizzard could finds that the clearly finding bypass issue is not erroneous. on this passed was to muster continue have practices unlawfully argued supervisors motion disfavored black tiffs that D & L’s both employees. racially derogatory used language speak- employees physically to black abused them, any repercussions without upper from management. reviewing After the anecdotal The district court considered evi presented by plaintiffs, evidence the dis- put by dence forward with re trict court concluded that D & L at no time gard alleged individual instances of dis encouraged or condoned mistreatment of gave crimination and that D L concluded employees supervisors. Specifical- black legitimate why promote reasons did not ly, the district court found that most of the individuals at issue. The district promote court also found that D & L did not [t]he sum total of the regard evidence with simply the other individuals because allegations to these of what constitutes a were overlooked or did not come to the at claim of hostile environment the form of management.32 part tention of See VIII.C disparate treatment of black Thus, not, the anecdotal did evidence infra. amounts, best, proof of isolated acts assert, support as the vulgar racially derogatory use of or engaged pattern practice & L in a or language by no more than two or three Dept. discrimination. Texas See Commu company single officials and a instance Burdine, nity v. Affairs physical contact company between a offi- 1089, 1097, (1981) 67 L.Ed.2d 207 single cial and a employee, all occur- (“The may fact that a court think that the ring period years. over a of three or more employer misjudged qualifications of the abundantly What is also clear is that when applicants expose does not in itself him to reported appropriate company offi- Frank, liability....”); Title VII Odom v. cials, every complained of incident was im- (5th Cir.1993) (noting F.3d 849-50 mediately followed company affirmative perpetuation “good boy” of a old network apologies action the form of to the em- discrimination). does not amount to racial involved, ployees counseling manag- Consequently, the district court did not involved, ers discipline of those and/or in concluding plaintiffs’ err 'management Moreover, officials. anecdotal support evidence did not their plaintiffs’ allegations pe- relate to the time Lewis, disparate claim of treatment. See riod Manager after the arrival of Plant the (finding F.2d at 1276-77 that anecdotal evi Grizzard, James being with there not a supported finding only dence of isolated scintilla of evidence that partici- discrimination). instances of pated in any or condoned of the com- acts, plained exception possi- with the E bly vulgar two instance of the use of lan- plaintiffs which, further guage, contend that both undisputed under the circumstances, rejecting court erred in their the court finds to have been attempt justified, one, to buttress their statistical evidence about which the union regarding discrimination with evidence D grievance, filed a for which Mr. Grizzard history & L’s plain- apologized.33 of discrimination. The plaintiffs practice

32. The contend that the district court’s or of discrimination in D & L’s findings policies would D & L allow to discriminate motion existed. Id. against many employees without hindrance long Company as as the treated other blacks the upon 33. Plaintiff Daniel Anderson testified that employees. same as white examination, "this attack returning to work from a doctor's impact overstates the of the district court’s ... supervisor his refused to let him take a break Lewis, analysis.” 750 at F.2d 1274 n. 11. D & L that he had missed. Anderson then clocked out demonstrating legiti- introduced evidence that plant. and left the Grizzard confronted lay delayed promotions mate reasons behind the day, asking Anderson the next him "who in the relatively early pro- for some just blacks and the hell did he think was be he able to leave evidence, motions for some whites. This as the because he couldn’t take his break.” On the found, occasion, properly negate district court tended to second Grizzard used the term "son-of- plaintiffs' attempted showing pattern that a a-bitch" when he discovered that someone had

1295 The district VI Excerpts at 129-30. Record that some of the individu- further found court finally plaintiffs argue The class L officials used alleged that als finding the district court erred that & L language were not credi-

racially derogatory against giv did not discriminate blacks when ble. assignments gen ing temporary upgrade province of the district is posit It within and rack eral maintenance maintenance credibility disputes. resolve They ions.35 contend that the statistical evi 564, 105 Anderson, at S.Ct. at 1504. 470 U.S. gross disparity in dence demonstrated a Moreover, an ethnic “mere utterance of employees, treatment of white and black thus engenders offensive epithet which establishing prima or racial facie case of discriminat employee” not indicative of a feelings in an appear ion.36 It does practice any testimony of racial discrimination pattern presented expert explaining or EEOC, Rogers v. 454 significance Title VII. meaning violation of or statistical of this (5th denied, Cir.1971), 234, Consequently, cert. our F.2d evidence. review of the unduly hampered by 32 L.Ed.2d evidence been has (1972). undertaking independent plaintiffs’ failure to establish whether their After record, meaningful significant or conclude that the statistics were review of the we fight particular facts of this case. See that the court did not err district Teamsters, best, demonstrated, at 1857 isolated S.Ct. evidence (noting that racially [of statistics] acts insuffi “usefulness instances of motivated depends surrounding all of the facts and disparate treatm cient to establish class-wide circumstances”). (hold Nevertheless, Bernard, closely after F.2d at 569 ent.34 See evidence, reviewing plaintiffs’ we con testimony plain of the ing that the anecdotal failed to establish a class-wide clude tiffs’ witnesses did not establish treatment). prima facie case of discrimination. disparate company Watson was an "out- bulletin Tolbert stated that Howard a notice from the removed standing production manager” grievance regard- who also was not board. When the union filed incident, apolo- prejudiced. Grizzard issued a written gy explained were not direct- his comments positions and rack maintenance within earshot. 35. General ed to plant job two classifications at D & L’s positions. supervi- considered to be "craft” that two black salaried 34.We also note Tolbert —and a sors —Charlie West and E.W. following statisti- 36. The claim that O’Neal —testified black foreman —Monroe man,” prima good super plant established a facie case of "a cal evidence Grizzard was "a Moreover, manager,” prejudiced. discrimination: and was not Percentage Temporary Upgrades of Total Hours As- to Gi eral Maintenance: Table signed Black White Time Period % % 11.60% 88.40% Pre-Grizzard Period [1/15/81-10/2/82] 92.59% 7.40% Pre-Suit Period Grizzard [10/2/82-7/4/85] 46.60% 53.40% Period Post-Suit [7/5/85-9/19/85] Assigned Percentage Upgrades Temporary of Total Hours Maintenance: to Rack Table 7. White Black %

Time Period % 14.61% 85.40% Pre-Grizzard Period Pre-Suit Period Grizzard 0.00% 100.00% Period Post-Suit 100.00% 0.00% pre- undergirding these statistics presenting data these raw Brief for Plaintiffs statistics, only through joint stipu- plaintiffs that the statis- stated to the district court sented stipulated data.” [the] "drawn from tics were lation. searching, we discovered that After much fairly.40 employees, three of whom were treated him also the lack Eleven We note black, upgrades gen- temporary regarding received employees, evidence the number of jobs during the eral maintenance Grizzard white, temporary black and who received During period, pre-suit period. two upgrades positions in to craft non-mainte- Myers employees and James white Cove, departments. nance See Wards —William tempo- Partridge 68% the total 651-53, (noting at 2122 S.Ct. —received By comparison, rary-upgrade hours. two plaintiffs unnecessarily segment should not employees Triplett and Alfred black employer’s —James attempt work force in an Kemp, approximately 88% of Sr.—received discrimination). prove agree We therefore temporary-upgrade during pre- hours with district court’s conclusion that D & post-suit period, period; in the one L selecting employees did not discriminate in Haynes, employee white Jr. —and —Robert temporary upgrade assignments employee Perry one black —Arthur —re- general rack maintenance and maintenance approximately temporary- ceived 94% of all departments. upgrade plaintiffs hours. The introduced no demonstrating evidence that other individu- VII qualified temporary- als desired or were upgrade positions.37 Consequently, have we presented by The statistical evidence nothing compare plaintiffs’ with which to present gross does not statisti- result, statistics. As a the statistics are that, disparity cal light considered meaningless.38 record, the entire require would a reversal of the district findings court’s ultimate that no supports

Other evidence pattern of regard D L discrimination existed with court’s determination that & did not giving employees temporary promotion practices. discriminate in to & L’s upgrade assignments. example, Nor per- For is the historical and anecdotal evidence centage permanent posi- blacks craft introduced sup- sufficient to during years ranged tions the Grizzard from port disparate treatment. Ac- During immediately 66.7% to 73.7%. cordingly, we AFFIRM the district court’s preceding years, comprised 67.1% to judgment on the class claims. employees positions.39

63% the in craft Moreover, D & L the Grizzard VIII eight permanent awarded seven out of the judge rejected The district the indi promotions positions to craft to black em- vidual plaintiffs challenging claims of various ployees. Moreover, Kemp, Alfred one of the the treatment pursuant received to & temporary who received no hiring, promotion, L’s upgrades during poli pre-suit period termination disparate but cies. “In a pre-Grizzard did both the treatment case involv individual, post-suit periods, unequivocally plaintiff initially testified has the good place D & L always was a to work and showing prima burden of facie case of example, Kemp, 37. For did introduce evi- temporaiy up- Sr.—received 95% pertaining qualifications dence grade of Thomas hours. Anderson, Hardy temporary and Daniel Jr. for individuals, upgrade assignments. These two figures County 39. The 1980 census for Bolivar however, qualified assign- were not for such preci- indicate that blacks held of all 33.4% part ments. See VIII.D infra. production repair positions sion craft and coun- ty-wide. plaintiffs' regarding 38. The statistics the number temporary-upgrade of rack-maintenance hours Watson, approve Kemp's 40. Howard who had to by problems. worked blacks suffer from the same availability temporaiy assignments, supplied for example, employees given For out of the three legitimate, nondiscriminatoiy a reason for temporaiy upgrades pre-Grizzard period, Kemp's temporary upgrade assignments lack of employee ap- a white Mullen —worked —Willie during proximately pre-suit period. the Grizzard of the total Watson tes- 82% hours. In the Griz- pre-suit period, employee approve zard tified that he one white sometimes refused to —Ed- Kemp temporary ward upgrades Kemp Otto—worked the hours. In the because 100% post-suit period, employee production department. one black was needed in a —Alfred

1297 Carroll, A 708 F.2d at 195. fered reason was not the true reason for the discrimination.” decision,’ initial burden demon- plaintiff employment meets this and that race was.” — strating at-, Mary’s, St. U.S. 113 S.Ct. at (ii) (i) Burdine, belongs minority; 256, a racial (quoting that he to 2747 450 at U.S. 101 job 1095). qualified and for a applied that he was plaintiff at at all S.Ct. times seeking appli- employer was for which the persuading the burden of the court bears (iii) cants; that, despite qualifications, his that he or she was the victim of intentional (iv) that, — rejected; after his he was and at-, discrimination.41 Id. U.S. rejection, position open remained the at 2747-48. S.Ct. appli- employer continued to seek the possessed the persons [who from cations] A plaintiff’s] qualifications. A plaintiffs number of individual Green, Douglas v. Corp. McDonnell unlawfully D L claim that discriminated 1817, 1824, 792, 802, 36 L.Ed.2d 668 against by refusing them on the basis of race (1973). prima of the facie “Establishment give applications during to them presumption the creates a that case effect through April —October unlawfully against employer discriminated The district court found that some the Burdine, 254, employee.” the because, plaintiffs were not credible plaintiff If 101 S.Ct. at 1094. establishes they allegedly ap the time could not obtain discrimination, “the prima facie case of plications, given applicat other blacks were employer to rebut burden shifts Furthermore, ions.42 by producing presumption of discrimination properly give D L found that & refused to rejected, plaintiff that the or evidence many plaintiffs applications because legitimate, preferred, for a someone else Company either was not when Carroll, non-discriminatory reason.” seeking applications or was employer If F.2d at 195. “carries already hiring but had received a sufficient production, presumption raised burden of applications positions number for the to be prima facie case is rebutted” findings filled. district court’s are not Burdine, drops ease. 450 U.S. at from the An extensive review of erroneous. 10, 255 & n. 101 S.Ct. at 1095 & n. 10. “The contains the record has convinced us plaintiff opportuni- full and fair then has ‘the indicating demonstrate,’ no evidence ty through presentation of given applications of their through were not because his own case and cross-examination witnesses, supporting the prof- race.43 The evidence of the defendant’s ‘that the fully while Malcolm Smith testified that he ob- a case has tried on the 41. "When been merits, application adequacy party’s showing 1982 and was tained an October of a Williams, Kimroy D & stage Douglas next particular ritual- hired the month. of the McDonnell inquiiy employee, that three of his relatives unimportant. on wheth- L testified We focus our employment upon applied with D & L record contains evidence er the years. of fact could have concluded as reasonable trier Constructors, juty did.” Molnar v. Ebasco (citation Inc., (5th Cir.1993) 986 F.2d applica- complain 43.The also omitted). by Haynes she procedures instituted tion appli- personnel manager making all became — plant, refusing to allow em- plaintiff cants come to the example, Chris Harris testified 42. For applications ployees home with to take blank were able to obtain that she and two other blacks potential to harm them—were intended applications October Plaintiff Linda applicants. D & L demonstrated although she was not able Davis testified nondiscriminatory legitimate, reasons for application, it had procure her sister and another process. example, changing application For person applications. Bobbie obtained Plaintiff employ- Haynes allowed testified that when she Sharp was not able to testified that while she mid-1984, plant, many applications from the application ees to take she did ob- obtain an mid-1983; that mak- She also stated were never returned. application from D & L in tain an occasion, plant potential employees come to Sharp testified that on the latter also helped applications complete en- filling applications. obtain other blacks out she saw actually people applying wanted the Hardy, employee, that sure that the a D & L testified Thomas plant, job, transportation to and had from procured application from D & L in his wife plaintiffs’ allegations significantly, that D & L refused to More it was uncontradieted. As Odom adduced no give employment applications is the evidence favorable to ..., position compelled accept his we are plaintiffs’ unsupported concerning assertions ex-planation_”); legitimate, non-discriminatory Strader’s collective belief that D & L discrimi their — Mary’s, St. against appli D nated blacks.44 Under & L’s cf . 5, 113 (“The -n. S.Ct. at 2751 n. 5 notion part supra, policy, cation see I.A there were every employer keeps that ‘per reasonable nondiscriminatory why legitimate, reasons D people sonnel records’ on who never became give applications plaintiffs. L & did not personnel, showing why they did not become Because the failed to demonstrate (ie., personnel respects in what all other that race was the true reason for D & L’s better) people who were hired were seems give applications, uphold refusal to them we highly us fanciful —or for the sake of Ameri finding the district court’s that D & L did not is.”). Likewise, hope can business we against plaintiffs. discriminate See Sharp failed to adduce evidence indicat Frank, (5th Odom v. 3 F.3d Cir. unlawfully & L considered her 1993) (holding that anecdotal evidence and deciding race when not to hire her. More speculation support finding mere will not over, at Sharp the same time Harris and employer “poli that an followed an unwritten hired, were not D L& hired other blacks as discrimination); Carroll, cy” of 708 F.2d at production Thus, workers. (noting suspicion 195-96 the mere finding court’s that D L& did not discrimi discrimination not is sufficient to establish against clearly nate her is not erroneous. treatment). disparate Carroll, (noting See F.2d that the suspicion mere of discrimination does “not B treatment”). disparate establish Harris and Chris Bobbie Plaintiffs Cannon, Plaintiffs Nathaniel Michael Sharp unlawfully contend that D & L dis O’Neal, Gregory Townsend contend that against criminated them on the basis of race unlawfully D & L against discriminated them by refusing production posi to hire them for by refusing to rehire them after had tions. The district court found that D & L laid been off.45 The district court found that valid, Sharp did not hire Harris and race- these adduced no evidence indicat Harris, example, neutral For reasons. who ing that race was a factor in D & L’s decision applied hiring only when D L& was for the fact, not to rehire them. the record indi shift, night application indicated on her twenty-two blacks, cates that D L& rehired she evening attended afternoon and classes whites, fourteen the Griz- college. Haynes at a local testified that she years. zard We hold that the district court believed she did not hire Harris because of finding did not err in & L did not the conflict between Harris’s classes and the Cannon, O’Neal, rehire and Townsend be hiring. shift for which she was The district legitimate, cause of race-neutral reasons. although Haynes’s explana Carroll, court noted that See (noting 708 F.2d at 195 absolute, tion was not suggested suspicion no evidence mere of discrimination does “not treatment”). disparate establish Haynes’s that race affected decision not to finding clearly hire Harris. This is not erro C Odom, (“Although neous. See 3 F.3d at 845 testimony ... Strader’s was less than Coleman, abso Plaintiffs Moses Richard lute, Grant, Hall, Melvin, it constituted at least some evidence. Earnest Robert Alexan- qualified could read and write. The Dhe] district court to work [at L] D & but ... testimony, Moreover, credited her application.” and this didn't even receive an plaintiffs uniformly erroneous. testified that neither Haynes security guards Dnor & L’s treated them derogatory in a harsh manner or directed or example, although Slaughter 44. For Miller testi- vulgar language at them. discriminating," fied that D & L did “a lot of she company large would not consider a that has a Cannon, O'Neal, percentage acting of black to be dis- and Townsend were not enti- criminatorily. Melvin Shaw testified that he tled to be recalled under the terms of D & L's against simply bargaining agreement. discriminated [he] “because knew collective steward, Smith, Kimroy claim that D union testified that and Williams Smith had an der L, promote them to leaders failing problem. to & attitude foremen, against them on discriminated and failed to demonstrate that The district court of their race. account played any their race role in D & L’s failure plaintiffs failed to demonstrate found Carroll, promote to them. See 708 F.2d at played any part in D & L’s failure that race (rejecting employee a claim that an was supported This promote to them. diseriminatorily promotion denied because evidence. the record supervisors his testified that his attitude and foremen, and D & L selecting leaders dependability lack of were the actual rea- attitude, employee’s work rec- considered sons); Pouncy, (rejecting 668 F.2d at 795 an abilities, ord, leadership experience, work employee’s promotion claim because the em- employees,

willingness help other commit- ployee possess did not the initiative or the Company, seniority to be and ment ability to communicate with and train his co- very important. record reflects that workers). most, plaintiffs proved At lacking was in one or more of plaintiff each L may D & have overlooked one or more of years. these attributes considering potential them when candidates U.S.A, Inc., 14 F.3d Davis v. Chevron See promotion. showing for a is insufficient Such (5th Cir.1994) (noting that to es- Odom, prove discrimination. F.3d Cf. discrimination, prima a facie case of tablish (noting perpetuation 849-50 aof sought for the applicant qualified must be “good boy” old network does not amount to it). applies for position at the time he or she discrimination). Accordingly, racial Bardwell, example, For Gerald Coleman’s su- prove failed to that D & L did not Coleman, tempo- pervisor, stated promote them on account of their race. foremen, rarily filling in for was unable to job employees “goofing had handle the him.” testified that

off on Howard Watson supervisors department never recommended Additionally, promotion. for both Coleman Anderson, Tommy Daniel Jr. and previ- and Hall had both turned down Grant Hardy argue give that D & L refused positions, promotion ous offers of to leader temporary upgrade assignments them indicating a lack of commitment to & L.46 general maintenance or rack maintenance de West, supervisor, a black testified Charlie partments Welding because of their race. that he did not consider Grant for one experience necessary many By- not know motion because he did Grant. performed by general maintenance em tasks Kyle, department superintendent, a testi- ron ployees, welding primary duty any leadership fied that Grant did not have employees. rack maintenance The district Williams, super- Richard a black abilities.47 give L court found that D & did not visor, Howard testified that Mel- Watson Hardy temporary upgrade as Anderson and Kimroy vin and did show Williams necessary signments they because lacked the being promoted only and did interest experience to work in the maintenance de enough get by, thereby demonstrat- work partments. ing little desire for advancement and no com- Lucas, Hardy Both Anderson and testified Company.48 Tommy mitment to the welding department had a limited amount of paint foreman in the and former Moreover, challenges admitted that he was not D & L's that he Grant 46.Hall assertion However, qualified promotion until 1985. he rejected promotion. an offer of charge alleging in 1984 that D Watson, filed an EEOC production Grizzard and Howard discriminatory promote him for rea- L failed to sons. Grant ing time, manager testified that Hall did at the both inconsistency by explained stat- promotion promotion turn down a because charge read the EEOC before that he did not required have Hall to work on the second would signed had no idea what the he it and thus testimony district credited this shift. The charge alleged. was not over that of Hall. This determination Anderson, 470 U.S. at erroneous. See change in Williams's 48. After D & L noticed 574, 105 S.Ct. at 1511-12. Company promoted him. attitude Moreover, they presented preponderance reject- experience.49 no of the evidence that it demonstrating legitimate that what little weld- evidence ed Anderson for reasons. Accord- they ing experience qualified had them to ingly, properly the district court found that general perform the duties of and rack main- Anderson failed demonstrate that he was Finally, they employees. tenance introduced the victim of discrimination. suggesting

no evidence white em- ployee upgrade unquali- an who received E (or qualified plaintiffs). than fied less finding Because the district court’s Cox, Four —Belinda Hardy qualified Anderson and were not III, Haywood Synovia Robert and Jackson clearly temporary upgrades receive is not Jesse Overstreet —contend that erroneous, district court’s conclusion court erred in that D L& did not Hardy Anderson and failed to establish a discharge employees garnish under its prima sup- facie ease discrimination is policy, part supra, ment see I.D in retaliation by ported the evidence. filing charges.52 To EEOC establish a prima retaliatory discharge facie case of un Anderson further contends that 2000e-3(a) 42,53 § against plaintiff L der of Title by & discriminated him must (1) permanently promoting engaged activity demonstrate that he him to the rack (2) VII, protected by department. maintenance Title that an adverse Anderson and (3) Bruce, employment occurred, action employee, James a white that a submitted permanent participation pro causal link posi bids for rack between maintenance activity employment tion. Both tected and the Anderson and Bruce were adverse inter First, Shirley Chrysler decision viewed. The exists. v. activity ed because D & L discharge did not by arguing D L counters that fired. & there employees other garnish who violated the plaintiffs’ is no causal connection between the policy employees ment if those or their rela protected terminations their activities. tives had not charges. filed EEOC argues plain- D L if & further that even the plaintiffs point to seven individuals whom prima tiffs did establish facie case retali- they argue discharged should have been un ation, legitimate, it articulated nondiserimi- policy der the but were not. Six of the natory discharges for their reason —viola- however, plaintiffs, individuals named the garnishment policy. tions of the garnishment did not policy.56 violate the Moreover, plaintiffs The district court found that the the mere fact that one individual policy fired, failed to show a causal connection being between violated the without does filing charges the of the and their termi- not itself amount to race-based discrimi Lewis, nations. The district court further found nation. See 750 F.2d at 1278 n. 19 legitimate, (finding that the evidence demonstrated irregular exceptions that to a valid nondiscriminatory plaintiffs’ policy claim); reasons for the support did not a retaliation cf. Frank, (5th garnishment terminations —violations of the Odom v. 3 F.3d Cir. 1993)

policy. findings clearly “misfeasance, (noting malfeasance, These are not erro- undisputed plaintiffs neous. It is the or nonfeasance —without nexus to ... race— actionable”). garnishment policy.55 Furthermore, did fact violate the is not contrary penalty contention, plaintiffs’ Because the for such a violation was the evidence in termination, D L properly & terminated the dicates that D L discharged under the (hold- Lewis, plaintiffs. garnishment policy See F.2d at 1279 employees other black ing that when some evidence substantiates a who had not filed EEOC claims.57 Addition ally, claim of plaintiffs presented retaliation while other evidence dem- sug no evidence employer gesting onstrates that the for inno- discharge acted that & L did not white Haywood charges, thirty-day grace period 54. did not file EEOC but his because the for the fourth garnishment father did. beyond extended over two garnishment; date that received their first policy Williams did not violate the his because argues that D & Jackson L terminated her one garnishment fourth was released before the end day 30-day grace period before the for her fourth thirty-day grace period; Stapleton filed however, garnishment expired. Haynes, testified bankruptcy petition, barred D & L from grace period monthly — n that she measured the her; counting garnishments against and Wes- i.e., employee garnishment if an was notified of a ley garnishment policy did not violate D & L's month, day grace period on the fifth of one Haynes garnish- either because did not count expired day following on the sixth month. personnel ments received before she became Thus, grace period usually days, while was 30 manager Haynes simply misapplied or because actually in some instances was either 29 or 31 record, reviewing policy. After we have days. Because Jackson was notified of her findings determined that these garnishment February fourth 1985 and Moreover, erroneous. Lofton was the sister of totally fired on March her termination was Melvin, only charges Robert who not filed EEOC customary application consistent with D & L's plaintiff but was a named in this suit. fact This garnishment policy. wholly plaintiffs' inconsistent with the conten- discharged garnish- tion that & L under the D & contend that L would have employees policy ment those black who filed Daniels, Drake, discharged Larry Charles Hazel charges had EEOC themselves or close relatives Lofton, Stapleton, Larry Wesley, Alice and Eddie do so. garnishment policy ap- Williams had the been plied nondiscriminatorily. the district Archie, discharged 57.D & L Lester James Hum- poli- court found that Daniels did not violate the phrey, cy Haynes, applying policy, and Jimmie James for violations of the because did Moreover, garnishment employees policy. garnishments D & L dis- not count received Drake, charged Humphrey day personnel manager; on the same that it dis- before she became Lofton, policy charged plaintiff and Williams did not violate the Belinda Cox. intent, if maintained cannot be in terms of policy. Conse- who violated quo status they operate to ‘freeze’ the not err in court did the district quently, employment practices.” discriminatory prior link between causal existed finding that no 849, 853, 429-30, charges and the EEOC filing of the Court, “good According to the L.Ed.2d 158. uphold discharges. We therefore plaintiffs’ discriminatory intent intent or absence D L did not finding that & court’s employment procedures or does not redeem for their in retaliation discharge operate as ‘built-in testing mechanisms activity. protected participation minority groups and are unre- headwinds’ measuring job capability.” Id. at lated IX 432, 91 S.Ct. at 854. produce failed to sufficient proof require elucidated the The Court engaged demonstrating D & L evidence in disparate impact cases Wards ments According- discrimination. race-based Atonio, Co., Packing Inc. v. Cove judgment of the district ly, AFFIRM the we (1989). 104 L.Ed.2d respects. plaintiff Title VII to that for a It ruled claim, disparate impact successfully present a JOHNSON, dissenting in Judge, Circuit employ point particular he or she must to part: dispa such practice that has created ment there show The facts here 657, 109 impact. at 2124-25. rate Id. S.Ct. have scores of individuals who have been alleged plaintiffs in Cove that the Wards at the behest of racial discrimination endured alia, were, nepo practices question inter (“D L”). Many of Douglas and Lomason tism, separate hiring chan the creation of pressured by the have been these individuals nels, preferential treatment in practice of company drop to this cause president of the subjective rehiring, and the use of decision company.” good “for the of the of action Supreme Court concluded making. Id. The company president’s bowing Instead presented a properly had laying closing plant and off threats of by pointing disparate impact claim to such lawsuit, people of this hundreds of because which, support a practices proved, if would up they stood for what have Id. finding of racial discrimination. rights equal employ- to their to believed be plethora which Despite a of evidence ment. Alleged Discriminatory Employment Plain- overwhelmingly support the seems to Practices contentions, court found that the district tiffs case, court con- In the instant the district occurred. The ma- no racial discrimination identify Plaintiffs had failed to cluded that carefully reviewing jority After affirms. practices ad- employment of D&L’s which record, to dissent. this writer is constrained versely impacted African Americans. The did, in majority the Plaintiffs asserts Hiring Impact in Disparate A. fact, practice identify employment an —name- requires applicants Law to ly, policy D&L’s which complete applications plant. at the place begin to an appropriate The most majority practice asserts that such disparate impact claims is the law. analysis of justify disparate im- could not resort to case, Griggs impact v. disparate The seminal theory. Maj. op. at 1284. pact Co., Congress’ explained that Duke Power belies the conclusions both enacting The record objective in Title VII was majority court. The and the district employment opportuni- equality “achieve court, order, by signed operated pretrial barriers that have ties and remove party, and filed group agreed to counsel for each past to favor identifiable trial, just prior to the states employees.” 401 over one month employees over other white (1971). allege that the defendant plainly, “Plaintiffs The Court went on 429-30 against blacks in company has discriminated “practices, procedures, or to determine that applica- face, by restricting access to hiring, both and even neutral tests neutral their by failing refusing to hire tion forms and D&L’s—and Plaintiffs’ —contention that Haynes permitted practice are to fill out Ms. of restricting those blacks who access to application application changes on the same basis as whites forms caused the forms black application plant fill out forms.” R. at 2590. hires at the completely Cleveland application Restricting access to forms and substantiated the record.2 D&L and blacks, using different standards Plaintiffs introduced into evidence dozens of whites, opposed employment depositions as both black individuals who had at- justify obtain, “practices” application tempted refused, but who were theory disparate impact applications to the facts of from company D&L when the hiring. Many deponents this case. identi- family fied black friends members and/or opportunity who were also denied the to file 3. Anecdotal Evidence all, applications.3 In Plaintiffs identified fifty more than blacks to whom D&L refused majority “found” that the “record con- applications. majority peo- The vast of these indicating ap- tains no evidence D&L’s ple attempted applications to obtain numer- procedures potential plication affected Many ous times. of them asserted that differently applicants any potential than discouraged stopped trying became Maj. applicants.” op. white at 1285-86. Further, applications. obtain there is no dis-

Reading the record reveals otherwise. The pute great that sought number of them to quite readily appli- record shows that D&L’s apply jobs company at D&L when exclusively the procedures cation almost affected evidence, Haworth, hiring. Reviewing this expert, blacks. D&L’s Dr. Joan the majority properly claims that hiring practices by that initiated “the black indi- testified away viduals Patty Háynes specifically gate by turned at the restriction D&L — guards application give any applications never saw D&L forms —caused “wild fluctuations” percent applicants seeking employment.” Maj. op. in the of black who were to whites added). “hung However, (emphasis hired.1 The defense its hat” on this 1285-86 that explanation. argument In probably provides its to the district little comfort to motion, during only court its Rule D&L’s attor- the blacks who testified not that ney Patty Haynes’ change obtaining applications, contended that in saw whites but also hiring procedures enough being alone was to saw whites hired at the same time that change proportion being away person- cause the of black blacks were turned in the pre-Grizzard opposed pate.4 hires from the era. nel office—as Evans, man, example, 1. Prior to James Grizzard's arrival at the Cleve- 4.For J.C. stated black general plant, land of those hired for 64.6% Bobby Jolly, that he followed a white school factory jobs at D&L were black. When Mr. mate, personnel Patty into the office. He saw plant manager, Grizzard became that number Haynes talking Jolly group approxi- and a left, dropped to After Mr. Grizzard 46.5%. mately ten other white individuals. Mr. Evans black hires rose 61.1%. Haynes could not Ms. the white hear what told However, applicants. Jolly leave as started to recognized Patty Even the district court that area, passed he Evans and informed Evans Haynes' employment practices may have ad- he, Patty Haynes versely discussing Jolly, affected blacks. the fact had been hired. told signif- percentage dropped that the icantly during of black hires day. Evans to come back the next He left with- years, compared the Grizzard as application, out an when he returned the concluded, pre-Grizzard years, with the following day pursuant Haynes' to Ms. instruc- "Finally, plaintiffs' analysis in this instance fails tions, company he was told take into consideration difference hiring. Haynes practices procedures initiated Ms. during which were in effect Likewise, White, woman, Ms. Lizzie a substantially and differed from what was done explained that she visited the in an affidavit predecessor." Op. her at 18-19. personnel office at the same time that Teresa white, Boswell, acquaintance visited an who was correctly recognized 3. The district court only an the office. Ms. Boswell not obtained hearsay Howev- these statements are in nature. day. application, but she was also hired same er, neglected recognize that D&L never away applica- without an Ms. White was turned objected depositions to the introduction of the on that basis or on basis. tion. certainly that she started to retain all of the deposition think —and the law would One applications the restriction after she learned that some em- requires applications —that equally. whites impact on charges would ployees filed EEOC October compared case here. As far from the such is away explained that she threw 1984. She score blacks who and one-half to the two just prior applications to that number of presented applications, D&L were refused time, she also and she was not sure whether resounding number of three whites away applications some of the threw applications. Although refused had been preceding the five months were filed applications to obtain sought dozens of blacks filing charges. of the EEOC hiring, not one of the company was while the Statistics, dispute, sought applica- which D&L did not con- applicants potential white periods. Haynes tion such that Ms. failed to retain all of firm testified, Haynes applications. Ms. Evidence Statistical found, she, gener- the district court as a Law

a. The rule, appli- accept al would twelve to fifteen fairly per cations seven hires. This was a usually focus- disparate impact model practice, adopted constant and once she analysis. v. Fort on statistical Watson es Trust, 977, 987, practice, she did not deviate therefrom Bank and Worth 2784-85, 2777, throughout personnel manager. 101 L.Ed.2d 827 her tenure as (1988). probative, howev- Statistical data is application data which exists after the er, compares properly (October 1984) if it relevant ele- charges EEOC were filed District v. ments. See Hazelwood School Haynes’ explanation are consistent with Ms. States, United S.Ct. hand, practice: applications of her Of the (1977). “proper comparison L.Ed.2d 768 forty percent ap- almost came from unhired composition [the racial at- [is] between the Hence, hires, plicants. every six Ms. jobs] composition of the issue and the racial Haynes applicants. was able to ten consider qualified population ... in the relevant labor application period information for the Cove, 651, at market.” 490 U.S. 109 Wards prior filing charges of the EEOC Hazelwood, 433 U.S. at (quoting S.Ct. at 2121 vastly presents picture. different Of the 2741-42). 308, However, 97 at where S.Ct. applications prior on hand which were filed impossible to labor market statistics are as- 17, 1984, only twenty-two percent to October certain, Supreme recognized Court has by applicants. Contrary were filed unhired mea- that “certain other statistics —such as Haynes’ explication practice to Ms. of her of indicating composition of sures the racial accepting applications per twelve to fifteen ‘otherwise-qualified applicants’ for at-issue hires, contrary to the district seven jobs pur- equally probative for this —are acceptance explanation, court’s of that Ms. pose.” Id. Haynes required would have to hire been people every applicants four out of five b. Facts throughout period. agree comparison All in this best composition would be the racial of the case applications The difference in the is not factory applicant general flow into D&L for substantial, only statistically but it is also jobs compared composition racial as with the Mann, statistician, significant. Dr. Charles general factory of The district hires. dispute testified —and D&L did not all —that comparison possi- court found that such a things being equal according to Ms. —and majority ble. The error in found no clear Haynes, things equal were —there finding. wrong. Both are only four chances in one thousand that all of applications existed.5 Dr. Mann’s calcu- filing Fewer than two of after lawsuit, Patty Haynes Haynes’ uncertainty testified lation reveals that Ms. Notably, applications when she testified in a sex termined the existence of all for a discrimina- case, prior to Dr. period tion case the trial of this certain of time. performed analysis a similar which de- Haworth

1305 applications people might of all of the discouraged about her retention ified be ap- from plying” was well-founded.6 alleged discriminatory because of the Cove, see Wards practices); 653, U.S. applications Even if all of the were avail- (“As long as there are S.Ct. 2122-23 they elearly are on able —and not —reliance practices no barriers or deterring qualified applications inappropriate would be applying nonwhites ... if percent- light company dispropor- of the fact that the from age applicants of selected who are nonwhite tionately potential appli- restricted significantly is not less than percentage to the forms. cants’ access Barriers and/or qualified nonwhite, applicants who'are practices qualified which deter minorities employer’s jobs probably selection mechanism applying impermissibly from taints operate disparate impact does not with a analysis employs the use of actual on Rawlinson, added)). v. (emphasis minorities.” Dothard applicant-flow light data. 330, 2720, attempts 53 D&L’s restrictions S.Ct. blacks’ (1977) (“The application process applications,

L.Ed.2d 786 obtain applicant-flow true might adequately itself not reflect actual information cannot form the basis of the potential applicant pool, qual- Hence, since analysis otherwise statistical in this case.7 Haworth, personal knowledge proportion 6. Dr. who had no applicant increase the black of the applications, of the reasoning retention was the logic flow. The court's has no basis in applications witness to claim that all of the exist- or in the record. spoke great ed. she with diffidence. Initially importantly, Supreme Court See, (“For e.g., Trans. Vol. 22 at 1714-15 supply explana- has discommended courts’ 4,May through period January, [of of time disparities expla- tions for statistical when those applications I have the 1985] the nonhired support nations neither have in the record nor people applications people, and the of the hired judicial presumptions. constitute valid See infra as far as I can tell to the extent that exist.” 13; Partida, note see also Castaneda v. 430 U.S. added)); (”[T]hese (emphasis Id. at 1718 are the 482, 500, (1977). 97 S.Ct. 51 L.Ed.2d 498 pools for which I believe I have the nonhired Speculative suppositions prof- inferences and applications they’re extent that available." courts, opposed parties, fered as are im- added)). (emphasis proper impermissible. Mayor Philadel- phia Equality League, v. Educational analysis 7. Dr. Haworth's statistical is infirm for a 605, 619, 621, 1323, 1332-33, 1333-34, reasons, First, number other as well. (1974). 39 L.Ed.2d 630 counsel, cross-examined Plaintiffs’ she admit- case, explication rejec- In this the court’s of its applications ted that she excluded 11 which were tion of the October-November data finds abso- (Failing application undated. to date an was not lutely support no in the record. There is no Patty Haynes people fatal. hired several who suggestion, testimony, no no intimation in the forgot place application.) the date on their remotely implies record which that numerous applications Nine of those came from blacks. applied jobs appli- blacks for D&L to affect Additionally, applica- Dr. Haworth included the suggestion cant-flow data. There is not even a person sought general a tion of white who a greater proportion a of blacks than normal position, general maintenance mechanic not applied general factory positions for in October factory position position at issue here. Fi- —the fact, specifically 1984. In about informed nally, analysis hiring in her statistical of D&L’s applications 1984, the number of filed blacks and practices for Dr. Haworth excluded the whites in October 1984 and her selection rate of hiring period. October-November She claimed applicants, Haynes during those Ms. testified her applicants period the white in that had deposition that she knew no reason other greater manufacturing experi- blue collar and/or disparity hiring. than race for the in Mr. Griz- applicants. Patty Haynes, ence than the black however, similarly deposition plant zard testified in his that he prior production testified that why hiring knew no experience applicants of reason rates for Octo- none of the white —which possessed manufacturing ber were so favorable toward whites. Dr. blue collar or ex- —not investigation perience, seeking applicants Haworth testified that the EEOC of was beneficial for general applicant- factory D&L did not the October work at D&L. The district court make specifically adopt did not this rationale. It was correct in so flow data unreliable. She stated that doing. those numbers should not be discarded because erred, however, brought the EEOC had been into D&L The district court in substitut- Finally, matters. Dr. Haworth testified its own rationale for the exclusion of the wildly” applicant-flow did not "fluctuate applications. data in October-November The court rea- blacks, availability according began 1984. The of soned that because blacks to file EEOC Haworth, charges constant. of racial discrimination in October Dr. was explication rejection community conspired the black to flood D&L The district court's for his applications applications with black individuals so as to of the October 1984 once —that analysis “pools” appro- worth’s statistical of her have resorted to court should district information, Supreme proxy applications, pool which such was smaller priate each “equally probative.” to be has declared Court in pool than the this first standard.11 Cove, 490 U.S. at Wards rejection district court’s of this standard 2121-22. directly acceptance Dr. antithetical to its view, pools. it is Haworth’s this writer’s seven different stan- proffered Plaintiffs Furthermore, proxy this stan- availability served as a erroneous. dards information. applicant-flow provides actual barometer of the for dard rehable particularly were Four of those standards general applicant during peri- D&L flow 1) Employ- Mississippi State persuasive: Indeed, Haworth, question. her- ods in Dr. (“MSES”) referrals to D&L for System ment standard, self, though this not testified 2) fully registered persons with perfect, proportion was valid.12 The of black 3) MSES,8 general factory the number of applicants in standard was 65.6%. When three-quarter hires the four compared with the actual number blacks 4) arrival, and preceding Mr. Grizzard’s hired, signifi- statistical the standard showed County people Bolivar who are number cance at standard 8.8 deviations.13 fabricators, employed operators, as D&L, rejected stan- The district court the second other than as laborers factories dard, “heavily claiming that this standard is shown the 1980 census.9 weighted” in favor of blacks because of their appli- The first standard contains all of the over-utilization of MSES. Dr. Haworth tes- applicants cation forms of all interested D&L study tified that a national indicated that period in the time issue which were large generally urban areas on file with This standard is untaint- MSES. likely public employment more to use servic- ability by unfair on blacks’ ed restrictions testimony tempered es than whites. her She applications.10 obtain such The district court with the caveat that she did not know wheth- rejected “an this standard because study applicable er this to Bolivar Coun- extremely sample.” Interestingly small however, enough, ty, the court credited Dr. Ha- which is small and rural. Dr. Haworth charges availability. My blacks heard of the of discrimination at second and third standards of D&L, they apply positions flocked to with the third and fourth standards correlate to their sixth illogical. Among things company' also other and seventh standards. —is explanation the district court’s assumes that the applicants' people applied jobs — According deponents, to some hlack MSES require high which did even school edu- permit applications. would not them to file D&L proffering cation—understood statistical restrictions, though appar- the MSES case, proof disparate impact in a would blacks, are, ently only affecting way, in no as greater minority presence want to have a pervasive as D&L’s restrictions. court, applicant According to flow. Robinson, stipu- parties Oliver a man whom both applicants 11. MSES referred How- D&L. paranoid schizophrenic, encouraged lated was a ever, largest application pool which Dr. Ha- apply applicant blacks to so as to skew the flow. analyzed applicants. worth contained 80 unsupported suppositions way These much, assume too largest pool just applicants. second contained 57 clear, my and in view constitute indeed pools, descending The other in order contained patent, error. 48, 43, 42, applicants. and 18 any 8. This standard excluded individuals Although including she had trouble certain 12. younger any were than 18 old and who had ("OJT’s”) pool, MSES referrals Dr. Ha- might imply one of seven characteristics which recommending worth testified she was not they genuinely were not interested in em- OJT referrals be excluded from this ployment. they Individuals were excluded if re- standard. Stamps, they ceived welfare benefits or Food if applying Stamps, they if were in were for Food Project program, the WIN or Summer Youth Supreme 13. Court has reiterated numerous they migrant previously if were workers or had greater times that standard deviations than two worker, migrant been coded as a or if were suggest suspect activity. or three See Castaneda receiving unemployment Partida, 482, 17, 1272, insurance. v. 496 n. 97 S.Ct. 17, Hazelwood, (1977); 1281 n. 51 L.Ed.2d 498 299, 17, 2736, My U.S. 312 2743-44 9. first and second relevant standards corre- n. 17, (1977). spond to the Plaintiffs’ and the district court’s n. 53 L.Ed.2d 768 rejecting registrant in the us- the overall claimed that the difference MSES never This, age public employment services between information. the second standard of statistically significant was availability, blacks and whites reveals that 65% of the MSES did, on the other or even substantial. She registrants eighteen who were over and had hand, testify might be a there none of the earlier-described characteristics difference, difference, in trivial or even no just percent six-tenths of one black — in usage of MSES blacks and whites proportion than the in smaller of blacks County. Bolivar first standard. This standard demonstrates significance statistical at 8.5 standard devia- Bendick, Plaintiff’s Dr. Mark labor econo- tions. mist, study usage that a on the testified southern, agencies in public employment ru- in The district ruled that the third County, including Mis- ral Sunflower areas — availability, relevant standard of Plaintiffs county adjacent sissippi, a rural which is to failed “to take possi- into account the distinct County public that the em- Bolivar —revealed bility hiring prior period rate southern, usage in ployment service rural 1982, October, beginning may indeed have quite usage different from such in areas is discriminatory appli- been in favor of black study large urban areas. While the did not added). Op. (emphasis Again, cants.” at 18 compare usage employ- blacks’ and whites’ improperly supplied the district court rea- areas, in Dr. ment services rural Bendick ignoring analy- sons for Plaintiffs’ statistical study testified that no shows that rural party espoused no sis which time unemployment services at a blacks use dis- throughout during depositions. the trial or higher than proportionately rate rural supra See note 13. The district court had no evidence

whites. explained The court further that the differ- contrary testimony to Dr. Bendick’s on this Patty Haynes’ employment proce- ences subject, company for the did not even at- n perhaps change dures accounted for the tempt dispute testimony. to The district employment rates of blacks. The district court’s conclusion that the second standard is actually court’s statement of the obvious is “heavily weighted” in favor of blacks due disparate impact theory. of the recitation Its their “over-utilization” of MSES is therefore attempt explain away Plaintiffs’ reliance highly supported by speculative and is not upon standard three in fact embraces Plain- light Supreme the record. Court’s tiffs’ Title claim. There is no valid rea- VII supply explana- disapproval of court’s Indeed, disregarding son for this standard. tions of statistical information which is not record,14 light of the district court’s back-door rec- supported by the this Court should may judicial ognition Haynes’ practices that Ms. have not countenance the activism the disparities, every caused the there is district court. No sufficient reason exists otherwise, this standard as a valid the record —or for that matter— reason to consider cases, Supreme though theory generally appli- Court has that even In Title VII courts, cable, consistently required parties, inadequate support district record upon theory. circuit courts to base their decisions credi- use of the The Court concluded that case, arguments presented ble evidence or which exist in the facts in this the ... "under See, e.g., Mayor Philadelphia, fully enough satisfy record. theory developed is not 1332-33, 619, 621, (noting S.Ct. at 1333-34 The Court chose to State's burden rebuttal.” explanations Castaneda, that the circuit court’s constituted rely 430 U.S. at on the record. "speculative “supposition” inference” which 97 S.Ct. at support charge); could not serious Albemarle cases, Consistent with these then Justice Rehn- Paper Moody, 422 U.S. 438 n. Co. v. quist concurrence in Dothard v. elucidated in his 2362, 23, (1975) n. L.Ed.2d 280 S.Ct. 2382 45 may employer in that case Rawlinson job (asserting that a defendant’s relatedness justifiable employing the have had a reason proved through vague and un- claim "cannot be alleged discriminatory hiring practice. Howev- hearsay”). substantiated er, Rehnquist explained the that once Justice employer, employer Castaneda, the burden shifts to the Ap- the Texas Court of Criminal In —not Dothard, reason. the court—must articulate that peals proffered theory explained away a a case, 340, this S.Ct. at 2732. In charge in 433 U.S. criminal defendant’s of discrimination downplayed Plaintiffs' case grand jury practices. Supreme court has selection proffered by theory, deciding the defense. rejected with reasons not Court the State court's job analysis accurately pool qualified reflect the of a statistical of the dis- reference for standard, impact claim. This third parate applicants, permitted plain- ... we have ... of the hires were which shows 64.6% prima facie cases on such tiffs to rest their black, just percent of lower is four-tenths one 6, statistics as well.” 490 U.S. at 651 n. equates to 8.3 than the second standard S.Ct. at n. 6. consistency of standard deviations. the in The final statistical standard this case is and, amazing in first three standards is this probative general population much than more view, greatly to writer’s lends their credibili- statistics,15 for it hones that number down ty probative value. only the actual number of blacks who are not standard, though rejected by The final the general population, in in the who are not court, precise district is much more than market, the labor but who civilian also are accepted by Supreme in standards Court qualified gen- in and have shown interest Dothard v. Rawlinson and International factory-type eral work at in case.16 issue this Teamsters v.

Brotherhood United States. of rejecting The district court’s basis for compared make-up Those cases of defen support standard finds no in the record. The employees minority presence dants’ Dothard, spurned court first this standard because it general population. the Su comparison that a preme penitentia Court reviewed claimed between hired Alabama’s ry system hiring. people people inherently for sex discrimination in and unhired was upheld comparison The Court the district court’s reliance unreliable. While such certain- upon generalized, national statistics which ly perfect, is not the facts of this ease hiring system demonstrated Alabama’s prove comparison greatly that such a favors Dothard, against discriminated women. 433 D&L, Plaintiffs, unemploy- not the for the 329-331, U.S. at 97 S.Ct. at 2726-28. Simi among County ment rate blacks in Bolivar larly, compared the Teamsters Court far relevant times has exceeded the unem- minority proportion of with the ployment county.17 of rates whites proportion general, of minorities in the area- Hence, proportion black-unemployed Teamsters, population. 324, wide job-seekers obviously would exceed the 337 n. 97 S.Ct. at 1855 n. 17. The Court portion general factory of black workers. in Hazelwood School District v. United rejected The District also Court this last explained that States the Teamsters Court’s applicants standard because not all D&L general population figures “highly use of prior production-worker experience.18 have jobs probative” question because the were many applicants The court noted that have jobs, everyone low-skill which almost in the employment backgrounds in the service or population acquire. could 433 U.S. at 308 n. manufacturing industries. 97 S.Ct. at 2741-42 n. 13. While the court’s The Wards true, is, explicated entirely Cove Court further that “where observation is at the same figures general time, population might completely for the ... irrelevant to the issue at comparison 15. A hiring number of black hires at one chance in a thousand that D&L’s years general years happened by D&L with the the Grizzard chance— County population people Bolivar 2.4 standard deviations. Dr. Haworth found that comparison statistically signifi- and older also between the labor market and reveals disparities hiring. probative. D&L's analysis cant in the re- veals difference which constitutes 4.5 standard deviations. 17. In of the white work force was 4.3% unemployed. unemployment ap- The black rate proached five times that number: patently 20.7% 16. It is obvious—and even Dr. Haworth county unemployed. blacks in the were Like- acknowledged because the black labor —that wise, unemployment in 1984 rates showed that County heavily weighted force in Bolivar is in the unemployed of the were whites while 4% 19.9% jobs presence "top low-skills and has little unemployed. blacks ladder, rungs” employment number eligible production-worker jobs blacks greater much than the number of disregarded any comparison 18. The court with County manufacturing Bolivar civilian labor market. blue collar work force for comparison just applicants pos- even with the black labor mar- the similar reason that not all just manufacturing experience. ket reveals that there is a little than more sessed blue collar or deviations, that, plainly proportion applicants at D&L dard shows like hand —what black, proper standards, is a stan- hiring and hence what other were results availability appli- for this case. Each happen by dard of Grizzard did not chance.19 cant, experi- regardless of his or her work standards, any Under it seems abun- these ence, factory displayed general interest in dantly successfully clear that Plaintiffs have work, just working production- in as those prima made out a case of race discrimi- facie jobs type displayed such interest. While hiring practices,20 in nation D&L’s and the production comparison of the number of district court so found trial. applicants the number of to workers with jobs industry manufacturing or would service 5. Conclusion in interested unpersuasive be those —since jobs might have no interest in the the latter D&L, essence, in conceded Plaintiffs’ dis- job upon former standard here focuses parate impact21 hiring by —the in agreeing case person, the interest which each whether changes Patty Haynes in made her general applicant, or has shown in worker hiring practices namely, her restriction of — work, factory case. the work issue this application significant forms —resulted in the rejecting rationale for The district court’s change proportion in the of blacks hired. entirely final standard misses the boat. concession, Even without such a all credible standard, evidence demonstrates that D&L’s The last which to me is com- valid, during the employ- discriminatory Grizzard era was pletely shows that 70.1% of the jobs respect with to race. general factory-type in Bolivar This Court should ees County, excluding employees, therefore reverse on this D&L are issue and remand significance, damages. at 11.3 stan- for the district court to black. Statistical decide Travenol, factory Comparison County's general Bolivar civilian la- work in Baxter with which (2.4 deviations), D&L, directly manufactur- bor force standard located is across the street from (5.2 deviations), standard and labor force were black. (6.4 deviations) force standard blue collar labor Haworth's data is also inconsistent with Dr. statistically significant hir- likewise demonstrates Patty Haynes' experience. example, Haynes For ing disparities. Dr. Haworth found such com- occasion, every- testified that on one she invited parisons to valid. be apply general factory jobs to at a one mass application distribution. Three to four hundred upon Basing her conclusion the actual num- people February stood in snow on remaining applications, Dr. Haworth con- ber of morning apply job. Haynes for a testified that only the D&L tended 46.7% of 45.5% whites; many many there were more blacks than applications came from blacks. The district applicants more than of the were black. 51% numbers, accepted these which should be all, any comparison All in between the stan- only applicant-flow rejected, because generated availability dard of which Dr. Haworth tainted, incomplete irrevocably data is and as valid, propor- and reliable indicator of the discussed, are earlier but because the numbers produc- tion of blacks who showed interest in the demographics thoroughly illogical light of the jobs at D&L confirm that Dr. Ha- tion worker County. in Bolivar wrong. analysis numbers are Her worth's Although majority people in the civilian rejected be as fallacious and contrived. should force, force, manufacturing labor labor County the blue collar labor force in Bolivar majority disparate Plaintiffs’ 21.The rebuffed 51.9%, 58.2%, respectively— and 60.7% black — claim, finding basically treatment no use of ra- only professionals managers of the 3% however, epithets. Notably, cial the record black, county are 45.5% 46.7% fact, were, epithets used certain shows that range implies that the black labor force in the management personnel. James Grizzard testi- Mississippi County rural Delta of Bolivar is deposition, in a which neither the district fied heavily rungs" weighted "top more at the of the read, majority regularly that he court nor the rungs. employment at the bottom ladder than "boy.” data, Haworth, called black Grizzard believed Logic, census and even Dr. see friendly Patty Haynes to be a one. the term supra agree opposite n. is true. complain that Tom testified that she heard blacks Dr. Haworth’s numbers are not “out of Gamer, manager during plant production County, demographics touch” with the of Bolivar Further, years, "boy.” called them one vastly applicant- but are different from the supervisors had a of D&L’sblack testified that he proximi- in close flow data of factories who, during years, told foreman ty people to D&L. Of the MSES referred majority ap- jokes. factory jobs, The district court and general racist Colortile for 68.7% testimony. parently ignored Additionally, applicants black. 84.4% than the entire work pool motion was smaller B. Promotions Pouncy, employed those In force. Pool Proper Labor 1. promoted to level twen could be level eleven analysis and the court’s Like the district (S.D.Tex. ty. Pouncy, F.Supp. data, this majority’s review of the (5th Cir.1982). 1980), affd, 668 F.2d 795 analysis review of and finds that writer Lewis, only employees had Similarly, in greatly mislead- promotion statistics or field examiner level reached the GS-12 v. Nation- majority, quoting Lewis ing. The attorney “pre level were the GS-13 field Board, correctly sets Relations al Labor Lewis, eligible” promotion. sumptively for upon framework proper statistical forth the Comparison en at 1275. with the 750 F.2d promotion discrimination which claims improper, in those cases was tire work force establishing an “In must be based: claims had reached since not all of from statistical inference of discrimination employment promotion. level for required evidence, to a ‘required comparison [is] employees presumptively qualified pool Here, every quite different. This case is ” Maj. op. at 1290 eligible promotion.’ for eligible” “presumptively is production worker (5th Lewis, 1266, 1275 (quoting 750 F.2d positions ques- promotion to the leader for Cir.1985) (emphasis original)). tion, every “presumptively eligi- and leader is positions promotion to the foreman Lewis, ble” for case, this this in both Unlike educational, skill, or question. There is no Pouncy authored, v. Prudential writer America, objective requirement promotion for other Company to which Insurance excep- positions at D&L.22 With the refers, those minimum majority were there record,23 of the of the attendance tion objective qualifications promotions which for “qualifications” supervisors which the D&L comparison proportion of precluded of the subjec- entirely thought important were were proportion of mi- promoted to the minorities unsupported claim that tive.24 An in the overall work force. norities fact, experience big manager deposition plant a plant in a was James of the In Grizzard— so, people plant Even with fewer to 1986 and executive consideration. nine from 1982 .Cleveland experience promoted manager plants trial— were over two at the time of than five months' high grade years. dropped after the tenth Four of the out of school leader pos- plant not attained a GED thereafter. He did worked at the fewer than two nine had Indeed, any type college degree. he never sess All nine were white. months. college. even attended Tolbert, super- Watson and E.W. both Howard visors, prior testified that to James Grizzard's good Although was a reason for attendance arrival, they inexperienced had never seen such apparently rejecting promotion, blacks for Bailey, people promoted. the assistant Wallace rejecting a reason for whites. For sufficient department, supervisor over the maintenance purported- example, Plaintiff Richard Grant anyone pre- he did not know testified that ly good Grizzard recom- a worker whom James promoted period hav- who was without According promotion. to Robert mended for " ing years' experience production as a "some Goodman, supervisor, good plant Grant had worker. promotion attendance record. experience of the leaders and fore- The lack of Jerry Swin- which Grant was considered went to ford, plant productivity men in the hurt employee. worked at a white Swinford had arrived, Harry plant. Soon after James Grizzard just 14.6 months at the time of his D&L for D&L, Lomason, president of wrote Mr. Grizzard months, During promotion. Swinford those 14 manufacturing quality to inform him that the times, reported reported off off for sickness August systems control. were out of times, personal came to work business six Company it had Motor informed D&L that Ford times, twice, early left and was AWOL late three quality of work com- serious concerns about once. plant. One of the D&L from the Cleveland Bowen, example James who was Another plant managers presidents vice wrote D&L meeting Mr. Grizzard in hired as a foreman after give guidance fixing February 1985 to them During eigh- first the all-white VFW Club. his on-going quality One of problems. early job, Bowen left or left teen months on plants’ problems was "insufficient causes of the to work late and returned 19 times and came supervi- training personnel (production and or eight times. sion).'' deposition testified in his Mr. Grizzard nothing receiving experi- after supervisors plant that he did different 24. D&L testified that later, Harry Eight Lomason seniority important months consider- letter. ence and and three other individuals promotion. testified wrote Mr. Grizzard ations for James Grizzard

13H subjective plant, prior at a area in the all testified that to characteristics possess these is, arrival, proportion than substantially lower whites Mr. blacks and Grizzard’s whites itself, discriminatory in claim which promotions. They turned down asserted finds solace. apparently district people of that the number who turned down promotions change during did not such Mr. force, ap- which was entire work D&L’s departure. In Grizzard’s tenure or after his black, proper pool is the to proximately 70% fact, a substantial number of blacks who promotions compared,25 should be which the promotions during turned down the Grizzard promotions pool from which for that is the promotions had also turned down dur- promotions approx- At no time have came.26 however, years. pre-Grizzard employ- These Interestingly enough, imated 70%. D&L, repeatedly promotions.28 at 63.8% ees were offered

prior to Mr. Grizzard’s arrival promotions rejection to blacks. That num- upon testimony, of the went Based blacks’ such during his tenure plummeted ber to 22.7%27 promotions could not be the reason for the manager and returned to 63.7% plant as Indeed, promotions. stark decline lawsuit. after Plaintiffs filed When supervisor no ever even intimated that such why promotions asked the number of black was the case. drastically, had declined so neither Mr. Griz- questioned could supervisor

zard nor so Comparison Supervisors With in Boli- explanation. an offer County var Contrary majority’s to the reliance on agree Dr. Haworth and the district court turned fact that a number of blacks had comparison promotions that a of the at D&L Byron Kyle, supervisor promotions, down Watson, percentage supervisory with the of blacks plant, over one half of the Howard positions County manager plant, Bolivar is more relevant production the Cleveland Lofton, supervisor press comparison promotions than a and Charlie over the to the standing reputation compari- D&L’s with Ford 26. The district court stated that such a that disregarded practice company's “at an all-time low." Mr. Grizzard affirmed son directly appli- plant. that Mr. Lomason's statement was individuals from outside the notable, plant. disregarded deposi- cable to the Cleveland 'Tis in- district court Mr. Grizzard’s deed, written, the testimony practice pro- that soon after this letter was tion that D&L's was to supervisors plant began pro- at the Cleveland company. mote from within the experienced mote more workers. Finally, computing notes that in this writer supervi- 27. The district salaried court included promoted experience number of months’ of those promotion sors in its review of Plaintiffs’ claims. years, majority apparent- supervisors is errone- Such inclusion Vardaman, ly Larry included the demotion of ous. Plaintiffs have never claimed that D&L plant who had been at the for 194.8 months promoting discriminated on the basis of race in position before he was demoted from the of fore- supervisors. promoted man. Vardaman had twice been prior majori- foreman to Grizzard's arrival. The Hence, rejection promotion clearly did of a ty's inclusion of Vardaman in its calculation is employee pro- disqualify future an from demotion, the erroneous. Absent Vardaman's Watson, fact, production motions. Howard pro- evidence shows that the whites who were manager plant, testified that after he entire tenure, pos- moted Mr. Grizzard's indeed leader, promoted requested he that was first average approximately 15 months' sessed production-worker level he be returned promotion. experience at the time of their could handle the because he did not think that he again. Possessing job. promoted production- He was later proportion 25.The of blacks plant, highest jobs proportion one of the in the Cleveland worker level exceeds himself, Watson, rejec- position. Mr. demonstrated that his hired in that Dr. Haworth and Dr. pro- explained greater promotion a much tion of a was not fatal to his Bendick number voluntarily employment opportunities. quit It should not have been of whites their with motion this, Haynes or Richard Grant's D&L than blacks. Ms. confirmed fatal for Earnest Hall’s Indeed, opportunities testifying many quit Mr. Griz- whites soon after motion either. deposition average, employed that after Richard blacks are zard asserted in his are hired. On time, promotion long down the to which the D&L twice as as whites. Over Grant turned refers, high- longevity in a much district court Grant informed difference has resulted Grant, he, being again percentage company interested in er of blacks in the than promoted. whites. majority’s welding experience pool. comparison Such a is contention applicable labor general was needed for both maintenance highly questionable. positions completely and rack maintenance Haworth, According Dr. most of the Bailey, the assistant su- refuted Wallace compared she supervisory positions to which pervisor department. maintenance over the precision posi- craft promotions were D&L’s Bailey sought at trial that who testified one positions of those were tions. The “bulk” temporary upgrade needed to have a me- Dr. by craftsmen and craftswomen. held background, familiarity have a with chanical only held 33.4% Haworth testified that blacks ability equipment, electrical have the to un- jobs County. in precision craft Bolivar of the instructions, derstand written and verbal general factory contrary, 70.1% of the On rack, general, as maintenance opposed jobs To therefore com- belonged to blacks. positions, possess welding experience. Bai- positions to pare the leader and foreman ley disquali- averred that one would not be craft, precision opposed general to the as possess quali- if fied he did not of these factory promotions pool from which the in fact, ties. In he asserted that he would not comparing apples question came is like person simply even exclude a because that comparison simply is ir- watermelons. The Nevertheless, person could not weld. relevant.29 majority quali- contends that no blacks were general fied for the or rack maintenance 3. Conclusion positions. ap- maintenance The evidence proper analysis A of this reveals that issue n pears contrary. to be to the improperly in pro- D&L considered race its practices. provide a motion D&L did not Tommy Hardy: Contrary majori- why, neutral reason race 70% facts, ty’s Tommy Hardy recitation of the did up presumptively promotion pool is made of welding experience. not have “limited” To blacks, qualified only 22.7% the lead- contrary, Hardy that ran testified he promotions during went to blacks er/foreman body shop parts. in which he welded auto Mr. Grizzard’s tenure —a number which had body shop He started the in 1976 or dropped appears from 64.8%. It in operation and was still at the time of company promotion dispar- tried to make the Hardy’s welding experience trial. was there- ity disappear. legitimate is Because there no very Hardy fore extensive. averred that the explanation minority pro- for the decrease in type welding done at D&L was the same motions, prevail. Plaintiffs must This Court type welding performed in which he his reverse and remand for' a trial on the should body shop/ body shop in His work his also damage issue here. ability. revealed he had mechanical Hardy testified that he informed Lamar Temporary Upgrades to C. Rack/General Lamb, Hays Terry supervisors in the Maintenance30 shop, tempo- in maintenance of his interest rary upgrades numerous times. Qualified Employees majority’s temporary The review Daniel Anderson: Anderson testified that upgrade questionable. welding experience. evidence is also he had limited Howev- 29. The district court's were that blacks That number decreased his tenure to favorably during pre-Grizzard years Also, treated black. there were three 54.1% promotions is likewise unfounded. The positions during motions in the maintenance pre-Grizzard years roughly pool reflects the years. proper comparison is there- presumptively eligible employees. promotions fore the which went to blacks 33.3% compared as with the of the blacks who 54.1% majority 30. Both the district court and included presumptively eligible promotion. Al- positions analysis, opposed all craft in their as flawed, though majority's analysis this general positions and rack maintenance agrees prove writer that Plaintiffs did not wheth- They analyzed which issue here. also promotion disparity er the in the maintenance promotions positions, for all craft instead of the positions statistically significant that race or positions question. maintenance When James played part promotion discrimination plant, Grizzard came to the Cleveland 66.7% general for rack maintenance. decisions positions these maintenance were held blacks. upgrades during pre- me- hours of er, he had substantial he asserted period. taken mechanics suit No blacks were awarded such ability. He had chanical high period. majority upgrade four throughout his classes import import appears other weekly repaired places his or some and had school — mainte- unsupported when D&L’s to be the record —on the fact employees’ machines only person unavailable. Wallace personnel were that Mr. Wolfe was the nance be interested Bailey upgrades during that he would testified received rack maintenance apart could take production worker who in a the Grizzard era. James Grizzard testified machines, could. depart- as Anderson deposition the maintenance solely ment should not have awarded the Bailey testified Kemp, Wallace Sr.: Alfred Likewise, upgrades Mr. Wolfe. Wallace Sr., Kemp, had an during trial that Alfred why Bailey asserted that he did not know in maintenance. background excellent was awarded all of the rack main- Mr. Wolfe fact, at the arrival prior to James Grizzard’s majority’s upgrades. supply tenance general rack and Kemp both plant, received reason, albeit an and unfounded irrelevant The district court upgrades. maintenance reason, improper. appear would to be See every blocked found that Howard Watson supra note 13. Kemp could have temporary upgrade which question. in That years received unsupported the record. While 3. Conclusion prohibit testified that he would

Watson gen- Why whites received 1199.4 hours of if receiving temporary upgrades Kemp from upgrades and 745 hours of eral maintenance general factory for his Kemp were needed upgrades maintenance when blacks re- rack work, not know conceded that he did Watson just ninety-seven general hours of ceived upgrades such he had blocked whether hours of rack mainte- maintenance no Bailey question. Wallace likely upgrades will never be known nance why of no reason that he knew asserted deposi- any other court. one of his this or given single upgrade from Kemp was not tions, unequivocally refused to Mr. Grizzard August 1985. August 1982 to why company provide explicate failed According to Bai- Perry: Wallace Arthur compete for tem- everyone opportunity general ley, Perry qualified Arthur provide porary upgrades. He refused However, Bailey could work. maintenance give one explanation even when instructed why Perry had not been not state a reason attorney. by D&L’s filing prior to the of this given upgrades named Plaintiffs’ The numbers —and the lawsuit. loudly enough for them- protests speak — Only upgrade claim. temporary Analysis selves 2. Statistical Therefore, this comes from D&L. silence explanation some other Absent —which and remand for trial should reverse Court temporary up- give company did —D&L’s damages. disparities. racial grades grave reveal temporary majority correctly sets forth the D. Conclusion in footnote 36. Were

upgrade information insubstantial, temporary upgrade hours stated, For the reasons heretofore majority in by the percentages set forth and remand. writer reverse would consequential. might not seem that footnote I dissent. of hours involved the number up- general rack maintenance

both During the Grizzard

grades is considerable. given 1199.4

years, employees were white general maintenance. upgrades

hours just ninety-seven.

Blacks received area, maintenance Edward

In the rack Wolfe, employee, received a white

Otto notes the interviewer indicat (5th Inc., Cir.1992). 970 F.2d ed that he considered Anderson’s Once a limited case, welding experience, plaintiff prima establishes a more facie Bruce’s extensive employer welding experience, articulating bears the burden of “poor Anderson’s attend ance,” legitimate, nondiscriminatory some “very good” reason Bruce’s attendance in so, determining for its action. Id. If prom who to recommend for the the defendant does Moreover, employee otion.50 the notes also bears ultimate burden of indicate showing given by employee seniority that a black that the reasons with more em job. ployer pretext than were a Bruce turned down the for retaliation. Id. Thus prima Anderson did not establish a facie case quali of discrimination he parties appear agree because was not promotion fied for sought.51 plaintiffs passed prongs which he the first two Moreover, test, even if they charges Anderson did establish a retaliation as filed EEOC case, prima facie D L subsequently & demonstrated were terminated from their However, Hardy qualified promotion. 49. testified that he had done some weld- Wallace body shop high Bailey, dent, superinten- department in an auto the maintenance and took school welding experience testified that early Anderson’s "fix- classes 1970s. Anderson ing” simply preven- machines involved routine welding experi- also admitted that he had little ence, performed by tive maintenance duties all D & L although he had taken "mechanics" classes employees and was not similar to the duties high school. performed by department employ- maintenance ees. plaintiffs, citing any authority, 50. The without relying contend that the court erred in district appeal 52. The do not the district court's these notes because the interviewer never testi- conclusion, holding our based on in Carter v. fied. the notes were attached an as Bell, (5th Cir.1990), South Central 912 F.2d 832 joint stipulation exhibit to a submitted to the denied, cert. parties. stipulation both Because (1991) L.Ed.2d 1079 that their claims of retalia stated that both it and the attached "exhibits tory discharge § were not actionable under shall be received in evidence at the trial of this case," plaintiffs’ contention is without merit. prohibits employer section This from "dis- criminating] against any of his ... argues experience 51. Anderson also had employee] charge he [the because has made a ... 2000e-3(a). "fixing” subchapter.” § D & L machines and therefore was 42 U.S.C. under however, reasons, They disagree, positions.54 about cent the district court did not claim). rejecting the causal link between their terminations err in the retaliation filing charges. and the of the EEOC plaintiffs argue their termi argue filing that but for the causally nations are protect related to their complaints, would not have been EEOC

Case Details

Case Name: Daniel Anderson, Jr. v. Douglas & Lomason Co., Inc., Douglas & Lomason Co., Inc.

Court Name: Court of Appeals for the Fifth Circuit

Date Published: Sep 9, 1994

Citation: 26 F.3d 1277

Docket Number: 92-7554

Court Abbreviation: 5th Cir.