Pollocks v. Sunland Training Center (85 F. Supp. 2d 1236, 2000)

United States District Court for the Northern District of Florida 2000-02-23

Citations

Full opinion text

ORDER ON LIABILITY — FIRST GROUP OF PLAINTIFFS

HINKLE, District Judge.

This is a Title VII action in which 46 plaintiffs allege a pattern of racial discrimination ' in hiring and promotions by the Sunland Training Center in. Marianna, Florida, a state facility for the mentally disabled, spanning the years 1976-1984. By agreement of both sides, the claims have been divided into five discrete groups for purposes of trial on the issue of liability. Trial has been completed on the liability issue with respect to the first group, consisting of five African American plaintiffs who assert they suffered racial discrimination in hiring. This order sets forth the court’s findings of fact and conclusions of law on the issue of liability regarding this first group of plaintiffs. I conclude' that each of these five plaintiffs was the victim of intentional racial discrimination in hiring in violation of Title VII.

Introduction

During the period at issue, Sunland was a residential facility for the profoundly retarded. The defendant State of Florida owned and operated the facility. Residents lived in houses, sometimes referred to as “cottages.” Basic care for residents, including assistance with such necessities as eating, dressing, bathing and toilet functions, was provided by employees assigned to the various cottages;, these care-giving employees were known as “cottage parents.”

The five plaintiffs now before the court applied many times during the period at issue for the job of “cottage parent.” The only stated qualifications for the job were a high school diploma or “[experience in rearing a family or in related custodial or nursing environments”; such experience could be “substituted on a year-for-year basis for the high school education requirement.” The five plaintiffs all had high school diplomas as well as experience of this type, but they were passed over many times for the position of cottage parent. They also applied for but failed to obtain other jobs at Sunland for which they were qualified.

When it opened, Sunland had separate houses for African Americans and whites, respectively. African Americans were substantially under-represented among Sunland employees. By 1976, the first year at issue in the case at bar, however, segregated housing had been eliminated, and African American hiring was on the upswing. Indeed, Sunland made substantial and commendable efforts to increase African American hiring throughout the 1976-1984 period.

Still, African American applicants did not receive the same treatment as whites with respect to hiring for the positions at issue. African Americans, including these five plaintiffs, suffered intentional racial discrimination in hiring for these positions throughout this period. This is shown by (1) statistical evidence indicating that Sun-land hired a substantially lower percentage of qualified African Americans for these positions than would have been expected if decisions had been made without intentional discrimination, (2) evidence of the goals and policies that Sunland administrators followed during this period intended to raise African American hiring but to do so to a level significantly below the level that would have been achieved by non-diserimi-natory hiring, and (3) evidence of the specific treatment of the plaintiffs as analyzed under the familiar burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802-04, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). This order begins by recounting briefly the procedural history of this action, then summarizes the applicable legal principles, and then addresses in turn each of these three sources of proof.

Procedural Background

This litigation traces its origin to a class action complaint filed in this court on February 11, 1977, alleging a pattern of racial discrimination in initial hiring and promotions at Sunland in violation of Title VII of the Civil Rights Act of 1964, as amended. The first named plaintiff was Mary Worlds; the action became known as the Worlds case. The court initially certified a class but later decertified the class, effective as of June 1, 1985. The court notified former class members that they could seek to intervene by not later than July 31, 1985.

The plaintiffs in the case at bar were members of the Worlds class but were not named plaintiffs in that case. When the class was decertified, they filed a timely motion to intervene in Worlds. By order dated April 24, 1987, the court denied the motion. Six days later, on April 30, 1987, plaintiffs filed the case at bar.

The court in Worlds ultimately ruled that one plaintiff had been the subject of racial discrimination in hiring at Sunland but was entitled to no relief because he had made only token efforts to secure other, readily available employment over a period of some 13 years. The court ruled that the other Worlds plaintiffs had not been victims of racial discrimination. The Worlds plaintiffs appealed, and the state cross-appealed. The parties in the case at bar agreed that this action should be stayed pending resolution of the Worlds appeal.

By order dated May 13, 1997, the Eleventh Circuit affirmed this court “on all rulings.” Worlds v. Department of Health, 116 F.3d 491 (11th Cir.1997) (unpublished opinion at 2). In due course, litigation of the case at bar resumed. Defenses arising from the requirement for the filing of administrative charges as a prerequisite to a Title VII action have been addressed by separate orders and remain pending. Subject to those defenses, this order resolves the issue of liability for the five plaintiffs whose cases have been tried.

Applicable Law

Title VII of the Civil Rights Act of 1964, as amended, prohibits discrimination in employment based on race or other specified characteristics. See 42 U.S.C. § 2000e-2(a)(l). If, as plaintiffs claim, the persons making hiring decisions at Sun-land considered plaintiffs’ race as an adverse factor in the employment process, then the state violated Title VII. This legal principle is beyond dispute and, indeed, is not disputed by the state in this case.

It also is undisputed that plaintiffs who allege intentional discrimination in hiring may prove their cases in at least two ways: through direct evidence of discrimination (for example, testimony or other statements of the employer that race was a reason for the hiring decision at issue) or through the familiar burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802-04, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), which focuses on the employer’s hiring decision with respect to a specific position on a specific occasion. At least in class actions, plaintiffs also may prove their case in a third way: by proof that the employer has followed a pattern or practice of intentional discrimination. See, e.g., Franks v. Bowman Transp. Co., 424 U.S. 747, 96 S.Ct. 1251, 47 L.Ed.2d 444 (1976). The parties disagree on whether the pattern or practice method of proof is also available in a case that, like the case at bar, is not a class action.

It probably would be futile, and in any event it clearly is unnecessary in the case at bar, to attempt to set down an exhaustive list of, and definitive rules governing, the manner in which Title VII plaintiffs may prove their cases. The issue ultimately is whether, in making the employment decision at issue, the employer engaged in intentional discrimination based on a prohibited factor such as race. McDonnell Douglas sets forth a convention by which this issue may be addressed. But McDonnell Douglas does not create additional substantive elements of a Title VII case, over and above intentional discrimination, that a plaintiff must prove in order to prevail. Thus a plaintiff who proves intentional racial discrimination through admissible evidence is entitled to prevail without regard to whether he or she proves a “prima facie case” as defined in McDonnell Douglas. Such evidence may include proof of a pattern or practice, so long as the pattern or practice evidence, together with all the other evidence in the case, shows that the employer engaged in intentional racial discrimination with respect to the particular employment decision at issue.

In the case at bar, plaintiffs have produced no “direct evidence” of intentional racial discrimination in hiring during the period at issue. Plaintiffs have proceeded under McDonnell Douglas and also have presented statistical and other evidence of a pattern and practice of racial discrimination in hiring. I find that there was a pattern and practice of racially discriminatory hiring for the positions at issue at Sunland during the period 1976-1984. I also conclude that plaintiffs have proven their entitlement to prevail under McDonnell Douglas with respect to numerous individual positions. And even if, as the state asserts, plaintiffs properly could proceed only under McDonnell Douglas, plaintiffs’ statistical and other evidence would remain admissible on the ultimate question of whether the hiring decisions at issue resulted from intentional racial discrimination. For the reasons that follow, I find that plaintiffs were victims of intentional racial discrimination in violation of Title VII and that, but for that discrimination, they would have been hired for specific positions they have identified.

Proof of Discrimination

I. STATISTICAL PROOF

Courts long have recognized that statistical evidence may be relevant in assessing whether an employer has engaged in intentional discrimination. See, e.g., Hazelwood School Dist. v. United States, 433 U.S. 299, 307, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977). Such evidence is relevant in hiring cases because one would expect that, if an employer makes hiring decisions without regard to race, then given enough hiring decisions, the proportion of hires of a given race would approximate the proportion of qualified , applicants of that race, other things being equal. Id.; International Bhd. of Teamsters v. United States, 431 U.S. 324, 339-40 n. 20, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977). Some deviation between the percentage of hires and the percentage of qualified applicants, by race, of course would be expected, simply as a result of chance. But the odds that a deviation of a given magnitude would result solely as a matter of chance are also subject to calculation.

The relevant statistical analysis starts, then, with (1) a compilation of the employer’s actual hiring record by race for the type of position at issue during the time period at issue, and (2) a determination of the racial composition of the pool of qualified applicants to the employer for the type of position at issue during the time period at issue. The racial composition of the pool of qualified applicants is sometimes referred to as the “benchmark,” because it is the approximate result that race-neutral hiring would achieve, on average, given enough hiring decisions, and it therefore provides a reference or “benchmark” to which the employer’s actual hiring results can be compared.

In the case at bar, the quality of available information on hiring for the positions at issue, by race, and on qualified applicants for the positions at issue, by race, is far from perfect. Both sides have submitted expert testimony compiling and interpreting statistical evidence derived from the imperfect available information. Not surprisingly, the parties and their experts disagree in various respects.

The parties’ principal disagreement is on the appropriate source that should be used to determine the benchmark. Plaintiffs’ expert Dr. David W. Rasmussen has compiled data on actual applications to the Sunland facility for positions of the types and during the periods at issue. Defendants’ expert, Dr. Paul H. White, asserts that the data on actual applications are too unreliable to be used, so he has used census data (that is, information on persons designated in the category of jobs at issue in the same geographic area) as a proxy for data on actual applications.

I find Dr. Rasmussen’s analysis credible. I conclude that the actual applicant flow data used by Dr. Rasmussen, while imperfect, are data of a type reasonably relied upon by experts in the field and thus provide a proper basis for his expert testimony. See Fed.R.Evid. 703; see also Bazemore v. Friday, 478 U.S. 385, 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986) (recognizing that statistical evidence need not be perfect to have probative value and holding erroneous district court’s failure to consider probative but imperfect statistical evidence). I find that these actual applicant flow data provide a far better benchmark against which to measure Sunland’s hiring record than the census data used by Dr. White. I find that, under the circumstances o'f this particular case, the census data relied upon by Dr. White provide an unreliable benchmark.

The positions most at issue with respect to the five plaintiffs whose cases have been tried include the position of cottage parent and other positions within the “paraprofessional” category denominated as category EEO 5. Dr. Rasmussen’s analysis indicates that the appropriate benchmark for hiring of African Americans in category EEO 5 jobs at Sunland, calculated based on actual applications, was 50%. For the period 1977-1980, however, only 66 of 274 persons hired for such jobs at Sunland (or 24%) were African American. For the period 1981-1984, only 52 of 176 persons hired for such jobs at Sunland (or 29.5%) were African American. These statistics produce standard deviations for these periods of -8,58 and -5.43, respectively, which in turn indicate that, other things being equal, such a low percentage of African American hires would result by chance far less than one time in a thousand. The odds of such a low percentage of African American hires for the 1977-1980 period occurring by chance, and being followed by such a low percentage of African American hires for the 1981-1984 period again by chance, are long indeed.

Dr. Rasmussen’s analysis, including his use of a 50% benchmark, draw additional support from the apparently undisputed evidence of actual hiring at Sunland beginning in 1985. In 1985, Sunland hired 50 new employees in category EEO 5, of which 27 (or 54%) were African American. In 1986, Sunland hired 20 new EEO 5 employees, of which 11 (or 55%) were African American. In 1987, Sunland hired 33 new EEO 5 employees, of which 19 (or 57.6%) were African American. A reasonable inference is that the available pool of EEO 5 applicants at Sunland has approximated 50%, just as Dr. Rasmussen concluded, that from 1976-1984 the race of African American applicants was a negative factor in the hiring process, and that the discrimination ended in 1985.

I do not find credible Dr. White’s contrary analysis, which resulted in a 80.9% benchmark for African American hiring in category EEO 5 based on census data showing that, in the appropriately-weighted relevant geographic area, 30.9% of persons designated in category EEO 5 jobs were African American. First, as a matter of principle, actual applicant flow data (if available), not census data, provide the appropriate benchmark. See, e.g., Payne v. Travenol Lab., Inc., 673 F.2d 798, 823-24 (5th Cir.1982) (recognizing that reliable applicant flow data rather than population data provide appropriate benchmark against which hiring record should be measured). This is because hiring decisions at Sunland were made from applicants for positions at Sunland, not from persons counted in the applicable job category in the census. If, as appears to have been the case, 50% of the qualified applicants for EEO 5 jobs at Sunland were African American while 30.9% of persons counted in the census in category EEO 5 in the geographic area were African American, one would expect Sunland to have hired roughly 50% African Americans for EEO 5 jobs at Sunland, not 30.9%, if indeed hiring at Sunland was done on a race-neutral basis. Id.

Dr. White does not dispute this. He says, however, that Dr. Rasmussen’s data are so unreliable they should not be used. Dr. White thus says there is no reliable evidence that in fact 50% of qualified applicants for EEO 5 positions at Sunland during the 1976-1984 period were African American. Dr. White says the best reliable data are the census data, showing that 30.9% of EEO 5 persons in the appropriately-weighted geographic area were African American. Dr. White thus says the appropriate benchmark is 30.9%.

Some of Dr. White’s criticisms of Dr. Rasmussen’s data are incorrect, and some deal with alleged errors so small that they would have no material effect on Dr. Rasmussen’s ultimate conclusions. More fundamentally, Dr. White has identified no race-sensitive errors (that is, errors that would be expected to skew Dr. Rasmussen’s results along racial lines), no errors that explain the striking increase in African American hiring for the years 1985-1987, no errors that explain the close correlation between Dr. Rasmussen’s benchmark and actual hiring for the years 1985-1987, and no errors that explain the substantial divergence between Dr. White’s benchmark and actual hiring for the years 1985-1987.

This latter point is especially significant. Sunland’s hiring record for 1985-1987 is much more consistent with Dr. Rasmussen’s view that an appropriate benchmark for African American hiring in category EEO 5 at Sunland was 50% than with Dr. White’s view based on census data that an appropriate benchmark was only 30.9%. Dr. White has offered no explanation for why, if his analysis is correct, the hiring rate for African Americans in category EEO 5 at Sunland consistently exceeded 50% for the years 1985-1987.

In short, I find that Dr. Rasmussen’s 50% benchmark is analytically sound, supported by imperfect but reasonably reliable data, and far more consistent with the undisputed evidence regarding actual hiring during the years 1985-1987 than Dr. White’s proposed benchmark.

That the 50% benchmark dramatically exceeds the 30.9% African American share of EEO 5 persons in the geographic area does not undermine these conclusions. A reasonable hypothesis is that African Americans applied for cottage parent positions at Sunland in numbers disproportionate to their representation in the census. This would not be surprising for at least two reasons. First, as this record confirms without dispute, African Americans faced a substantially higher unemployment rate in the relevant geographic area at the relevant time than whites. There were thus a higher percentage of African Americans looking for work than projected by Dr. White’s analysis, which took no account of the relative unemployment rates by race. Second, the position of cottage parent is a demanding and sometimes unpleasant and even demeaning position that one reasonably would expect to attract a disproportionate share of applicants with historically lower expectations and fewer good alternatives; this in turn would include a disproportionate share of African Americans, who in Jackson and surrounding counties of Florida in 1976 and even in 1984 faced not only higher unemployment but also continuing racial discrimination, at least from some employers. Thus one would not reasonably expect the racial breakdown of applicants for the position of cottage parent at Sunland to mirror the census data used by Dr. White, and one would not reasonably be surprised if, as Dr. Rasmussen determined for the period 1976-1984 and as the actual hiring record suggested for the period 1985-1987, as many as 50% or more of the applicants for the position were African American.

Finally, even Dr. White’s own analysis suggests there was racial discrimination in hiring for EEO 5 positions at Sunland for the years 1976-1980. His report concludes that Sunland hired 211 persons in category EEO 5 during that period, only 50 of whom were African American. Using Dr. White’s census-based benchmark of 30.9%, one would have expected Sunland to hire 65.2 African Americans, not 50. This produces a standard deviation of -2.187, which social scientists would regard as statistically significant; the odds of such a result occurring by chance, other things being equal, are roughly 3/100. This evidence standing alone is hardly overwhelming, but it does support the conclusion that Sunland’s hiring in category EEO 5 was not race neutral.

In sum, the statistical evidence indicates that Sunland’s pattern and practice during the years 1976-1984 was to engage in intentional discrimination against African Americans who, like plaintiffs, applied for positions in category EEO 5.

II. SUNLAND’S EXPRESSED GOALS

Further support for this conclusion comes from the testimony of Sunland administrators concerning their hiring goals during the 1976-1984 period.

As of 1976, the Civil Rights Act of 1964 had been in effect for only 12 years and had 'been applicable to the states for only four years. The Constitution itself and other statutes had of course predated Title VII, but decisions recognizing their effect on state hiring were not of long standing. As of 1976, the workforce at Sunland still reflected the effects of earlier race-based hiring.

Commendably, Sunland administrators and state officials in Tallahassee were engaged throughout the 1976-1984 period in a good faith effort to improve African American hiring. Unfortunately, they set as a goal bringing African American hiring up to a level that would match the African American population in the geographic area. The result was that supervisors who received 50% African American applications were nonetheless under instructions to bring African American hiring up to a level of around 30%. Phrased differently, the message to hiring authorities was that they could discriminate against African American applicants, so long as the frequency of the discrimination was not too severe.

The actual results obtained at Sunland are consistent with a finding that the announced goal had a discriminatory effect on hiring. Dr. White’s benchmark, based on census data for employment category EEO 5, was not substantially different from Sunland’s announced goal, based on population. Dr. White’s analysis showed that Sunland’s African American EEO 5 hiring for the period of 1976-1980 fell significantly (although not drastically) below his benchmark, while by the period 1981— 1984, hiring approached his benchmark. A reasonable inference is that African American hiring improved from 1976 forward, reaching the announced goal for the period 1981-1984. The announced goal, however, fell substantially below what Title VII required: non-discriminatory hiring.

In short, Sunland’s administrators had the laudable goal of improving African American hiring. They chose a target employment rate, however, that fell far short of what would be produced by the nondiscriminatory hiring the law required, because of a significant divergence between the African American share of the population in the geographic area and the African American share of applications for the positions here at issue. When the flawed target was implemented by hiring authorities, the result was discrimination against African Americans, albeit discrimination less severe than had been in place earlier, and less severe than undoubtedly still was in place among some other employers in Jackson and surrounding counties.

III. MCDONNELL-DO UGLAS

That there was employment discrimination at Sunland in the types of position at issue of course does not necessarily mean that the five plaintiffs now before the court were victims of that discrimination. Their specific circumstances may be analyzed under the burden-shifting framework set forth in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802-04, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973) and Texas Dep’t of Community Affairs v. Burdine, 450 U.S. 248, 253-54, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981).

Under that framework, the initial burden is on the plaintiff to establish a “prima facie case.” The Eleventh Circuit has defined a prima facie case for this purpose as follows:

[Plaintiffs may establish a prima facie violation by showing that they are members of a group protected by title VII, that they sought and were qualified for positions that [the defendant employer] was attempting to fill, that despite their qualifications they were rejected, and that after their rejection [the employer] either continued to attempt to fill the positions or in fact filled the positions with [persons outside plaintiffs protected class].

Walker v. Mortham, 158 F.3d 1177, 1187 (11th Cir.1998) (bracketing in original in part), quoting Crawford v. Western Elec. Co., 614 F.2d 1300, 1315 (5th Cir.1980), citing McDonnell Douglas Corp. v. Green, 411 U.S. 792, 804; 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973).

If the plaintiff establishes a prima facie ease, the burden shifts to the defendant to present evidence that it acted for a legitimate, non-discriminatory reason. While this burden is “exceedingly light,” Holifield v. Reno, 115 F.3d 1555, 1564 (11th Cir.1997), it cannot be met solely by a defendant’s presentation of legal argument or after-the-fact rationalizations; the defendant must present “admissible evidence” of the actual reason for the action at issue. See, e.g., Texas Dep’t of Community Affairs v. Burdine, 450 U.S. 248, 255, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981). So long as the articulated reason is not based on race or other prohibited characteristics, it need not be a reason that seems wise or logical; under Title VII, an employer lawfully may take action for any non-discriminatory reason, good or bad, fair or unfair, or for no reason at all. See, e.g., Combs v. Plantation Patterns, 106 F.3d 1519, 1543 (11th Cir.1997) (noting that employee cannot overcome employer’s proffered reason “merely by questioning the wisdom of the employer’s reason”).

If the defendant meets the requirement of presenting evidence of a non-discriminatory reason for its action, the burden is on the plaintiff to establish that defendant’s proffered reason is pretextual and that the plaintiff has been the victim of discrimination. On this issue, the statistical and other evidence addressed in earlier sections of this order, as well as all other relevant evidence, may be considered.

I find that each of the five plaintiffs now before the court has established a McDonnell Douglas prima facie case with respect to various positions; that the state has provided legitimate, non-discriminatory reasons for some of the hiring decisions at issue but has failed to provide such reasons for many others;' and that many of the proffered legitimate, non-discriminatory reasons are clearly pretextual. I find that each plaintiff was the victim of intentional discrimination and that absent such discrimination each plaintiff would have been hired at Sunland during the period at issue. The reasons underlying these conclusions, with specific examples illustrating the reasoning, are set forth in the text below. An appendix to this order sets forth my conclusions with respect to each plaintiff and each specific position at issue.

A. Prima Facie Case

Plaintiff Gwenevere Long is African American. From July 1977 through some time in 1983, she submitted 15 to 20 applications per year to Sunland for positions including cottage parent, residential training instructor, dietary worker, and laundry worker. She was qualified for these positions. She was never offered any job. Many of the successful applicants were white. Ms. Long has met the McDonnell Douglas four-part test establishing a pri-ma facie case with respect to each such specifically identified position filled with a white person.

Plaintiff Queen Granberry is African American. She repeatedly applied for positions at Sunland from May 1969 (well before the beginning of the period at issue) through some time in 1980. Among the positions for which she applied were cottage parent, food service aide, residential training instructor, and beautician. She was qualified for these positions. She was never offered any job. Many of the successful applicants were white. Ms. Granberry has met the McDonnell Douglas four-part test establishing a prima fa-cie case with respect to each such specifically identified position filled with a white person.

Plaintiff Marinda Hayes is African American. She- applied for positions at Sunland approximately eight or nine times per year from the early 1970s (well before the beginning of the period at issue) through some time in 1982. Among the positions for which she applied were cottage parent, food service aide, and cook. She was qualified for these positions. She was never offered any job. At least some of the successful applicants were white. Ms. Hayes has met the McDonnell Douglas four-part test establishing a prima facie case with respect to the specifically identified positions filled with a white person.

Plaintiff Jennifer Pittman is African American. She applied for positions beginning in June 1977, left the geographic area and stopped applying later that year, returned and began applying again in March 1980. Among the positions for which she applied were cottage parent, food service aide, and custodian. She was qualified for these positions but was rejected many times. Many of the successful applicants were white. Ms. Pittman has met the McDonnell Douglas four-part test establishing a prima facie case with respect to each such specifically identified position filled with a white person.

Plaintiff Erma Calhoun is African American. She constantly applied for positions at Sunland beginning in 1972 (well before the beginning of the period at issue) and continuing throughout the period at issue. Among the positions for which she applied were cottage parent, food service aide, nursing assistant, and laundry worker. She was qualified for these positions. She was never offered any job. Many of the successful applicants were white. Ms. Calhoun has met the McDonnell Douglas four-part test establishing a pri-ma facie case with respect to each such specifically identified position filled with a white person.

B. Alleged Non-Discriminatory Reasons

Prior to trial, plaintiffs propounded to the state an interrogatory asking for Sunland’s alleged legitimate, nondiserimi-natory reason for choosing the successful white applicant for every position at issue. The state responded with no specific reason for any decision. The state said that, if plaintiffs “by competent substantial evidence can establish having applied for any other positions, then Defendants respond upon their best information and belief that the best qualified applicant was chosen.” See, e.g. PLEx. 71. Under the law of the circuit, such a general, conclusory assertion is not sufficient to meet the employer’s burden of providing a legitimate, nondiscriminatory reason for a decision, in response to an employee’s prima facie case. See, e.g., IMPACT v. Firestone, 893 F.2d 1189, 1194 (11th Cir.1990).

For some positions, the state failed to proffer any legitimate, non-discriminatory reason for selecting a white applicant over the plaintiff, other than the conclusory interrogatory answer. With respect to those positions, plaintiffs are entitled to prevail, based on their presentation of an unrebutted prima facie case.

For other positions, the state presented evidence at trial over and above its interrogatory answer. I admitted the evidence so that the record would be complete, notwithstanding the unfairness of allowing the state to proceed in this manner. The state’s failure to answer the interrogatory with a candid description of the position it intended to present at trial was inexcusable.

In addition, in many instances the non-diseriminatory reason offered by the state at trial was merely an after-the-fact rationalization conjured by a person who did not participate in the hiring decision and did not know the true basis of the decision to hire the white winner. An employer does not meet its burden by proffering such an after-the-fact hypothesis; the employer’s burden is to present “admissible evidence” of the actual reason for the action at issue. See, e.g., Texas Dep’t of Community Affairs v. Burdine, 450 U.S. 248, 255, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981); Lee v. Russell County Bd. of Educ., 684 F.2d 769, 775 (11th Cir. 1982).

These procedural flaws in the state’s presentation would justify rejecting the state’s proffered explanations. The unavailability of better proof, however, is not solely the state’s fault. In a case alleging racial discrimination in a major state institution and seeking equitable relief, the state ought not lightly be foreclosed from presenting its best case on the merits. I thus have considered on the merits the evidence presented by the state at trial regarding Sunland’s reasons for the hiring decisions at issue. For the most part, those reasons (if supported by evidence that they were the true reasons) would constitute legitimate, non-discriminatory reasons sufficient to carry the state’s intermediate McDonnell Douglas burden.

Thus, for most positions for which the state offered testimony at trial of a reason why the white winner was selected over the plaintiff, I proceed on the assumption that the state has successfully met its intermediate burden, thus placing on plaintiffs the burden of proving, by the greater weight of the evidence, that each proffered reason was pretextual and that intentional discrimination was a reason for the hiring decision at issue.

C. Pretext

Some of the state’s proffered reasons for hiring decisions were not pretextual. The record establishes that, for the most part, Sunland gave a very strong preference to current Sunland employees over applicants from outside. Sunland gave a substantial, although not necessarily dispositive, preference to veterans. Sunland often, although not always, gave a preference to male applicants for some cottage parent positions (in at least some male cottages) and to female applicants for some cottage parent positions (in at least some female cottages).

Beyond these objectively verifiable preferences, however, the state’s proffered explanations do not fare as well. The after-the-fact explanations offered by the state’s witnesses did not have the ring of truth; they were unpersuasive and often contradictory.

For example, Sunland hired Jeanette Bruner over plaintiffs Gwenevere Long and Jennifer Pittman for cottage parent position 11986 on January 23, 1981. The state proffered explanations that, first, Ms. Bruner was 40 years old and thus more mature, and second, Ms. Bruner raised six children, which is important experience. I find these explanations pretextual. Plaintiff Marinda Hayes was also approximately 40 years old when she applied for various positions, but that never seemed to help her. Plaintiff Erma Calhoun raised five children, but that never seemed to help her. Factors that are cited only when they help whites, not African Americans, are often pretexts for racial discrimination; I find that to be the case with respect to the reasons given for selecting Ms. Bruner. In fact, Ms. Long was a temporary employee at Sunland at the time of the application and had experience at other nursing facilities; Ms. Pittman had been a temporary employee at Sunland and a teacher’s aide at an elementary school; and Ms. Bruner was an Avon salesperson. If there was a legitimate race-neutral reason for hiring Ms. Bruner over Ms. Long and Ms. Pittman, this record does not reflect it.

Similarly, the state’s proffered explanation for choosing William Pierce over Ms. Long and Ms. Pittman for cottage parent position 11946 on May 12, 1982, was that he had experience with Sunland clients and staff through his experience with the Washington County School Board. But as of the date of his application, Mr. Pierce had been a teacher’s aide for only a month. His other work experience was less than a year shipping and receiving at a uniform company in Tennessee. Ms. Long had spent six months as a CETA employee at Sunland itself and had two years experience at a nearby nursing facility. Ms. Pittman had three months experience at Sunland itself and a year as a teacher’s aide in an elementary school. If there was a legitimate race-neutral reason for hiring Mr. Pierce over Ms. Long and Ms. Pittman, this record does not reflect it.

The state’s proffered explanation for choosing Rhonda Reddoch over plaintiff Queen Granberry for cottage parent position 12054 filled October 26, 1979, was that Ms. Reddoch had two children and a high school diploma. Ms. Reddoch had not worked for seven years as of that time. Plaintiff Erma Calhoun had five children and a high school diploma but was never hired; the state did not attempt to explain why Ms. Reddoch’s children and high school diploma netted her a job while Ms. Calhoun’s similar credentials did not. In any event, Ms. Granberry, who lost out to Ms. Reddoch, also had a high school diploma, and, in addition, Ms. Granberry had cared for her grandmother and a deaf mute aunt for many years, had worked for a year as a teacher’s aide and at Sunland as a temporary CETA employee, and had been continuously employed for eight years (a factor cited by the state as a positive factor for successful white applicants). If there was a legitimate race-neutral reason for hiring Ms. Reddoch over Ms. Granberry, this record does not reflect it.

The state’s proffered reason for choosing Mary Lou Elmore over Ms. Granberry for cottage parent position 12081 on February 15, 1980, was that Ms. Elmore was at that time a temporary worker at Sun-land. From Ms. Elmore’s application, however, it appears she did not even meet the minimum qualifications for the cottage parent position; she had neither a high school education nor enough experience to substitute year for year for the missing education, as required by the written job qualifications. In any event, Ms. Granber-ry was clearly better qualified than Ms. Elmore. I find the explanation for the hiring of Ms. Elmore pretextual.

The state’s proffered reason for hiring James Gesslein over plaintiff Marinda Hayes for cottage parent position 12005 was that he had 40 hours of junior college credit and had been an orderly at a nursing home. Mr. Gesslein’s application reflected one semester of junior college and six months work at the nursing home. Ms. Hayes had two years of college and almost 18 months as a part-time care giver. If there was a legitimate race-neutral reason for hiring Mr. Gesslein over Ms. Hayes, this record does not reflect it.

The state’s proffered reason for choosing Donald Perry over plaintiff Jennifer Pittman for service aide position 11680 on April 15, 1983, was that Mr. Perry had done a good job while working at Sunland on a prison release program. Mr. Perry had no high school diploma, had served time in prison for grand theft, and listed as work experience only two day laborer positions of less than a year combined. Ms. Pittman had a high school diploma, was at that time a current Sunland custodial employee (thus qualifying for the virtually dispositive preference for current employees), had additional prior experience at Sunland, and had additional experience as a teacher’s aide. There is no evidence that Ms. Pittman had not done good work in her two Sunland positions. I find pretex-tual the assertion that Mr. Perry was hired because of his good work in the prison release program.

Finally, the state’s proffered reasons for choosing Linwood Daffin over plaintiffs Erma Calhoun and Queen Granberry for cottage parent position 11963 on November 16, 1979, included that this was a position in a male cottage, that Mr. Daffin had a bachelor’s degree, that Mr. Daffin worked with other inmates as a chaplain while he was serving a two-year drug trafficking sentence, and that Mr. Daffin had changed jobs four times in three years, showing his willingness to work. But the state interviewed at least three women for the position (including Ms. Granberry), suggesting this was not a “male position.” Ms. Calhoun and Ms. Granberry both had experience in patient care facilities much more relevant than experience as a chaplain to fellow prisoners. Mr. Daffin’s frequent job changes were hardly a positive factor. And in any event, Ms. Calhoun and Ms. Pittman also were willing to work (as confirmed by their own work histories and by their many applications to Sun-land). Commendable as Sunland’s willingness to offer employment to a convicted felon may have been, the state' did not suggest this was a reason for preferring Mr. Daffin over applicants with no criminal record. I find pretextual the state’s proffered reasons for selecting Mr. Daffin over Ms. Calhoun and Ms. Pittman.

D. Intentional Discrimination

That plaintiffs have shown that some of the state’s proffered explanations for hiring decisions were pretextual does not end the matter. A finding of pretext permits, but does not require, a finding that a plaintiff has been the victim of intentional discrimination.

I find that each of these five plaintiffs was well qualified for the positions at issue. Each had a high school diploma and significant care-giving experience. Each was intelligent, poised and articulate when she testified at trial.

I find that each plaintiff would have been hired if the hiring process had been conducted without racial discrimination. I find that each plaintiff was a victim of intentional racial discrimination in each instance for which she established a prima facie case and either (1) the state failed to present a legitimate non-discriminatory reason for its decision or (2) the state’s proffered reason was pretextual. Intentional racial discrimination is the only reasonable explanation for the clear pattern of hiring decisions at Sunland during the period 1976-1984. Had it not been for that discrimination, plaintiffs would have been hired for jobs that instead went to whites.

Conclusion

The five plaintiffs now before the court were well qualified for positions for which they applied at Sunland during the 1976-1984 period. But for intentional racial discrimination, they would have been hired. Accordingly,

IT IS ORDERED:

Plaintiffs Gwenevere Long, Queen Gran-berry, Marinda Hayes, Jennifer Pittman and Erma Calhoun are hereby determined to be entitled to prevail on the issue of liability, subject to defenses relating to the filing of administrative charges. I do not direct the entry of judgment under Federal Rule of Civil Procedure 64(b). By separate notice, the clerk shall set a scheduling conference, at which the scope, sequence and timing of further proceedings, including the desirability of further mediation, will be addressed. The attorneys shall confer with one another at least five days prior to the scheduling conference in a good faith effort to reach agreement on all such issues and shall be prepared to announce any agreements on such issues at the scheduling conference.

APPENDIX 1

GWENEVERE LONG

I.Successful Claims

A. Claims for which no race neutral reason is given:

1. Cottage parent position 12031 awarded to Mary Elmore on 2/15/80

2. Cottage parent position 12014 awarded to Syble Lawrence on 10/17/80

3. Cottage parent position 12006 awarded to John Turnbow in January 1981 (The state’s exhibits concerning this position indicate that Mr. Turnbow was a veteran, but the state has presented no testimony that this was a reason Mr. Turnbow was selected. In light of Sunland’s inconsistent record with respect to preferring veterans, it cannot be assumed that this is why Mr. Turnbow was chosen.)

4. Cottage parent position 12005 awarded to Luther Medlock on or about 7/31/81

5. Cottage parent position 12052 Robert Edwards on 8/24/81

6. Cottage parent position 12004 awarded to Arthur Ingram on or about 6/11/82

7. Cottage parent positions 11972 and 12007 awarded to Evla Woods and Judy Chambliss respectively on or about 6/2/81

8. Cottage parent position 12086 awarded to Penny Crutchfield on 8/17/82

9. Cottage parent position 12050 awarded to William Pratt in March 1983

10. Cottage parent position 11928 awarded to Lori Green on 3/18/83

11. Laundry worker I position 11607 awarded to Wanda Poole on 3/24/83

12. Cottage parent position 39433 awarded to Lucille Harrison on or about 7/21/81

13. Cottage parent position 12036 filled by Voncile Murdock on or about 8/21/81

14. SSA position 11626 filled by Nettie Daniels on or about 8/14/81

15. SSA position 11655 filled by Sara Standland on or about 8/14/81

16. SSA position 11659 filled by Elizabeth Hamilton on or about 8/14/81

17. Cottage parent position 11921 filled by Rickey Jones on an unspecified date

18. Cottage parent position 12048 filled by George K. Williams on or about 9/26/80

19. Cottage parent position 12006 filled by Mary Medlock on or about 3/20/81

B. Claims for which the reason given is pretextual:

1. Cottage parent position 12017 awarded to Annie Roberts on or about 10/24/80 (alleged reason is pretextual; Ms. Long had more relevant job experience)

2. Cottage parent position 11986 awarded to Jeanette Bruner on 1/23/81 (alleged reason is pretextual: see text)

3. Cottage parent position 11946 awarded to William Pierce on 5/12/82 (alleged reason is pretextual: see text)

4. Resident training instructor position 16843 awarded to Jennie Jackson in or about October 1980 (alleged reason is pretextual; Ms. Jackson’s qualifications were not superior to Ms. Long’s)

II. Unsuccessful Claims

A. Claims for which Ms. Long fails to make a prima facie case:

1. Cottage parent position 11984 advertised on or about 5/20/82 (unclear position was ever filled, or by whom)

2. Cottage parent position 11968 awarded to Lola Daniels on 2/15/80 (Ms. Daniels is black)

3. Cottage parent position 11923 awarded to Ola Smith on 9/26/80 (Mr. Smith is black)

4. Cottage parent position 11921 awarded to Dennis Hines on 1/23/81 (Mr. Hines is black)

5. Cottage parent position 11974 advertised in July 1981 (position went unfilled)

6. Cottage parent position 12043 awarded to Ophelia Thompson on 1/16/81 (Ms. Thompson is black)

7. Cottage parent position 12054 advertised on or about 8/12/82 (position was unfilled)

8. Cottage parent position 11974 filled in July 1982 (winning applicant was black)

9. Nursing assistant position 12110 advertised on or about 6/14/83 (unclear if filled, or by whom)

10. Resident training instructor position 42224 awarded to Wilma Robinson on 8/3/82 (Ms. Robinson is black)

11. Cottage parent position 12018 awarded to Crushwell Swilley on 7/24/81 (Mr. Swilley is black)

12. Laundry worker I position 11603 awarded to Frances Calhoun on 3/24/83 (Ms. Calhoun is black)

13. Resident training instructor position 40012 awarded to Mildred McLeroy on or about 9/26/80 (Ms. McLeroy is black)

14. Resident training instructor position 40002 awarded to Mary Neal (Ms. Neal is black)

15. Cottage parent position 11959 filled by Danny Sylvester on or about 7/2/82 (Mr. Sylvester is black)

16. Cottage parent position 11964 filled by Lorraine Daniels on or about 7/2/82 (Ms. Daniels is black)

17. Cottage parent position 11977 filled by Sharon Lettinhand on or about 7/23/82 (Ms. Lettinhand is black)

18. Cottage parent position 12052 filled by Oliver Williams on or about 7/23/82 (Mr. Williams is black)

19. RTI position 16286 filled by Lucille Gilbert on or about 11/7/80 (Ms. Gilbert is black)

20. Cottage parent position 12033 filled by Lyndell Johnson on or about 4/1/83 (Mr. Johnson is black)

21. Cottage parent position 11948 filled by Patricia Evans on or about 4/1/83 (Ms. Evans is black)

22. Resident training instructor position 16297 filled by Plassie Rhynes on or about 12/5/80 (Ms. Rhynes is black)

23. Cottage parent position 12013 filled by Dorothy Brown on or about 8/14/81 (Ms. Brown is black)

24. SSA position 11656 filled by Carrie Godwin on or about 8/14/81 (Ms. Godwin is black)

25. SSA position 11658 filled by Nina Long on or about 8/14/81 (Ms. Long is black)

26. RTI position 16287 filled by Ralph Borders on or about 6/18/82 (Mr. Borders is black)

27. Cottage parent position 11931 filled by Gloria Williams on or about 6/25/82 (Ms. Williams is black)

28. Laundry worker I position 11605 filled by Edna Truitt on or about 8/13/82 (Ms. Truitt is black)

29. RTI position 32339 filled by Ruth Paramore on or about 8/27/82 (Ms. Paramore is black)

30. Cottage parent position 11919 filled by Lillie Blue on or about 3/25/83 (Ms. Blue is black)

31. Cottage parent position 11910 filled by Otis Pittman on or about 7/2/82 (Mr. Pittman is black)

32. RTI position 16298 filled by Mary Highsmith on or about 2/2/80 (Ms. Highsmith is black)

33. Cottage parent position 12043 filled by Earlene Paramore on or about 8/1/80 (Ms. Paramore is black)

34. Cottage parent position 11990 filled by Joseph Davis on or about 11/11/80 (Mr. Davis is black)

35. RTI position 16276 advertised on or about 11/11/80 (position frozen)

36. RTI position 32341 advertised on or about 12/12/80 (position readver-tized)

37. RTI position 39435 advertised on or about 12/19/80 (position frozen)

38. Cottage parent position 11938 advertised on or about 9/26/80 (position never opened)

39. Cottage parent position 12042 advertized on or about 9/26/80 (position never opened)

40. Cottage parent position 12041 filled by Donald Williams at an unspecified date in 1980 (Mr. Williams is black)

41. Service aide position 11648 filled by Gerald Williams on or about 4/20/82 (Mr. Williams is black)

42. Service aide position 11654 filled by Dorothy Mindy on or about 4/20/82 (Ms. Mindy is black)

48. RTI position 40035 advertised on or about 7/13/82 (unclear if ever filled, or by whom)

44. Cottage parent position 11944 filled by Hosie McCollough on or about 7/2/82 (Mr. McCollough is black)

B. Claims for which there is a legitimate, non-pretextual reason:

1. Cottage parent position 11909 awarded to Ricky Bragg on or about 1/16/81 (non-pretextual reasons given: existing employee preference; gender preference)

2. Cottage parent position 12055 awarded to Chaimas Stone on 1/16/81 (non-pretextual reason given: existing employee preference)

3. Cottage parent position 11943 awarded to John Hagan on 7/20/81 (non-pretextual reason given: existing employee preference)

4. Cottage parent position 12034 awarded to Virgil Rowe on 7/20/81 (non-pretextual reason given: gender preference)

5. Cottage parent position 12008 awarded to Clara Williams on 6/22/82 (non-pretextual reasons given: existing employee preference, and Ms. Long was precluded from getting this position by anti-nepotism rules because she had a relative working at this cottage)

6. Cottage parent position 11990 awarded to Larry Raines on 7/23/82 (non-pretextual reason given: existing employee preference)

7. Cottage parent position 11940 awarded to James Smith on 8/17/82 (non-pretextual reason given: Mr. Smith received stellar recommendation from current employer)

8. Nursing assistant position 12115 awarded to Michael Sellers on 9/26/80 (non-pretextual reason given: gender preference)

9. Nursing assistant position 32289 awarded to Judson Hand on 6/14/83 (non-pretextual reason given: gender preference)

10. Resident training instructor position 40011 awarded to Kathryn Neel on 9/26/80 (non-pretextual reason given: existing employee preference)

11. Resident training instructor position 40014 awarded to Ann Frick on 9/26/80 (non-pretextual reason given: existing employee preference)

12. Resident training instructor position 12137 awarded to Quentin Long on 10/24/80 (non-pretextual reason given: veteran’s preference and military record)

13. Cottage parent position 12001 awarded to Monroe Bryan on 5/12/82 (non-pretextual reason given: Mr. Bryan spent a number of years at Sunland in vocational training)

14. Cottage parent position 12041 awarded to Alva Mercer on 3/23/83 (non-pretextual reason given: Mr. Mercer had been resident training instructor for three years)

15. Resident training instructor position 32336 awarded to Edward Tyus on 11/7/80 (non-pretextual reason given: existing employee preference)

16. Resident training instructor position 32347 awarded to Addie Bellamy on 11/7/80 (non-pretextual reason given: existing employee preference)

17. Resident training instructor position 40013 awarded to Olive Green (non-pretextual reason given: existing employee preference)

18. Resident training instructor position 42225 awarded to Billy Lipford on 8/3/82 (non-pretextual reason given: existing employee preference)

19. Resident training instructor position 42212 awarded to Catherine Nichols on 3/21/83 (non-pretextual reason given: preference for hiring full-time state employees)

20. Cottage parent position 12011 filled by Martha Wade on or about 3/5/80 (non-pretextual reason given: existing employee preference)

21. Cottage parent position 11966 awarded to Johnny Baxley on 10/24/80 (non-pretextual reason given: gender preference)

22. Cottage parent position 11961 awarded to Michael King and filled by Anthony McMillan on or about 8/24/81 (non-pretextual reason given: gender preference)

23. Food service aide position 11662 awarded to Ellen Galland on 3/14/83 (non-pretextual reason given: Ms. Galland had prior food service experience)

24. Nursing assistant position 32289 filled by Kenneth Grantham on 3/9/83 (non-pretextual reason for not hiring Ms. Long: another applicant, not the winner, had superi- or qualifications)

25. Cottage parent position 11918 awarded to Hilda Smith on 3/25/83 (non-pretextual reason for not hiring Ms. Long: another applicant, not the winner, had superior qualifications)

26. Resident training instructor position 32335 awarded to Linda Wright on May 1982 (non-pretextual reason: Ms. Wright, who had been a cottage parent for over a year, was more qualified that Ms. Long)

27. Resident training instructor position 32326 awarded to Susan Weeks on 6/18/82 (non-pretextual reason: Ms. Weeks, who had been a cottage parent for over two years, was more qualified than Ms. Long)

28.Cottage parent position 12017 filled by Zadie Lewis on or about 9/19/80 (non-pretextual reason: Ms. Lewis, who had been an attendant at FSU hospital caring for the mentally ill for seven to eight years, was more qualified than Ms. Long)

APPENDIX 2

QUEEN GRANBERRY

I.Successful Claims

A. Claims for which no race neutral reason is given:

1. Cottage parent position 12107 awarded to Carlos Harris on 10/5/79

2. Cottage parent position 12030 awarded to John Dunaway on 11/23/79

3. Cottage parent position 12053 awarded to Joe Harlow on 3/14/80

4. Cottage parent position 11986 awarded to Nancy Dykes on 2/29/80

5. Cottage parent position 12052 filled by Larry Higginbotham in February 1980

6. Cottage parent position 12015 filled by Eunice Porter in February 1980

7. Cottage parent position 11966 filled by Jerry Shelby in or about July 1980

8. Cottage parent position 11932 filled by Larry Krause on or about July 1980

9. Cottage parent position 11952 filled by Robert Perdue on or about July 1980

10. RTI position 40024 filled by Betty Brogdon on or about 8/21/80

11. RTI position 40025 filled by Doris New on or about 8/21/80

12. RTI position 40026 filled by Dianne Worthington on or about 8/21/80

B. Claims for which the reason given is pretextual:

1. Cottage parent position 11968 awarded to Linwood Daffin on 11/16/79 (alleged reason is pretextual: see text)

2. Cottage parent position 12054 awarded to Rhonda Reddoch on 10/26/79 (alleged reasonds pretextual: see text)

3. Cottage parent position 12036 awarded to Carol Davis on 5/9/80 (alleged reason is pretextual: Ms. Davis had no relevant job experience and was not more qualified)

4. Cottage parent position 12031 awarded to Mary Lou Elmore on 2/15/80 (alleged reason is pretextual: see text)

II. Unsuccessful Claims

A. Claims for which Ms. Granberry fails to make a prima facie case:

1. Unspecified cottage parent position filled in April 1978

2. Unspecified cottage parent position filled in May 1978

3. Cottage parent position 11927 listed on 10/15/79 (position was reclassified)

4. Cottage parent position 12043 awarded to Earlene Paramore on 8/1/80 (Ms. Paramore is black)

5. Cottage parent position 11923 awarded to Ola Smith on 9/26/80 (Mr. Smith is black)

6. Unspecified food service aide position filled in April 1978

7. Clerk typist II position 32244 awarded to Gail Edenfield on 10/26/79 (Ms. Granberry did not prove she was qualified for this position)

8. Resident training instructor position 40012 awarded to Mildred McLeroy on or about 9/26/80 (Ms. McLeroy is black)

9. Resident training instructor position 40002 awarded to Mary Neal on or about 9/26/80 (Ms. Neal is black)

10. Unspecified cottage life position filled in September 1977 (job qualifications and race unknown)

11. Cottage parent position 12038 filled by Donald Williams on or about 10/26/79 (Mr. Williams is black)

12. Cottage parent position 11968 filled by Lola Daniels on or about 2/15/80 (Ms. Daniels is black)

13. RTI position 39383 filled by David Hunter on or about 8/21/80 (Mr. Hunter is black)

14. RTI position 40027 advertised on or about 8/21/80 (position lost)

15. RTI position 40028 advertised on or about 8/21/80 (position lost)

B. Claims for which there is a legitimate, non-pretextual reason:

1. Cottage parent position 12003 awarded to James Goodwin on 3/7/80 (non-pretextual reason given: gender preference)

2. Cottage parent position 11967 awarded to Haise Eddins on 3/7/80 (non-pretextual reason given: gender preference)

3. Cottage parent position 11977 awarded to Harry Williams on 2/29/80 (non-pretextual reason given: gender preference)

4. Cottage parent position 11959 awarded to Edgar Glisson on 8/1/80 (non-pretextual reason given: gender preference)

5. Resident training instructor position 40011 awarded to Kathryn Neel on or about 9/26/80 (non-pretextual reason given: existing employee preference)

6. Resident training instructor position 40014 awarded to Ann Frick on or about 9/26/80 (non-pretextual reason given: existing employee preference)

7. Resident training instructor position 40013 awarded to Olive Green on or about 9/26/80 (non-pretextual reason given: existing employee preference)

8. Cottage parent position 12004 awarded to Edward McCoy on 10/26/79 (non-pretextual reason given: gender preference)

9. Cottage parent position 12053 awarded to Charles Harrell on 10/26/79 (non-pretextual reason: Mr. Harrell was current, five-year employee of Florida State Hospital)

10. Cottage parent position 12011 awarded to Martha Wade on 3/5/80 (non-pretextual reason given: existing employee preference)

11. Cottage parent position 12017 awarded to Patricia Anglin on 5/9/80 (non-pretextual reason: Ms. Anglin’s experience as state employee in field of retardation)

12. Cottage parent position 12051 awarded to Sue Weeks on 8/1/80 (non-pretextual reason: Ms. Weeks’ experience and qualifications)

13. Cottage parent position 12037 awarded to Sue Ditty on 9/12/80 (non-pretextual reason: Ms. Ditty’s experience and qualifications; existing employee preference)

14. Cottage .parent position 11986 awarded on 9/28/79 to Estelle God-frey (non-pretextual reason: Ms. Godfrey’s experience and qualifications (she had been a psych, aide at Florida State Hospital for 18 years))

15. Cottage parent position 11988 awarded to Linda Williams on 3/14/80 (non-pretextual reason: Ms. Williams’s experience and qualifications (she had been a social worker’s assistant for over two years))

16. Cottage parent position 12017 awarded to Zadie Lewis on 9/26/80 (non-pretextual reason: Ms. Lewis’s experience and qualifications (she had been an attendant at Florida State hospital caring for the mentally ill for seven to eight years))

APPENDIX 3

MARINDA HAYES

I. Successful Claim,

A. Claim for which the reason given is pretextual:

1. Cottage parent position 12005 awarded to James Gesslein on 2/26/82 (alleged reason is pretextual: see text)

II. Unsuccessful Claims

A. Claim for which Ms. Hayes fails to make a prima facie c