45 Fair Empl.Prac.Cas. 398,
Elise D. McINTOSH; Susan C. Sorrells; Odessa Hollis; Ann
E. Kennedy; and JoAnn Scherbring, individually and as
representatives of a class of individuals employed by the
United States Army who have been denied promotional
opportunities by reason of race, color, national origin, or
age, Appellants,
v.
Caspar W. WEINBERGER, Secretary of Defense; John O. Marsh,
Jr., Secretary of the Army; Edward Turner, individually and
in his capacity as Assistant Civilian Personnel Officer for
Troop Support and Aviation Materiel Readiness Command, Appellees.
Elise D. McINTOSH; Susan C. Sorrells; Odessa Hollis; Ann
E. Kennedy; and JoAnn Scherbring, individually and as
representatives of a class of individuals employed by the
United States Army who have been denied promotional
opportunities by reason of race, color, national origin, or
age, Appellees,
v.
Caspar W. WEINBERGER, Secretary of Defense; John O. Marsh,
Jr., Secretary of the Army; Edward Turner, individually and
in his capacity as Assistant Civilian Personnel Officer for
Troop Support and Aviation Materiel Readiness Command, Appellants.
Nos. 85-2023, 85-2086.
United States Court of Appeals,
Eighth Circuit.
Submitted Sept. 10, 1986.
Decided Jan. 14, 1987.
Rehearing and Rehearing En Banc Denied April 9, 1987 in No. 85-2086.
Louis Gilden, St. Louis, Mo., for appellants.
Joseph Moore, Asst. U.S. Atty., St. Louis, Mo., for appellees.
Before ARNOLD and FAGG, Circuit Judges, and BRIGHT, Senior Circuit Judge.
ARNOLD, Circuit Judge.
Plaintiffs Elise McIntosh, Susan Sorrells, Odessa Hollis, Ann Kennedy, and JoAnn Scherbring are current and former employees of the Automated Logistics Management Systems Activity (ALMSA), a branch of the United States Army located in St. Louis, Missouri. They brought suit in the District Court1 alleging that discrimination based on race, national origin, and age had occurred in certain ALMSA promotion actions, in violation of Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. Sec. 2000e-16, and the Age Discrimination in Employment Act of 1967, as amended, 29 U.S.C. Sec. 633a (Count I); that ALMSA officials had retaliated against them for filing Equal Employment Opportunity (EEO) complaints (Count II); that there had been systemic discrimination in ALMSA promotion practices and procedures, affecting a class of minority and older employees (Count III); and that individual defendant Edward Turner, an Army personnel official, had violated their due-process rights by concealing and destroying certain merit-promotion records (Count IV). The plaintiffs appeal from the District Court's adverse judgment on Counts I through III, and defendant Turner appeals from the judgment entered on a jury verdict for the plaintiffs on Count IV. Except as to one issue that has become moot, we affirm on all counts.
I.
A.
As background to the particular facts and issues of this case, it is necessary to discuss in some detail the system of personnel administration under which the federal government operates. The great majority of civilian positions in the Executive Branch, including those in this case, are in the "competitive service" and are administered under regulations issued by the Office of Personnel Management, the federal government's central personnel agency. The basic authority of OPM is found in 5 U.S.C. Secs. 1101 et seq. OPM issues government-wide regulations for use by other agencies and delegates to those agencies the authority to operate their own personnel programs within the limits of these regulations. OPM has the responsibility for overseeing agency personnel programs to ensure that its regulations are complied with.
Federal agency personnel administrators operate under a highly elaborate system of law and regulation. The basic statutes, found in Title 5 of the United States Code, define in general the responsibilities of the federal government as employer. Government-wide Civil Service Rules and Civil Service Regulations, promulgated by OPM and found in the Code of Federal Regulations, refine these principles and are binding on all agencies with employees in the competitive service. The Federal Personnel Manual (FPM), also published by OPM, contains the detailed day-to-day operating rules and is also binding on agencies. The FPM itself is further elaborated by FPM Supplements, which provide precise standards for certain categories of actions, and FPM Bulletins, which give additional, temporary guidance on selected issues. The regulatory authority of OPM extends throughout the various levels of regulations which have been described above. In addition to this body of government-wide regulations, each agency publishes its own instructions. These are supplementary to the OPM regulations, and relate those regulations to the particular organizational context of the agency. In the Department of the Army, each headquarters-level command issues instructions for the guidance of local units within the command. In the present case, the headquarters command is the U.S. Army Materiel Development and Readiness Command (DARCOM). Its written instructions are applicable to the constituent organizations under its control ("local activities"). It shares with OPM the responsibility for supervising personnel administration within the local activities. ALMSA is a local activity within DARCOM. For its own purposes, it has negotiated with its employees certain local procedures which further define the discretion of management in personnel matters. Although ALMSA had its own promotion plan, its personnel operations were performed by the Troop Support and Aviation Materiel Readiness Command (TSARCOM), another Army activity in St. Louis.
The federal personnel system is permeated by reliance on merit system principles. These principles are set out at 5 U.S.C. Sec. 2301, which includes the following language:
(b) Federal personnel management should be implemented consistent with the following merit system principles:(1) Recruitment should be from qualified individuals from appropriate sources in an endeavor to achieve a work force from all segments of society, and selection and advancement should be determined solely on the basis of relative ability, knowledge, and skills after fair and open competition which assures that all receive equal opportunity.
(2) All employees and applicants for employment should receive fair and equitable treatment in all aspects of personnel management without regard to political affiliation, race, color, religion, national origin, sex, marital status, age, or handicapping condition....
(3) Equal pay should be provided for work of equal value....
Much of the accumulated mass of federal personnel regulations is designed to enhance and protect these principles.
The promotion procedure at ALMSA, which is at issue here, is set up as follows: When a position vacancy arises, a personnel specialist consults with the supervisor over the position to determine the key performance requirements of the job. A crediting plan is assembled, which is to serve as a guide for the evaluation and ranking of eligible candidates for the vacancy. The vacancy is posted through a "Job Opportunity Announcement" (JOA). The JOA describes the position and its organizational location, the qualification requirements and the performance elements, and sets an initial closing date for receipt of applications. If the vacancy provides for a subsequent career promotion, i.e., promotion to a higher level without further competition, this is noted on the JOA. Applicants are requested to submit a resume of their qualifying experience along with a narrative description of their experience and training in the key performance areas of the job. The employee's current supervisor is also invited to comment on the employee's likelihood of successful performance.
After the closing date of the JOA, the personnel specialist determines the basic eligibility of each applicant. This is done by applying the OPM qualification standard (known in the trade as "handbook X-118") to each application. The X-118 standard describes for each occupation the kinds of experience which would be likely to provide the necessary knowledge and skill for entry and normal advancement. The standard specifies for each grade level of work within the occupation the number of years of general and specialized experience which must be shown for the applicant to be minimally qualified. Generally, the qualifying experience must have been progressively responsible, with at least one year of the specialized experience at a level of responsibility no more than two grades below the grade of the position to be filled. At the same time as the personnel specialist screens the qualifications of each applicant, he or she also checks the application materials to ensure that the applicant meets time-in-grade, time-after-competitive-appointment, and other regulatory prerequisites for promotion eligibility.
Once this screening process is completed, the personnel specialist assembles a panel of "subject matter experts." These persons, although not trained in personnel, are knowledgeable about the occupation for which applications have been taken. The panel is given the applications of eligible candidates and the crediting plan. Using the crediting plan as a guide, each panel member rates each application and the supervisors' remarks on the various performance elements which are identified in the plan. Then the panel discusses the ratings and reaches a consensus on a final numerical score for each candidate. The rated applications are then ranked and divided into three groups using a mathematical formula provided in the crediting plan. The highest group is the "best-qualified" pool. Candidates in this group are referred to the selecting official. Below this pool is a group of "highly qualified" (but not best-qualified) candidates. These are not referred, but they are available should subsequent vacancies arise. The lowest ranked candidates are designated as "qualified." They will not be considered for selection until the pool of highly qualified candidates has been depleted.
The list of best-qualified candidates is forwarded to the selecting official. Final selection from this pool is essentially a discretionary matter with the selecting official, provided, of course, that his or her choice is based on some job-related reason, and provided that the evaluation method gives a fair opportunity to each candidate. (E.g., ALMSA's promotion instruction provides that if the selecting official chooses to interview any candidate, he or she must attempt to interview all.)
B.
In February of 1981, two job vacancy announcements were posted for ALMSA employees. JOA 81-17 was for Management Analysts at grade 7 with career promotion potential to grade 9 ("GS-7 opt 9" in Army jargon). JOA 81-24 was for Management Analysts, GS-9 opt 11. All five plaintiffs applied for the GS-7 positions; plaintiffs Sorrells and Scherbring applied for the GS-9 position. Other persons applied as well. Plaintiffs Hollis, Kennedy, Sorrells, and McIntosh were determined by the personnel office to be eligible for the GS-7 position. Plaintiff Scherbring was determined to be ineligible. The four eligible plaintiffs were ultimately ranked by the rating panel among eleven best-qualified candidates for the five vacancies which were to be filled through JOA 81-17. Eventually, five persons, Judith Young, Rosemary Mest, Christy Wind, Maureen McNeill, and Carolyn Vierdag, were selected. All were white, and all but Ms. Mest were under 40 years old. None of the plaintiffs was selected.
On JOA 81-24 (the GS-9 position), plaintiff Sorrells was found to be highly qualified but not best qualified. Plaintiff Scherbring was found ineligible. Judith Young was the only person listed as best qualified. She was referred for selection, but instead the selecting official chose Marie Carver, a GS-7 personnel intern at TSARCOM. This choice was justified as a "lateral transfer," but it was in fact a promotion. Although the promotion did not violate OPM merit promotion regulations (Ms. Carver had acquired promotion potential to the GS-11 level when, on a previous occasion, she competed for the position in the personnel intern program), the action did violate ALMSA's own merit promotion procedures. In order to correct this procedural error, her application was rated by the same ranking panel which had earlier set up JOA 81-24. The panel determined that Ms. Carver would have been one of two best-qualified candidates for the position. Her promotion was therefore allowed to stand. ALMSA then created an identical additional position and placed the other best-qualified candidate (Ms. Young) in it at the GS-9 level.
This and other irregularities in the procedures used by ALMSA in filling these vacancies led the plaintiffs to believe that the successful candidates had been "preselected"2 for their promotions. They complained of discrimination to the Equal Employment Officer at ALMSA. Odessa Hollis alleged discrimination on account of her race (black); Elise McIntosh on account of being an American Indian; Susan Sorrells because of her Hispanic-American heritage; and Ann Kennedy and JoAnn Scherbring on account of their age (over 40). Their informal complaints were not resolved favorably. On June 11, 1981, they filed formal complaints of discrimination with the United States Army Civilian Appellate Review Agency (USACARA). These complaints eventually resulted in investigations and reports by ALMSA, TSARCOM, USACARA, DARCOM, and OPM.
In April 1981, the plaintiffs wrote to their representatives in Congress, complaining of irregularities and favoritism in ALMSA's promotion practices. This prompted congressional inquiries into the matter, which in turn prompted an investigation by DARCOM, the agency to which both ALMSA and TSARCOM report. In early June 1981, DARCOM sent personnel expert Elizabeth Bailey from its Alexandria, Virginia, offices to St. Louis to review ALMSA's personnel actions. Ms. Bailey's investigation lasted two days. She concluded from it that several deficiencies were evident in ALMSA's promotion actions, including improper rating of candidates' qualifications in JOAs 81-17 and 81-24.
The chief personnel officer at TSARCOM, Civilian Personnel Officer (CPO) Diane Ottolini, assigned defendant Edward Turner, Assistant CPO for TSARCOM, to handle matters connected with Ms. Bailey's investigation. When Ms. Bailey completed her investigation in early June, she directed Turner to undertake further review of the JOA 81-17 and JOA 81-24 promotion actions. Her oral instructions were followed up by a written directive from the Commander of DARCOM dated June 22, 1981. P.Ex. 10(a). Bailey's notes on her investigation, P.Ex. 10(b), which Turner acknowledged to reflect her orders accurately, J. Tr. IV:157-58,3 indicate that Turner's task was to "reconstruct" the two promotion actions. While Turner maintained that this meant that he was only to re-evaluate the basic qualifications of the applicants under OPM's X-118 standards, the plaintiffs presented extensive evidence showing that Turner was to redo both the minimum-qualification determination and the succeeding step, the rating and ranking of the qualified candidates by a panel of subject-matter experts. Expert witness Chris Kunz, a former OPM employee, testified that the term "reconstruct" had a generally accepted meaning in the federal personnel field, and that it meant that both the minimum-qualification determination and the rating-and-ranking phase were to be redone. Further, DARCOM's written directive of June 22, 1981 states: "reconstruction is directed only for the rating and ranking of all candidates who applied under each announcement." P.Ex. 10(a) at 2 (emphasis added). Finally, Francis "Bud" Steins and Valeda Henson, two former personnel staffing specialists drafted by Turner to carry out the review of the promotion actions, testified that Turner specifically directed them to "reconstruct" the actions, and that they understood from this that both a minimum-qualification screening and a new rating-and-ranking phase were required.
On June 11 and 12, 1981, Steins and Ms. Henson took the applications and, without reviewing the original determinations, applied the X-118 standards for the two positions, producing new minimum-qualification evaluations. Then, on June 16, 1981, they convened a subject-matter-expert panel whose members were three ALMSA Management Analysts who had no connection with the original actions; also present as observers were an EEO representative and a labor-union representative. The panel completed its rating-and-ranking task on June 17, 1981. The results of this Steins/Henson reconstruction were, as we further explain below, significantly different from the original evaluation of the applicants.
On the day the panel finished its work, Steins and Ms. Henson turned all of the panel materials, including the report of their results, over to Turner. Turner reviewed the results of the Steins/Henson reconstruction, and then, dissatisfied with it, convened a "third panel" to review the promotion actions. Two of the three persons Turner selected for this panel were Edward Jones, the personnel staffing specialist who had done or participated in the original minimum-qualification determinations for both JOAs, and Barry Bainter, Jones's immediate superior. Turner's panel did not re-rate the applications, but merely reviewed the original minimum-qualification determinations, referring at points to the Steins/Henson evaluations. This third panel concluded that the original determinations were "defensible." J. Tr. I:50.
This result was communicated to Ms. Bailey at DARCOM in a report dated July 25, 1981. Turner did not mention the Steins/Henson reconstruction in this report or in any other communication with DARCOM. In early August, DARCOM informed Turner that its response to the congressional inquiry had been drafted, and that it accepted the position taken in Turner's report. At DARCOM's request, Turner met with the five complainants and a number of other persons on August 11, 1981, to inform them of this result.
After Turner's panel had completed its work, he placed the Steins/Henson reconstruction materials on his credenza. Then, within a few months after DARCOM's response to his report, under Turner's orders the Steins/Henson results were shredded. Meanwhile, beginning on August 10, 1981, union representative Ralph Edwards lodged a series of requests and demands for the results of the reconstruction. In a September 9, 1981 meeting, Turner told union representatives that they would not receive the Steins/Henson results because they had not been used, and that he had not used them because the panel had exceeded its authority when it rated and ranked the applicants. J. Tr. II:156, P.Ex. 34 and 40. Turner did not indicate that the Steins/Henson reconstruction had been or would be destroyed. In October, in response to continuing union requests, Ray Fulkerson, a labor-relations specialist at TSARCOM, informed Edwards that he hoped to provide the union with the requested information. J. Tr. II:164, P.Ex. 42. In late November, Fulkerson gave Edwards a package of documents which Fulkerson presented as ALMSA's response to the request. However, the package did not contain materials from either the Steins/Henson panel or Turner's third panel. Instead, on December 9, 1981, Fulkerson informed Edwards that the materials from the Steins/Henson reconstruction no longer existed. In a February 9, 1982, letter to Edwards, Colonel Wayne C. Hogan, Commander of ALMSA, confirmed that the Steins/Henson results had been destroyed, noting that ALMSA efforts to locate the results had been unsuccessful.
At trial, several of the participants testified as to their recollection of the Steins/Henson reconstruction. Most significantly, they testified that with one exception, none of the persons actually selected for either position was rated even minimally qualified in the Steins/Henson reconstruction. The lone exception was Judith Young, who in the reconstruction survived the minimum-qualification screening for the JOA 81-17, GS-7 opt 9 position, but not for the position she ultimately obtained, the JOA 81-24, GS-9 opt 11 position. As to the plaintiffs, various witnesses recalled that Sorrells, Kennedy, and McIntosh were rated Best Qualified in JOA 81-17, and that Sorrells was rated Best Qualified in JOA 81-24. However, no one recalled the Steins/Henson reconstruction results for plaintiffs Hollis and Scherbring.
Turner's position at trial was that he had discarded the Steins/Henson reconstruction because he thought that Steins and Henson had been too strict in applying the X-118 standards and because they had convened a rating-and-ranking panel, a step that, as noted above, he maintained they were not supposed to take. He further asserted that the documents were destroyed because they had not provided the basis for his final decision in the matter, and because they contained confidential Privacy Act information, see 5 U.S.C. Sec. 552a, about the applicants. However, the plaintiffs' position, which a jury could and did accept, was that Turner rejected the Steins/Henson results because they demonstrated that ALMSA employees had engaged in wrongdoing; that for the purpose of covering this up and avoiding discredit to the agencies for which he worked and to his fellow employees, he convened a biased third panel to legitimate the original actions and then ordered the Steins/Henson results destroyed; and that he continued this cover-up and attempted to avoid discovery of his misdeeds by giving what were at best evasive responses to the requests of the plaintiffs and of union officials for production of the Steins/Henson results.
While DARCOM's August 1981 response to the congressional inquiry had accepted Turner's position, OPM and DARCOM continued to be concerned that something was amiss. In October and November 1981, OPM and DARCOM conducted a joint survey of ALMSA's entire personnel operation, including the JOAs challenged by the plaintiffs. In December 1981 and January 1982, OPM (P.Ex. 12) and DARCOM (P.Ex. 13), issued reports of their initial findings, stating that they found "strong indications of significant problems in the assignment of employees to positions." P.Ex. 12(a). OPM reported that it appeared that there had been cases involving "the use of non-competitive promotions as a means of preparing certain employees for future advancement," id., and a number of instances in which OPM's X-118 qualification standards had been ignored or misapplied. P.Ex. 12(b). With specific relevance to this case, OPM called into question the qualifications of selectees McNeill (P.Ex. 12(b) at 36), Vierdag (id. at 38), Wind (id. at 40), and Young (id. at 43), and questioned the prior experience progression of selectee Mest (id. at 29). ALMSA and TSARCOM did not acquiesce in these accusations; their responses to the OPM and DARCOM survey continued to adhere to Turner's position. See P.Ex. 49 and 77. Nothing in the record indicates that OPM or DARCOM took any further action on the matter.
The USACARA reports on the plaintiffs' formal agency complaints were completed on January 20, 1982. USACARA recommended that the complaints of Sorrells and McIntosh be sustained and that Hollis's complaint be denied. The record is unclear as to what was recommended on the ADEA complaints of Kennedy and Scherbring. Ultimately, ALMSA resolved part of Sorrells's complaint by retroactively reassigning her to a GS-7 Management Analyst Position and subsequently promoting her to GS-9. D.Ex. K and L. However, her complaint over non-selection for the GS-9 opt 11 position remained unresolved. Nor did ALMSA take any favorable action with respect to McIntosh.
After the USACARA report was completed, plaintiffs Hollis, McIntosh, Kennedy, and Scherbring requested an administrative hearing before the EEOC. See District Court Order of April 17, 1984. The hearing was scheduled for May 24, 1982. However, on April 5, 1982, this action was filed. Subsequently, EEOC temporarily stayed the administrative proceedings at the request of the plaintiffs, see P.Ex. 25, over the protest of the defendants, see Order of April 17, 1984.
The reprisal complaint (Count II) is based on the following facts: (1) At various times during the processing of these complaints, top management of ALMSA referred to the complainants as the "famous five," a term which the plaintiffs regard as demeaning under these circumstances. See, e.g., P.Ex. 16, ALMSA Board of Directors minutes dated March 4, 1982. (2) In 1981, after the initial discrimination complaints were filed, plaintiff Sorrells submitted an application to the TSARCOM personnel office for future vacancies in the Program Analyst occupation. The application was placed in the Applicant Supply File.4 At some point later in that year, Sorrells's application was overlooked for three vacancies in the Program Analyst occupation. Eventually, the error was corrected by her receiving priority consideration5 for three like vacancies. See B. Tr. VI:15-16. (3) In June of 1982, a vacancy for Management Analyst GS-9 opt 11 occurred. It was announced on JOA 82-19. Plaintiffs Hollis, Kennedy, Sorrells, and Scherbring applied. After the closing date of the announcement, but before the applications had been rated, the selecting official, Kenneth Dawley (who had also been the selecting official on JOA's 81-17 and 81-24 and who had been named in the administrative complaints as an alleged discriminating official) asked the personnel office to return to him the supervisory appraisals which he had completed or reviewed for certain of his employees who were applying for the job. This practice, although perhaps irregular, apparently did not directly violate any regulation. Mr. Dawley allegedly asked other supervisors to reconsider the appraisals which they had given certain of their employees. Ultimately, however, no changes were made and the applications and appraisal forms were returned to the personnel office. Mr. Dawley then requested that the vacancy be cancelled. This was done. In October 1982, another vacancy occurred in Mr. Dawley's office. JOA 82-19 was then reactivated and the applications were evaluated and ranked. Plaintiffs Sorrells and Hollis were among the best qualified, but were not selected for the position. Instead, Dawley selected Jane Null, a non-ALMSA employee who had applied after the cancellation of the initial vacancy.
C.
The initial complaint on April 5, 1982 alleged four counts: (I) individual Title VII claims of discrimination by Hollis, Sorrells, and McIntosh, and ADEA claims by Kennedy and Scherbring; (II) reprisal claims by all five plaintiffs; (III) a class count; and (IV) the records-destruction claim against Edward Turner. On March 21, 1984, the District Court denied the plaintiffs' motion to maintain a class action. D.R. at 114-21. On April 17, 1984, the Court dismissed the ADEA claims. D.R. at 122-26. The Court denied defendant Turner's motion for summary judgment on Count IV on May 24, 1984,
II.
In Count III of the complaint, the plaintiffs sought to maintain an action on behalf of a class of ALMSA employees who were American Indians, Hispanic-Americans, blacks, or over 40 years old. They alleged that ALMSA's promotion program was flawed by irregularities which systematically put members of these protected groups at a disadvantage. After two days of hearing on the class complaint, held during October 1983, the Court dismissed Count III on the ground that none of the plaintiffs had exhausted their administrative class remedies as required by 42 U.S.C. Sec. 2000e-16(c) and 29 C.F.R. Sec. 1613.601-.643.
The plaintiffs' position is, first, that their individual administrative complaints of discrimination were sufficient to give notice of a class claim since the matters complained of were inherently characteristic of a class claim. Second, they contend that even if their individual claims did not state the claim of the class, the federal courts may nonetheless entertain a class action despite the lack of an administrative record. We reject both arguments.
When a statute mandates that the complainant first seek relief through an established administrative mechanism before filing suit in the federal court, the administrative process must be fully exhausted before the court can act. Edwards v. Department of the Army,
The plaintiffs contend that exhaustion of their individual remedies is sufficient to confer jurisdiction on the District Court, not only to hear those claims, but also to hear a class claim which was never presented through the administrative channels, so long as the class claim has at its core the facts on which the individual complaints are based. They are mistaken. The administrative procedure established by EEOC for the processing of discrimination complaints includes a detailed method for pursuing claims of class-wide discrimination. 29 C.F.R. Secs. 1613.601-.643. The administrative complaint must include allegations of numerosity, commonality, typicality, and representativeness which parallel the requirements in Fed.R.Civ.P. 23(a). 29 C.F.R. Sec. 1613.601(b). The administrative complaint must set out specifically the facts on which the class seeks relief. 29 C.F.R. Sec. 1613.603(b). The class representatives must await either a final agency decision or a 180-day lapse without decision before commencing their suit in court. 29 C.F.R. Sec. 1613.641. Nowhere in the regulation is there any suggestion that processing of an individual complaint to exhaustion gives the complainant a right to carry an unadjudicated class count piggyback into federal court. See Lewis v. Smith,
The authority on which the plaintiffs rely is inapposite. They cite Chandler v. Roudebush,
Appellants further assert that even if an administrative class-action complaint is a prerequisite to this suit, they have fulfilled that requirement through their consolidated administrative complaint, which stated in part: "the complainants were not afforded fair and equal consideration without regard to political, religion, or labor organization affiliation, marital status, race, color, sex, age, national origin, non-disqualifying physical or mental handicaps or any other criteria [sic] that is not job related, including favoritism based on personal relationships or patronage." P.Ex. 83. They allege that this and other boilerplate stated "across-the-board allegations that fit the model for class actions," Appellants' Brief at 42, and thereby gave the Army adequate notice that it was contending with a class complaint. This argument is without merit. 29 C.F.R. Sec. 1613.603(b) requires that the administrative class complaint be stated specifically and in detail; Sec. 1613.601 requires that the Rule 23(a) criteria of numerosity, commonality, typicality, and representativeness be addressed. There is a good reason for this. If the agency is to attempt to resolve the grievance, and if a usable record is to be assembled, the nature of the complaint must be defined. An administrative proceeding which is based on guesswork by the defending party and by the adjudicating authority is likely to be a waste of time for all concerned. See Wade,
III.
Plaintiffs Ann Kennedy and JoAnn Scherbring appeal the dismissal of their ADEA complaint in Count I. They had filed consolidated administrative complaints along with the other plaintiffs. On January 20, 1982, the USACARA report was completed and they requested a hearing. Before the EEOC hearing (scheduled for May 24, 1982) could be held, this action was filed. They requested, and received, a temporary stay of proceedings from the EEOC, over the protest of the Army. By Order dated April 17, 1984, the District Court granted the defendants' motion to dismiss their complaint on exhaustion grounds.
Under the ADEA, aggrieved federal employees have two routes into federal court. They may avoid administrative processing of their complaints entirely or in part by filing a notice of intent to sue with the EEOC at least 30 days before filing the lawsuit (but within 180 days after the alleged acts of discrimination have occurred). 29 U.S.C. Sec. 633a(d). This requirement of advance notice to the EEOC gives the EEOC some time to attempt informal conciliation between the employer and the employee. See Ray v. Nimmo,
There is authority to support the government's position that exhaustion is required. See, e.g., Purtill v. Harris,
IV.
We turn now to the remaining individual claims under Title VII. These are Count I, alleging discrimination on account of race and national origin against plaintiffs Hollis, Sorrells, and McIntosh, and Count II, alleging reprisal. The District Court entered judgment for the defendants on July 12, 1985, following a bench trial lasting seven days. The Court also had before it all evidence which had been taken in the earlier hearings for class certification and the jury trial of the constitutional claim against defendant Turner.
For their appeal, the plaintiffs raise five alleged errors by the trial court: (1) The Court erred in ruling that the plaintiffs' adverse-impact theory was inapplicable; (2) the Court erred in not finding that the plaintiffs were necessarily better qualified than the persons actually promoted when the jury in the Turner trial had allegedly already determined that issue in the plaintiffs' favor; (3) the Court gave insufficient weight to the plaintiffs' statistical evidence; (4) the Court erred in not finding that the defendants' proffered explanation for the non-selection of the plaintiffs was pretextual when the selections were accomplished through violations of civil service laws and regulations; and (5) the Court erred in not finding that the plaintiffs were victims of reprisal when the Court did not make specific findings of fact on the question. We shall take up each of the plaintiffs' points in turn.
A.
The plaintiffs presented their Title VII case on both an adverse-treatment theory and an adverse-impact theory. The District Court held that a challenge to the employer's subjective decision-making processes was inappropriate for adverse-impact analysis. The plaintiffs contend that the District Court misapprehended the basis for their adverse-impact complaint. They say that they had no quarrel with ALMSA's promotion system "in [its] pristine form," the broad outlines of which conformed to OPM and Army regulatory guidelines, but rather pointed to ALMSA's "facially neutral ... violations of the legally mandated procedures." Appellant's Brief at 26-27. The plaintiffs complain of a systemic pattern of departures from the rules, including the use of uncontrolled details for the purpose of "trying out" prospective candidates for promotion, permissive misassignment of favored employees to jobs which would build their credentials for promotion, and final selection based on favoritism. There was ample evidence in the record that such practices did occur, and that ALMSA had been afflicted with such irregularities for a number of years. See, e.g., P.Ex. 12, 12a, 13, 14a, 14b, 14c, and 14d. The plaintiffs' theory was that this pattern of prohibited practices had become so ingrained at ALMSA that it had become a de facto promotion system in itself, without regard to the written promotion plan. We do not think, however, that it makes a difference whether the plaintiffs were challenging merely the subjective elements of the promotion system as written in the regulations or the de facto policies and practices which ALMSA used to carry out that system. The substance of this part of the complaint is the same in either instance: a challenge to the cumulative effect of an entire set of personnel operations, on the ground that members of minority groups were disadvantaged thereby.
The adverse-impact model is designed to give relief to plaintiffs who are the unintended victims of personnel practices which bear unevenly on members of protected minority groups. Such practices, if not justifiable on the ground of business necessity, operate as "built-in headwinds" for members of minority groups. Griggs v. Duke Power Co.,
Since intent is not an element of the adverse-impact case, this theory in effect gives the plaintiff a leg up on his proof. But the adverse-impact theory compensates for this advantage by imposing on the plaintiff a more exacting duty of showing the actual effect of the challenged practice. Lowe v. City of Monrovia,
The classic impact case is one in which a graded examination is used for selection purposes. This was the situation in Griggs. Impact analysis is also useful in assessing the effect of other objective measures. See, e.g., Green v. Missouri Pacific Railroad,
We are happy to say that the facts of the present case make it unnecessary to enter this labyrinth. Even if we were to hold that the plaintiffs' theory fits the adverse-impact model, we should still be constrained to uphold the decision of the District Court, because plaintiffs presented no substantial proof of disparate impact.
Proof of disparate impact was attempted through the use of workforce statistics. The plaintiffs' data purported to show the number of persons, identified by race, in each grade level at ALMSA. The data also purported to show the average number of years of federal service for each racial group in each grade. See P.Ex. 6a, 6b, and 6c. From these data, the plaintiffs sought to demonstrate that there was a disproportionately small number of blacks, Hispanic Americans, and American Indians in the upper grade levels at ALMSA, and that white persons were being promoted in less time than were minority group members. These data were fatally flawed. As a matter of causation, they did not isolate variables other than discrimination which could have caused the disparity (assuming that there was one). In algebraic terms, an adverse impact is shown through the equation (x)W/A = B/C, where W/A expresses the ratio of white persons promoted ("W") to the number of white applicants qualified for promotion at each grade level ("A"), while B/C represents the ratio of minority persons promoted at each grade level per the number of qualified minority applicants. If, when the equation is solved, the variable "x" turns out to be less than one, the fact of a disparity will appear. Further quantitative tests on "x" can be used to determine whether the disparity is significant. Instead of offering the foregoing analysis of promotion behavior, the plaintiffs offered data which gave a "snapshot" of the racial composition of each grade-level at ALMSA. Such figures do not take into account the important variables of how many minority group members were actually applying for promotion, or how many were actually qualified for promotion. Without isolation of these variables through determination of the actual ratio of persons promoted to persons qualified, any value found for "x" is unreliable to show that there was a disparity, let alone that defendants' promotion policies caused it. See Gilbert v. City of Little Rock,
Thus, even if the District Court erred in ruling that the plaintiffs' disparate-impact theory is inapplicable in this sort of case, the error would not lead to reversal. If we were to reverse now and remand to the District Court, its next step would be to make a finding of fact on the question of adverse impact. If it were to find that adverse impact existed, there would inevitably be a second appeal, and we would be compelled to reverse this finding, because, as we have just explained, there is no substantial evidence in this record to support it. (We have no reason to suppose that the parties have not already put in all of their proof relevant to this issue.) Accordingly, it is appropriate for us now to take the step of upholding, on other grounds, the District Court's rejection of the disparate-impact theory. By so doing, we simply avoid the rigmarole of a futile remand and second appeal, all ending in a result that is already clear on the record. The Supreme Court has approved such a use of appellate power. See Pullman-Standard v. Swint,
B.
The bulk of the plaintiffs' case proceeded under a theory of disparate treatment. In such a case, the key question, on which the plaintiff always bears the burden of proof, is whether the defendant intentionally discriminated against the plaintiffs.
The District Court found as fact that the plaintiffs were all members of protected classes; that they all applied and were qualified for the Management Analyst positions; none of them was selected; and the persons selected were all white. The plaintiffs therefore all established a prima facie case of discrimination. McIntosh v. Weinberger, No. 82-491 C (5), slip op. 5 (E.D.Mo. July 12, 1985). The District Court also ruled that the defendant had articulated a legitimate, non-discriminatory reason for the non-selection of the five plaintiffs, i.e., that the selecting officials believed that the persons actually promoted were better qualified for the positions. The Court ultimately held that the plaintiffs had not succeeded in proving that they had suffered intentional discrimination. Opinion of July 12, 1985, at 6. This is a finding of fact which is binding on us unless clearly erroneous. We are not convinced that appellants have met this burden.
We address in turn the major assaults made by plaintiffs on the District Court's findings. The plaintiffs first urge that the trial court erred in finding that the defendants had articulated a legitimate, nondiscriminatory reason for the plaintiffs' nonselection. The Court accepted the proffered reason, which was that the persons ultimately selected were better qualified than were the plaintiffs. The plaintiffs contend that in the earlier jury trial on Count IV (the records-destruction claim against Mr. Turner) the jury had necessarily found as a fact that the plaintiffs were among the best qualified for the positions and that the selected persons were not qualified at all. The plaintiffs argue that the jury finding precluded a contrary finding on the identical question by the Court in the Title VII bench trial.
Generally, when a case involves both a jury-trial portion and a bench trial, any essential factual issues which are central to both must be tried to the jury, so that the litigants' Seventh Amendment jury-trial rights are not foreclosed on the common factual issues. Sisco v. J.S. Alberici Const. Co.,
No such identity of issues is present here. The jury trial on the constitutional tort claim ended in a general verdict for plaintiffs. We cannot be sure what specific facts the jury found. We assume for present purposes that the Turner jury thought the personnel reconstruction that was destroyed had reached the conclusion that plaintiffs were "best qualified," and that the persons actually selected were not qualified at all. This is not at all the same issue that was before the Court in the bench trial on the Title VII claims. There, the issue was whether the promoting officials (acting long before the reconstruction overseen by Turner) believed (whatever the actual fact) that the persons promoted were better qualified than the plaintiffs. The issues are close to each other, and a finding as to either of them would be heavily influenced by one's view as to relative qualifications in fact, but the issues are not identical. There can therefore be no estoppel. We recall the language of Lord Coke: "Every estoppel ... must be certain to every intent, and not to be taken by argument or inference." Co. Lit. 352b.
Next, the plaintiffs assert that the Court clearly erred in not giving adequate weight to the plaintiffs' statistical evidence. The plaintiffs' objection really is that the District Court did not document with sufficient detail its findings on the statistical evidence. There is no merit in this contention. The Court is not required to render specific findings with respect to every item of evidence presented to it. Talley v. United States Postal Service,
We also reject the plaintiffs' argument that the defendants' explanation of their conduct was necessarily pretextual because "the selections were accomplished through deviations from legally mandated civil service procedures and requirements ..." Appellants' Reply Brief 11. To be sure, there was ample evidence from which the District Court could have found that ALMSA was regularly running afoul of OPM's regulations for conduct of its personnel operation. The Court in fact so found. Opinion of July 12, 1985, at 7-9. But the point of this case is not to determine whether ALMSA was running a tight ship; it is to determine whether ALMSA intentionally discriminated against the plaintiffs in their non-selection for the Management Analyst positions. To some degree, the bulk of evidence which the plaintiffs submitted to prove that ALMSA was disregarding personnel guidelines actually proves too much; ALMSA's pattern of disregard for the regulations was not directed against these plaintiffs or against the classes to which they belong; rather, it was general and pervasive. It affected all applicants, regardless of their race or national origin, by inserting excessive subjectivity, uncertainty of expectations, and (perhaps) favoritism into the promotion process. These are ills; but the general impact on all applicants does not require the factfinder to reach a conclusion that failure to follow guidelines amounted to intentional discrimination on the basis of a criterion prohibited by Title VII.
The plaintiffs argue, citing authority from a district court in another circuit, that "when the federal government is involved, deviation from legally mandated civil service procedures and requirements cannot form the basis of a legitimate, lawful explanation sufficient to overcome a prima facie case of discrimination." Hogan v. Pierce, 31 F.E.P. Cases 115, 126 (D.D.C.1983). While evidence of regulatory violations by a federal employer is probative of pretext, the statement quoted above paints too broadly. A personnel decision is not discriminatory simply because it is a bad personnel decision. The legal and regulatory guidelines administered by OPM serve worth-while purposes other than preventing discrimination. The bare fact of irregular personnel decisions does not foreclose the employer's opportunity to demonstrate a lack of discriminatory motive.
Finally, the plaintiffs assert error in the District Court's finding that they had not been victims of reprisal by ALMSA on account of their filing complaints of discrimination. There was contradictory evidence and testimony on all of the points relevant to this claim. Had the Court chosen to make all the inferences favorable to the plaintiffs and to discredit all the testimony favorable to defendants, which it could have done, a resulting judgment for the plaintiffs would not have been reversible. However, there was ample evidence and testimony from which the Court could also have found, as it did, for the defendants. Bearing in mind that the trial court is in the best position to judge for itself the weight of evidence and the credibility of witnesses, we see no clear error here.
The plaintiffs' objection to the lack of detailed findings on the selection made under JOA 82-19 is insubstantial. The Court had before it all the evidence from the class-action hearing, the jury trial, and the bench trial. It was unnecessary for the Court to comment on each and every datum of evidence.
For all the reasons which we have discussed, the judgments of the District Court for the defendants on the Title VII counts are affirmed.
V.
We next consider defendant Turner's appeal from the verdict against him in his individual capacity awarding the plaintiffs damages on their Bivens claims. Turner's principal arguments are that he is entitled to qualified immunity, that the plaintiffs' Bivens claims are precluded because other statutory remedies were available, and that the plaintiffs did not demonstrate that Turner's actions harmed them. He further argues that the verdict was excessive and the result of passion and prejudice, that there was insufficient evidence to support the jury's punitive-damages award, that the jury's verdict was inconsistent, and that the District Court erred in denying an instruction stating that Turner was personally and individually liable for any judgment.
A.
Turner maintains the plaintiffs' action for damages is barred under the qualified-immunity doctrine of Harlow v. Fitzgerald,
government officials performing discretionary functions, generally are shielded from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.
1.
The District Court expressly declined to reach the question whether Turner violated clearly established law; instead, citing Harlow's "discretionary function" language, the District Court rejected the qualified-immunity defense on the ground that Turner had acted in a ministerial rather than a discretionary capacity when he ordered the Steins/Henson documents destroyed. Order of May 24, 1984, at 5. The plaintiffs maintain that Turner violated a ministerial duty because he was acting pursuant to a clear DARCOM directive to reconstruct the ALMSA promotions, and, most importantly, because the Federal Personnel Manual, Army regulations, and ALMSA regulations imposed an absolute, non-discretionary duty to maintain all documents relating to placement actions for at least two years. See Federal Personnel Manual Ch. 335-29, 6-1.b. (3); Army Regulation C5, AR 690-300, Ch. 335.1, 1-4 Req. 5.a.; ALMSAR 690-12, Ch. 8-1.
The distinction between ministerial and discretionary duties of public officials has a long history. See, e.g., Amy v. The Supervisors,
2.
The question thus becomes whether Turner violated the plaintiffs' clearly established rights. Turner forfeited his immunity only if it was clearly established at the time of these events in 1981 that the property interests the plaintiffs claim in fact existed and that Turner's actions deprived them of these property interests without due process of law.
As we have said, it is not sufficient for this purpose to demonstrate simply that it was clearly established that Turner's actions contravened the federal personnel document-retention regulations. Again, Davis holds that "[n]either federal nor state officials lose their immunity by violating the clear command of a statute or regulation ... unless that statute or regulation provides the basis for the cause of action sued upon."
The plaintiffs claim that Turner deprived them of due process in connection with two separate property interests. First, the plaintiffs contend that they had a property interest in ALMSA's merit-promotion system, that this system, taken as a whole, amounted to a contract under which plaintiffs were assured that their applications for promotion would be fairly judged or reviewed on their merits. They maintain that by concealing the Steins/Henson results from DARCOM, substituting for them the results of a third panel with two members who were interested in the original placement action, and destroying the Steins/Henson documents, Turner denied them due process with respect to this interest, because he subverted DARCOM's investigation, an investigation that was aimed at assuring that the merit-promotion system's requirements had been followed. Second, the plaintiffs contend that they had a property interest in their EEO claims, and that Turner's destruction of the Steins/Henson materials denied them due process in this regard by depriving them of favorable evidence whose equivalent they could not obtain elsewhere. On appeal, Turner has addressed only the latter theory, arguing that it was not clearly established that a property interest existed in EEO claims. We address both theories.
" '[P]roperty' interests subject to procedural due process protection are not limited by a few rigid, technical forms;" instead, " 'property' denotes a broad range of interests that are secured by 'existing rules or understandings.' " Perry v. Sindermann,
Of course, absent statutes, regulations, or some other basis for such a legitimate claim of entitlement, government employment and promotion decisions are normally not subject to procedural due-process protections. See, e.g., Colm v. Vance,
We also conclude that the plaintiffs had a property interest in their EEO claims that was clearly established in 1981. Turner, as we noted above, has argued that this interest was not clearly established; his position is that not until Logan v. Zimmerman Brush Co.,
Turner has not argued that even if, as we hold, the property interests claimed by the plaintiffs were clearly established, it was not clearly established that his conduct deprived them of process which they were due, and we see no room for such a contention. For the issue here is not, as it is in many qualified-immunity cases, see, e.g., Davis,
We thus conclude that Turner was not entitled to qualified immunity.
B.
Turner next argues that the District Court erred in holding that the plaintiffs could maintain a Bivens constitutional tort action, citing Bush v. Lucas,
The general rule governing recognition of Bivens actions is set out in Carlson v. Green,
Thus, "Bush held only that where civil servants enjoy meaningful and constitutionally adequate ... statutory remedies for [constitutional] violations, federal courts should not imply an additional Bivens -type damages remedy under the Constitution." Doe v. United States Department of Justice,
Turner suggests that his actions may have been remediable through a discrimination or retaliation claim under Title VII and the ADEA. However, the plaintiffs have made no claim and adduced no evidence showing that Turner acted for discriminatory reasons or in retaliation for the filing of EEO claims, rather than out of a bureaucratic desire to squelch charges that ALMSA employees had not done their duty. Turner also suggests that he may have engaged in a prohibited personnel practice actionable by the Special Counsel of MSPB under 5 U.S.C. Sec. 1206. However, even if we assume this to be true, we note that the statute authorizing the Special Counsel to investigate and seek corrective action for such practices places the decision whether or not to do so in his discretion, 5 U.S.C. Secs. 1206(b)(3)(A), (h), that the injured employee is not a party to and has no right to appeal any action taken, 5 U.S.C. Sec. 1207(c), and, most importantly, that in an action filed by the Special Counsel the MSPB is authorized to take disciplinary action against Turner, but not to award any affirmative relief to the injured employee. 5 U.S.C. Sec. 1207(b). This is not adequate to bar a Bivens action. Kotarski,
C.
Turner asserts that the plaintiffs failed to prove that they were injured by his actions, that the award of actual damages was excessive, that there was no basis for an award of punitive damages, and that the jury verdict was inconsistent. Turner further contends that he was entitled to a jury instruction stating that he, rather than the government, would be liable for any damage award. We address each of these points in turn.
We think a reasonable jury could have found that had Turner ever reported the Steins/Henson results to DARCOM, corrective action would have been taken and the plaintiffs would have received the promotions they sought, or, perhaps, some other favorable action. There was evidence that the Steins/Henson reconstruction found the applicants who received jobs in the original actions not minimally qualified for the positions they received, and that Sorrells, Kennedy, and McIntosh should have been rated Best Qualified. Further, while they did not know how Hollis and Scherbring fared under the Steins/Henson reconstruction (though they knew that Hollis had been ranked Best Qualified in the original action), this was entirely due to Turner's destruction of the documents, and this was sufficient to support a jury inference that Hollis and Scherbring were also rated Best Qualified. The jury could have found that Sorrells was the highest-rated applicant for the JOA 81-24, GS-9 position, and that the remaining four plaintiffs were among six or seven applicants ranked Best Qualified for the five JOA 81-17, GS-7 positions. From this the jury could reasonably have concluded that but for Turner's actions, the plaintiffs would have received some relief as a result of DARCOM's investigation, resolving any doubts about what DARCOM would have done against Turner because, absent his wrongdoing, there would be no such doubts.
Thus, the plaintiffs did adequately demonstrate injury in connection with their DARCOM-investigation claim, and we therefore reject Turner's appeal on this point.
We further hold that there was ample evidence to support the jury's award of $10,000 in actual damages to each plaintiff. The plaintiffs presented evidence of lost earnings, adverse effects upon career potential, and emotional distress caused by Turner's acts; the jury could reasonably have found that some or all of these elements amounted to the damages awarded.
We also conclude that the punitive-damage award is supported by the evidence. The plaintiffs argued, and the jury could have found, that Turner, in breach of his official trust and responsibilities, intentionally concealed and later destroyed the Steins/Henson results in order to cover up wrongdoing at ALMSA. This sort of conduct justifies assessing punitive damages. See Smith v. Wade,
We do agree with Turner that the jury's verdict exhibits an inconsistency. While the jury awarded each plaintiff $10,000 in actual damages, it also placed a check mark on the portion of the verdict form providing that the plaintiff had suffered a constitutional violation but no actual injury, and should receive one dollar in nominal damages. However, as the plaintiffs point out, Turner did not object to this when the clerk announced the verdict, when the jury was polled, or at any other time before discharge, and did not request that the jury be returned to the jury room to reconsider its verdict; he thereby waived any objection to this inconsistency. Ludwig v. Marion Laboratories, Inc.,
Finally, we believe that the District Court properly refused Turner an instruction telling the jury that he would have to pay any damage award personally and that the government would not be liable for the award. At no point in the proceedings was it suggested to the jury that the government would be responsible; all that Turner has pointed to is that his attorney, though not in uniform, was sometimes addressed as "Captain Woodley." We think this inadequate to require that the jury be given an instruction upon an issue that has no bearing on fault, an instruction that might induce the jury to decide upon improper grounds, such as undue sympathy for the defendant. See Williams v. Bennett,
VI.
To summarize: with respect to the question of exhaustion of administrative remedies for plaintiffs' individual claims of age discrimination, the appeal is dismissed as moot, and the cause remanded to the District Court with directions to vacate as moot its order dismissing these claims, and thereafter to proceed as law and justice may require. In all other respects, the judgment of the District Court is
Affirmed.
Notes
The Hon. Stephen N. Limbaugh, United States District Judge for the Eastern and Western Districts of Missouri
"Preselection" is not defined in the statutes or in the Code of Federal Regulations. In common usage, as shown in the record and by its use in the FPM, it means favoritism toward one or more candidates who are competing for promotion. Preselection may involve giving a favored employee a competitive advantage through choice work assignments, or misuse of noncompetitive details to higher-graded work, or interference by a selecting official in the evaluation and rating process prior to actual selection. Although preselection is usually a violation of merit principles, it does not in itself amount to illegal discrimination against the employees who were not favored
As we further explain infra, p. 1423, there were a number of different hearings in this case, each dealing with a different aspect of the case. We will use the following designations to refer to various portions of the trial record: "J. Tr." for the jury trial transcript and "B. Tr." for the bench trial transcript
The Applicant Supply File is a pool of applications by outside candidates which is maintained by the personnel office for a limited period of time. When a vacancy occurs, these applications are supposed to be screened in order to determine whether any of the outside applicants is eligible. Any application that survives this screening is placed in competition with those of the internal candidates who have applied in specific response to the vacancy announcement
Priority consideration is a mechanism by which a candidate who has missed consideration is referred singly to the selecting official before other candidates are considered
The filing of a formal complaint with the employee's agency, as was done here, satisfies this requirement. See Purtill v. Harris,
The plaintiffs suggest that Turner surrendered his qualified-immunity defense by failing to appeal immediately the District Court's denial of summary judgment on this ground. While Turner was clearly entitled to take such an appeal, see Mitchell v. Forsyth,
While it rejected Turner's qualified-immunity defense on summary judgment, the District Court also submitted an instruction on the defense to the jury, requiring the jury to examine the objective reasonableness of Turner's conduct and his subjective good faith. J. Tr. V:74-75. Turner contends, and we agree, that submission of this instruction was error. Mitchell makes clear,
However, while the District Court erred in submitting the legal question of immunity to the jury, this presents us with no occasion to disturb the judgment. First, it appears that the contested instruction had its origin with Turner's counsel, and that he did not object to its submission. Further, the jury's rejection of the defense has not adversely affected Turner, since in any event our review of this legal question is de novo, and we conclude, as we explain below, that Turner was not entitled to qualified immunity.
The Davis Court did not derive this limitation on the ministerial-duty exception from the common-law, ministerial-discretionary function concept of official immunity, but instead imported the stricture from its analysis of the objective test of immunity, which focuses not upon whether the official was exercising discretion, but upon whether he violated clearly established law. As we explain below, infra p. 1432-33, earlier in its opinion, the Davis Court had ruled that the clearly established law which has been violated must be the basis of the plaintiff's cause of action, so that officials can anticipate when their actions may result in monetary liability.
The plaintiffs have not argued that Davis's requirements are met by violation of the document-retention regulations on the theory that the regulations are a source of the property interest that gives rise to their claim. We are not, in any event, inclined to believe that it is adequate under Davis simply to demonstrate the violation of such a regulation where the constitutionally-protected interest was not itself clearly established, since Davis emphasizes,