Full opinion text
OPINION AND ORDER DUNCAN, District Judge. I. Introduction For a number of years this Court has participated in litigation concerning the City of Columbus Division of Police. In Marbaugh v. Sensenbrenner, Case No. C-2-71-391 (S.D.Ohio 1975), Haynie v. Chupka, Case No. C-2-73-401 (S.D.Ohio 1975), Brant v. City of Columbus, Case No. C-2-75-425 (S.D.Ohio 1977), and now in the present lawsuit, I am again required to confront extremely difficult issues important to the Division and our community. After having spent hundreds of hours hearing and evaluating testimony in these cases, I believe I have some means of knowing the many strengths and some of the deficiencies of the Division. In most areas the Division has provided excellent service to this community. This long history of excellence has resulted from the combined efforts of women and men who performed this often dangerous and difficult work for less than desirable compensation. Nothing in this opinion should be read to denigrate the positive contribution of such individuals. However, this case concerns a number of problem areas in the Division having to do with race. It is also noted that consideration of certain historical facts concerning race relations in the Division is mandated and appears hereinafter. The Court is well aware that times have changed, highly significant new procedures are in place in the Division, and that any unnecessary recitation of the unfortunate past may be perceived as unfair. It is not my purpose to deprecate new efforts to achieve racial fairness by blaming today’s officials for past acts of yesterday’s officials. Nevertheless, history is important for the limited purpose explained in this opinion. The trial of this case was long. Therefore, this opinion is long even though the Court has not found it necessary to comment on each and every factual and legal dispute generated by the advocates. Included in the discussion which follows is an analysis of extensive statistical evidence. Although the Court has empathy for persons who may read this opinion who are not trained in or familiar with modern statistics, there is no avoiding an appraisal of the parties’ statistical evidence. In addition, the Court must refer to administrative agency guidelines and Title VI and Title VII case law, and in doing so speak in a language which to many, may be quite arcane. With the above considerations in mind, the reasons for the Court’s decision are set forth hereinafter. II. Procedural History of the Case Plaintiffs allege discriminatory treatment on the basis of race in their employment with the Columbus Division of Police. This Court has jurisdiction of the issues raised herein pursuant to 28 U.S.C. § 1343 and 42 U.S.C. § 2000e-5. The parties at the trial in this case are as follows and will be identified as such in this opinion: Plaintiffs. This matter has been certified as a class action pursuant to Rule 23 of the Federal Rules of Civil Procedure. The class consists of those black officers, past and present, employed by the Columbus Division of Police during the pendency of this action. Plaintiff Police Officers for Equal Rights, Inc. (POER) is an Ohio not-for-profit corporation whose members include certain past and present black officers of the Columbus Division of Police. One of the organization’s avowed goals is to promote equal opportunity for all officers in the Division of Police. Individually named plaintiffs include black male and female officers who allege to personally have been, and to represent other black officers who have been, the victims of discrimination. These individually named officers include Andrea Barrett, Ronald Bosley, David Crawford, George Garrett, Charles Martin, Jodie Reeder, David Vines, Judy Stubblefield, Ollie Stubblefield, and Clyde Haynie. Many of the named plaintiffs have filed charges with the Equal Employment Opportunity Commission (EEOC). See, e.g., PI. Ex. 427, 429-436. Many of these plaintiffs have received notices of a right to sue from the EEOC. See e.g., PI. Ex. 511, 513-519. Defendants. The defendants in this lawsuit are the City of Columbus, former May- or Tom Moody, succeeded by Dana Rinehart on January 1, 1984; former Police Chief Earl Burden, succeeded by Dwight Joseph, Jr. in April 1983; Public Safety Director Bernard Chupka, succeeded by Alphonso Montgomery in 1984; members of the Civil Service Commission, Thelma Schoonover, John Young and Walter Tarpley who succeeded Earl Sherard, and the Commission's executive secretary, Earl Murry. The City of Columbus is an employer subject to the terms of Title VII of the Civil Rights Act of 1964, as amended 42 U.S.C. § 2000e et seq. More specifically, Title VII was extended to public employers, 42 U.S.C. § 2000e-17, effective March 1972. The defendant City of Columbus has been sued in this Court on two other occasions for certain violations of Equal Employment Opportunity laws. In 1975, this Court ordered the City to strive to meet minority hiring goals to rectify Division hiring practices which were found to be racially discriminatory. Haynie v. Chupka, Case No. C-2-73-401 (S.D.Ohio 1975). In 1977, this Court concluded that the Division of Police had engaged in illegal gender-based discrimination in hiring. Brant v. City of Columbus, Case No. C-2-75-425 (S.D.Ohio 1977). The instant suit was filed on April 26, 1976, by Officer Jodie Reeder and others. Thereafter, POER and additional black police officers were added as parties plaintiff. On November 29, 1979, this action was certified as a class action. This action was originally brought pursuant to Title VI and Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq.; Title 42 U.S.C. §§ 1981 and 1983 and the Thirteenth and Fourteenth Amendments to the United States Constitution. On February 28, 1984, following notice to class members and an opportunity for those class members to be heard, plaintiffs’ motion to dismiss claims brought pursuant to 42 U.S.C. §§ 1981 and 1983 and the Constitution was granted. At that time plaintiffs’ motion to file a third amended complaint was also granted. This matter proceeded to trial on March 12, 1984, on the issue of liability only. The trial concluded on April 20, 1984. The record is extensive. Over 100 witnesses were heard and over 600 exhibits were admitted into evidence. The trial transcript is lengthy to say the least. Following the trial but prior to closing arguments of counsel, plaintiffs sought a preliminary injunction restraining the promotion of seven sergeants. That motion was denied on June 14, 1984. Closing arguments were heard on July 6, 1984. This Court is now prepared to rule on this matter. Pursuant to its obligations under Rule 52(a) of the Federal Rules of Civil Procedure, the Court makes the following findings of fact and conclusions of law. III. History of Discrimination in the Columbus Division of Police A. Organization of the Division A brief explanation or description of the organizational structure of the Division may prove to be helpful in understanding the discussion that follows. Section 97 of the City Charter (Jt.Ex. 1) establishes the Department of Public Safety composed of the Divisions of Police and Fire. Prior to January 1,1981, the Division of Police maintained a set of procedural rules known as General Orders and a set of policies known as Rules of Conduct. (Jt.Ex. 3 and 5.) On January 1, 1981, the Division replaced these with a formal directives system (Jt.Ex. 7). Each officer receives copies of these directives as well as Daily Bulletins printed by the Division. These various documents will be referred to from time to time. Additional sections of the City Charter describe in detail the organizational structure of the Police Department and the powers and duties of those individuals responsible for supervision of the Division, including the Mayor, the Safety Director, and the Chief of Police. Based upon the evidence adduced at trial (Jt.Ex. 1, 3, 5 and 7), flow charts, which graphically depict the organizational structure of the Division, have been appended to this opinion. That structure from time to time has changed, the most recent change was the implementation of a “midwatch,” a shift which reports during the peak crime hours of 6:30 p.m. to 3:30 a.m. In addition, the terms used to describe the organizational structure, or parts thereof, have changed from time to time. In an effort to be consistent, the Court has undertaken to use throughout this opinion the following set of descriptive terms and designations. The Division of Police is divided into approximately five subdivisions: Administrative, Investigative, Service, Special Operations, and Field Operations. Those subdivisions report to the Executive Branch of the Division of Police which operates under the command of the Chief of Police and his staff and includes three specialized bureaus: Internal Affairs Bureau, the Legal Bureau, and the Inspector’s Bureau. Each of the five subdivisions, reporting to the Executive Branch, are further divided into Bureaus. For example, the Investigative Subdivision is divided into the Detective Bureau, the Juvenile Bureau, the Narcotics Bureau, etc. Similarly, the Administrative Subdivision, for example, is divided into the Community Relations Bureau, the Training Bureau, etc. The Field Operations Subdivision is further divided into Companies, specifically A, B and C Company, each of which represents a shift to which uniformed personnel are assigned. In some cases, Bureaus are further divided into squads or units, denoting the smallest specialized group in the Division. In the past for operational purposes, the Division of Police divided the City of Columbus into four zones which approximated the four quadrants of the city — northeast, northwest, southeast and southwest. Those zones were further divided into approximately sixteen precincts. Officers from time to time refer to substations in these various precincts. As part of a recent reorganization effort, the City has been redistricted and these zones and precincts have been altered somewhat. A map of the precincts and zones in effect at the time of this litigation and a map of the precincts as they will appear under the proposed redistricting plan are appended hereto as well. B. History of Discrimination In order to fully understand the Court’s disposition of this matter it is necessary to take a hard look at the history of the Columbus Division of Police and its past record in the area of equal employment opportunity. A number of matters discussed herein predate March 1972, the date on which Title VII became applicable to municipalities like the City of Columbus. Others predate the original 1964 passage date of the Civil Rights Act and still other incidents are more recent. All the matters which will be discussed help place in context the Court’s later discussion of the liability of these defendants. At this point the Court pauses to note that many aspects of this historical discussion are disturbing. It is not the Court’s purpose herein to unearth long-since buried skeletons or to wag an accusatory finger at the defendants. As the Court has already noted, the defendants have made many strides in the equal employment opportunity area over the years. This progress will not and should not be overlooked. Nonetheless, it is the Court’s obligation to determine whether, keeping in mind the history of the Division, the defendants’ efforts have resulted in compliance with all the requirements of the law. The evidence at trial established a pattern of overt and frequent segregation and discrimination on the basis of race in the Division of Police both prior to and following the time that Title VII became applicable to municipalities in 1972. The first black police officer was hired by the Columbus Division of Police in 1895. And while blacks have been represented on the Division since that time, the number of black officers has been significantly out of proportion to the number of black citizens in the City of Columbus. This pattern of discrimination in hiring resulted in a judicial determination of unlawful discrimination in hiring in 1975 in Haynie v. Chupka, supra. Black officers who managed to overcome obstacles to hiring encountered difficulty as members of the Division. Beginning with the testimony of former Inspector Harvey Alston, the second highest ranking officer in the Division of Police between 1954 and 1962, the evidence of instances of discrimination in the Division is overwhelming. According to Inspector Alston, whose testimony was corroborated by other witnesses knowledgeable concerning Division policies, in particular journalist John Coombs and now Deputy Chief James Jackson, blacks were historically excluded from certain assignments in the Division. Initially, black officers were relegated to foot patrol in black neighborhoods. Later, in the late 1940’s and early 1950’s when blacks were eventually assigned to cruisers, they were assigned to patrol black neighborhoods with black partners in cruisers without radios. (Tr. 3/14 at 93-99.) Blacks were not permitted to work on police vans or wagons used to transport prisoners well into the late 1960’s. The opportunities for black officers in special bureaus were virtually nonexistent. Blacks were excluded entirely from the Detective Bureau with the exception of four slots set aside for black officers on the pawn shop detail. (Tr. 3/23 at 26-30.) Many veteran officers, who testified during the course of this lengthy trial, confirmed the existence of these and other discriminatory practices in the Division. Virtually every senior black officer testified that as a matter of general practice in the 1960’s and even to some extent in the 1970’s, blacks were excluded from wagon assignments, were not permitted to work integrated cruisers and were by and large assigned to patrol in predominantly black neighborhoods. There was testimony that when Officer David Vines had occasion to make inquiry about a wagon assignment, he was told by a lieutenant that “We [referring apparently to the Division] don’t need no sambos on the wagon.” (Tr. 3/13 at 77-84.) Other black officers’ requests for wagon assignments were also refused or ignored. More disturbing perhaps is the evidence that rather than assign a black officer to a wagon if a white officer were unavailable for that assignment, there were occasions when the wagon would not be sent out at all. (Tr. 3/30 at 5-7.) Veteran officers also substantiated the existence of a Division practice of refusing to integrate cruisers and wagons or to assign a black and a white officer to the same foot patrol. When in 1969 now Deputy Chief James Jackson questioned the Division’s refusal to integrate wagons, he was told by a captain that integration would only give you a “nigger and a hillbilly.” (Tr. 3/30 at 9-11.) Because of this policy, there was evidence that black officers would be sent out on patrol alone if another black officer was unavailable to work with that officer. (Tr. 3/13 at 105-106.) Blacks encountered similar obstacles in attempting to obtain other assignments outside of patrol. For a number of years blacks were excluded from the Traffic Bureau, as well as from Academy assignments. In addition, no black has ever been assigned to the crime scene search squad. Perhaps nowhere were obstacles to black officers more apparent than in assignments to SWAT. Numerous members of plaintiffs’ class testified concerning attempts in the 1970’s to obtain SWAT assignments. A black officer, now a sergeant, attempted in 1975 to apply for an assignment to SWAT. At that time, the commander of SWAT, Lieutenant Richard Foor, later Captain Foor, told the black officer that as long as he was the SWAT Commander he would have “no niggers on SWAT” unless it was ordered by the federal court. (Tr. 4/3 at 50.) With respect to uniformed officers, historically and to some degree at present, black uniformed police officers are also assigned to black neighborhoods. The defendants claim that this assignment pattern is an attempt to meet the needs and desires of the black community and individual black officers. Certainly, as a policy matter, it may make some sense to assign certain black officers to certain black neighborhoods in the hope that a common or shared experience will facilitate the development of better community relations. The defendants, however, overlook some of the more troublesome aspects of the Division’s historical practice of assigning black officers to black neighborhoods. The difficulty with such a practice arises because of the underlying assumption that black officers must work in black neighborhoods in order to be effective without any evidence of that being the case. From the facts of this case, it appears that the historical practice of assigning black officers to black neighborhoods is an anachronism which is being abandoned by the Division. In addition to the foregoing evidence of the historical discriminatory treatment of black officers, the Court, with some reluctance, believes that it is necessary to review at least some of the substantial evidence admitted at trial concerning overt expressions of racial animus. An exhaustive treatment of the racial attitudes of the members, and in some instances supervisors, of the Columbus Division of Police will not be undertaken here. Nonetheless a few pertinent examples will be helpful in understanding the Court’s findings and conclusions in this case. The value of this evidence is not limited to its historical significance. As will be noted, many examples of racial animus are of a more recent vintage. The evidence can only be described as overwhelming with respect to the frequent and unhesitating use of the word “nigger” as well as other racially derogatory expressions. These expressions were not simply used verbally but rather they were found scrawled on desks at the training academy, on bulletin boards, in the restrooms, and on the walls of various substations. (Tr. 3/14 at 129-134; Tr. 3/15 at 35-46; Tr. 3/27 at 127-129, 188-189; Tr. 4/2 at 166, 167.) Of perhaps the greatest concern was the testimony indicating the use of racial remarks or stereotypes, as well as the appearance of racial graffiti, at the training academy. (Tr. 3/28 at 50-56.) For example, in a welcoming speech given to at least one academy class a Captain reminded the cadets that while the Division of Police was forced to accept blacks and females, they did not have to keep them. (Tr. 3/27 at 86; Tr. 4/19 at 56-57.) After leaving the academy many officers encountered further, in some cases more egregious, examples of bigotry. A black female officer testified concerning an incident in which two white officers presented her with a bunch of bananas, their not so subtle way of indicating that they believed her to be of ape ancestry. (Tr. 3/12 at 4-9.) A former white officer testified about his futile attempt to be assigned to work with a black officer, and a lieutenant’s warning to the white officer that his continued association with black officers would result in his going nowhere in the Division of Police. This warning was given in 1973.' In 1978 the white officer resigned when he learned of the promotion of Lieutenant Richard Foor to captain. (Tr. 4/3 at 60-65.) Captain Foor was the former SWAT Commander who had on a prior occasion indicated that he would not hire blacks for SWAT. Captain Foor’s promotion occurred shortly after the Division received charges concerning Foor’s racist remarks. (Tr. 4/3 at 60; Tr. 3/30 at 76.) In addition, two other white officers testified concerning their encounters with racism in the Division of Police. Both officers testified that while being given a physical exam sometime in 1980 a city-employed physician made remarks concerning his efforts to find medical reasons to disqualify blacks for Police Division service. (Tr. 3/15 at 6-11.) Both officers wrote letters concerning this incident. One officer even advised the Civil Service Commission, both verbally and in writing, about the racist remarks of the city physician. Notably, neither officer received any response to his or her letter. The Court has found its review of much of this testimony bothersome and at no time was it more so than when the Court considered the experiences of another black officer. He testified about his encounters with what plaintiffs describe as racial hazing. While he was assigned to the 8th Precinct in 1979, the black officer was frequently referred to as “token” or “brillo pad” by his fellow officers and his supervisors, including a sergeant. He was often sent on false calls, sometimes at remote locations. On one occasion after being given directions to investigate a suspicious person, the black policeman arrived at the scene only to find a statue of a black man carrying a lantern. On yet another occasion, six of the black officer’s colleagues wore white sheets and staged a cross burning in what they characterized as “a joke.” IV. Other Findings Plaintiffs allege that defendants have engaged in racially discriminatory conduct in the areas of promotions, transfers and assignments, discipline, and with respect to other terms and conditions of employment. More specifically, plaintiffs allege that defendants have engaged in a pattern and practice of discrimination and further that various policies and practices of the Division of Police have a disparate impact on black officers in the Division. With respect to promotions, plaintiffs allege that the civil service sergeants examination, as well as the five- arid later three-year time-in-service eligibility requirements are discriminatory. Plaintiffs have produced statistical evidence to support these claims. Similarly, with respect to assignments and transfers, plaintiffs allege that a three-year time-in-grade requirement in order to be eligible for receipt of a transfer disparately impacts blacks. Plaintiffs also allege a pattern and practice of excluding black officers from or limiting their number with respect to certain, much-sought-after assignments. Plaintiffs further allege a pattern and practice of discriminating against black officers in the imposition and nature of discipline imposed. Statistical evidence was adduced in support of these latter claims as well. Plaintiffs allege discriminatory treatment with respect to various other terms and conditions of employment, such as shift assignments and days off. Also, plaintiffs claim to suffer from the discriminatory effects of defendants’ no-beard policy. Finally, plaintiffs allege an atmosphere of intolerable racism which makes it difficult, if not impossible, for black officers to perform their duties. A. Role of Statistics Statistical evidence is typically utilized in class action lawsuits similar to the case at bar in an effort to establish a prima facie case of a pattern or practice of unlawful discrimination or to prove that some selection device has a disparate impact on a class of protected persons. International Brotherhood of Teamsters v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977). There can be little doubt that statistical evidence plays and will continue to play an important role in cases in which the existence of discrimination is a disputed issue. Id. at 340, 97 S.Ct. at 1856; Contreras v. Los Angeles, 656 F.2d 1267 (9th Cir.1981); Davis v. Califano, 613 F.2d 957 (D.C.Cir.1979). As the Supreme Court has acknowledged Statistics showing racial or ethnic imbalance are probative in a case such as this one only because such imbalance is often a telltale sign of purposeful discrimination; absent explanation, it is ordinarily to be expected that nondiscriminatory hiring practices will in time result in a work force more or less representative of the racial and ethnic composition of the population in the community ... Teamsters, supra, 431 U.S. at 340, 97 S.Ct. at 1856. At the same time, however, to carry evidentiary weight statistics must be relevant, material and meaningful. Moreover, a court must keep in mind considerations which may detract from the value of such evidence. In this vein, the Supreme Court has also noted that statistics come in infinite varieties and “[i]n short their usefulness depends upon all of the surrounding facts and circumstances.” Teamsters, supra, 431 U.S. at 340, 97 S.Ct. at 1856; See also Stastny v. Southern Bell Telephone & Telegraph Co., 458 F.Supp. 314, 324 (W.D. N.C.1978) aff'd. in part and rev’d. in part, 628 F.2d 267 (4th Cir.1980). The determination of the role and weight of any statistical evidence must be made on a case-by-case basis. In reviewing statistical evidence and its supporting data the Court must give consideration to and evaluate fairly conflicting opinions and hypotheses and temper any of its conclusions or opinions with common sense. S. Agid, Fair Employment Litigation: Proving and Defending a Title VII Case, 540-41 (2d ed. 1979). While some guidance in this area is available from decisions of other courts which have engaged in careful and thorough analysis of statistical evidence and supporting data, see, e.g., New York City Transit Authority v. Beazer, 440 U.S. 568, 99 S.Ct. 1355, 59 L.Ed.2d 587 (1979) and Chance v. Board of Examiners, 458 F.2d 1167, 1173 (2d Cir.1972), even those courts have been forced to acknowledge that in the final analysis the task facing the court is often difficult. In the latter cited case, the court aptly noted that After all the technical statistical jargon like ‘one-tail’ or ‘two-tail’ tests and ‘Chi Square’ ... as well as the less esoteric numbers and percentages were placed before the trial judge, it was his job to resolve the issues. Chance, supra, 458 F.2d at 1173. In the discussion which follows, it should be remembered that legal significance is not necessarily the same thing as statistical significance. See Hallock, The Numbers Game — The Use and Misuse of Statistics in Civil Rights Litigation, 23 Villanova L.Rev. 5, 12 (1977-78). And while the Court concludes that the statistical evidence in this case aids plaintiffs in their proof of a prima facie case, the Court’s analysis of and conclusions concerning the issues in this case are not based solely upon the numbers and percentages discussed hereinafter. The Court heard testimony from a number of experts in this case. Needless to say, the conclusions reached by these experts appeared to depend in large part on which side produced them. Prior to analyzing the specific conclusions reached by each of the experts, the Court will attempt to summarize the way in which each of the experts performed his analysis of the data. Thereafter, the data as it relates to each specified area of employment will be discussed in greater detail. Plaintiffs presented the testimony of two statistics experts, Dr. Charles Cranny and Dr. Wayne Villemez. Basically Dr. Cranny testified concerning the promotional examinations, the time-in-service eligibility requirements, and the job relatedness of the examinations. Dr. Villemez’s testimony, on the other hand, was concerned by and large with an analysis of the data as it related to assignments, transfers, discipline and other terms and conditions of employment. Dr. Cranny rendered an opinion based in part upon the performance of a traditional Chi-square analysis of the promotional data. Dr. Villemez calculated binomial probabilities in determining whether there was any statistically significant difference between black and white officers in the area of assignments, transfers, discipline and terms and conditions of employment. Dr. Glen Milligan, defendants’ expert witness, also performed a Chi-square analysis of data in an effort to determine any statistically significant difference in the treatment of black and white officers in the areas of promotions, transfers, assignments, discipline and terms and conditions of employment. Dr. Milligan obtained much of the data, upon which he based his analysis, from Sergeant Renshaw of the Division of Police. Dr. Cranny also used data provided by the Division of Police when analyzing the impact of the 1976, 1978,1980 and 1982 sergeants examination. Dr. Villemez obtained the data upon which he based his analysis from plaintiffs’ witness, Tim Wagner. Tim Wagner, a computer consultant, designed a data base and placed therein what he believed to be the available relevant data. Wagner obtained at least some of his data from Seniority Rosters, various assignment bulletins from 1974-1983, and incident records for currently-employed officers. Using these various materials, Wagner created what he referred to as a P-Base file in which he entered the badge number, name, and seniority date for each officer. Then Wagner created an assignment file from the assignment bulletins referred to above. Units, subdivisions, days off and shift assignments were coded to simplify the task of entering this data into the assignment file. A third file was constructed in order to record the data relating to discipline. The computer tapes, upon which Wagner placed this data, were then sent to Dr. Villemez for analysis. It should be noted that Wagner stated that he took steps to validate the data which he entered on the tapes and everything placed on the tapes was double-checked. Despite these precautions, defendants claim that there are errors and inaccuracies in the data base which render the analysis based upon such data base suspect. The alleged errors in the data base can be classified as either (1) errors which were transposed from errors in the original materials provided by defendants or (2) errors which occurred when the data was transcribed from the defendants’ material to the computer tape and then again transcribed when received by the statistician. While the Court is not unconcerned with these matters, a few pertinent observations should serve to dispel any concerns with respect to the reliability of this data generally. At various points in this opinion, the Court discusses specific problem areas with this data. First, as a general matter, insofar as errors are traceable to errors in the materials provided by defendants to plaintiffs, it is the defendants and not the plaintiffs who bear the risk of any inaccuracy in some of the materials they provided to plaintiffs. If, for example, data entered .on the computer tape based upon defendants’ interrogatory responses proves to be inaccurate, plaintiffs should not suffer since they made the effort to obtain accurate information. With respect to the second type of error, i.e., possible errors in transcription of data from defendants’ record to the computer tapes, the Court finds insufficient evidence to suggest that such errors were frequent. Certainly, there is insufficient evidence of so many errors of this sort to warrant the wholesale rejection of plaintiffs’ testimony regarding this data. B. Promotions The Court will now discuss the statistical and other evidence as it relates to specific areas of the employment relationship being challenged by plaintiffs. The first area the Court will consider is promotions. A promotion is defined in Division directives as a “change from a position in one class to a position in a different class having greater authority and responsibility and a higher maximum rate of pay.” Jt. Ex. 7. All promotions, including the selection of the Chief of Police, are made by the Public Safety Director, at his discretion, and often with the advice and counsel of subordinates. Promotions in rank in the Division are generally made in the following order: from police officer to sergeant, from sergeant to lieutenant, from lieutenant to captain, from captain to deputy chief, and finally to chief of police. The chief is selected from among deputy chiefs who have been in that rank for at least one year. Promotions to deputy chief are made from among captains who have been in that rank for one year. No civil service exams are given for the chief or deputy chief position. Captains are selected from a Civil Service eligibility list consisting of lieutenants who have been in that rank for one year and who have passed the captain’s examination. Competitive examinations are given for lieutenants and sergeants positions as well. Given the almost complete absence of black officers in the upper ranks of the Division, there was virtually no evidence available concerning the effects, if any, of these testing procedures on black officers. Most of the evidence concerning promotions focused on the sergeants promotional examination. Police officers on the force three years are permitted to sit for the sergeants examination. Until 1977, police officers were required to have five years’ experience on the force to be eligible to sit for the sergeants examination. Sergeants in rank one year may take the lieutenants examination. After an examination is administered it is graded by the examiners from the Civil Service Commission. Those passing the exam are placed on an eligibility list which is then certified by the Civil Service Commission to the Safety Director, who then may make a promotion from the top three candidates. Prior to certifying the list to the Civil Service Commission but after the grading of the examinations, seniority points are added to each candidate’s raw score. No discussion has taken place nor have any objections been raised concerning the use of seniority points and indeed any attempt to do so would be difficult at best and futile at worst. However, other aspects of the promotion and examination procedure have been the subject of substantial testimony. At the outset, it should be noted that relevant to this case are the Civil Service sergeants promotional examinations administered in 1976, 1978,1980 and 1982. Eligibility lists were certified to the Safety Director following the examinations in each of those years. 1. Statistics Re: Promotions Plaintiffs’ chief expert witness concerning the sergeants promotional examinations for each of the above-noted years was Dr. Cranny, an industrial psychologist since 1967 and an associate professor of psychology at Bowling Green State University. Dr. Cranny has substantial expertise in statistics and EEO matters and has testified and done consulting with respect to the development of selection devices in the equal employment opportunity area. Dr. Cranny examined various aspects of the promotional procedure employed by defendants, more specifically the time-in-grade eligibility requirement, the examinations, and job analysis, in an effort to determine whether any aspect of the promotional procedure adversely affected black officers. Dr. Cranny concluded that the time-in-grade eligibility requirement and the sergeants promotional examinations for the years 1976, 1978 and 1982 had an adverse impact on blacks. First, with respect to the time-in-grade eligibility requirement for taking the sergeants examination, Dr. Cranny concluded that such a requirement, five years prior to 1977 and three years thereafter, had the effect of excluding greater numbers of blacks than whites from the promotional process. No one seriously disputed this fact. In fact, there was testimony that the Safety Director, the Chief of Police and the Civil Service Commission recognized that any time-in-grade requirement would adversely impact blacks who began entering the Division in 1975-1976. Dr. Kriska testified concerning the five-year and three-year time-in-grade requirements: Q. Now, in '76, there was a five-year requirement for those who wanted to take the police sergeant’s exam; isn’t that correct? ■ A. That’s correct. Q. There was a Civil Service requirement; isn’t that true? A. That’s correct. Q. Did the Civil Service Commission ever conduct an analysis or study to determine whether that five-year requirement was related to job performance? A. The police sergeant job analysis that was conducted in 1975 addressed the issue. It was certainly not the primary focus of that job analysis, but as I recall, some legal cases were reviewed, and based on that legal review, the requirement was changed from five years to three years. Q. Okay. The legal cases were reviewed, but you did not do a job analysis to determine, or any analysis or study to determine whether the five-year requirement was related to successful— A. There is not, no analysis specifically on the Columbus police sergeant. Q. And in ’78, of course — you did change it in ’77; right? A. Correct. Q. And it was changed from five years to three years? A. Correct. Q. And the reason for that change is that you decided that five years was too high; isn’t that correct? A. I think it’s very difficult to set a cutting point like that. Administratively, the Commission does need some criteria by which to decide who is eligible to take a promotional examination and who is not eligible, but it’s very hard to say that the five years is inappropriate and three years is appropriate. Q. But you — okay. You decided that three years was appropriate in ’77? A. The Commission did, yes. Q. Based— A. It took Commission action to make that change. Q. Based on the recommendation of the Division of Police? A. I’m not sure whether the police recommended it or whether the staff recommended it. It might have been a staff recommendation. In fact, as I recall, in the job observations, the police sergeants were very, very comfortable with the five-year requirement, and I think it’s probably more a Civil Service staff recommendation that the three years be considered as, basically, a form of affirmative action. Q. Okay. Then in ’78 and ’80 and ’82, there were — was a requirement — was there not — that one had to be a police officer for three years before being eligible to take the sergeant’s examination? A. That’s correct. Q. Was there ever any study done by the Civil Service Commission to determine whether that three-year requirement was related to successful job performance? A. Not that I — only ■ the review of what's happened in other cities. Q. The review of the literature; is that correct? A. That’s correct, yes. (Tr. 128-125.) Despite the Division’s recognition of the potential and actual adverse impact of this time-in-grade eligibility requirement, defendants contend that this requirement constitutes a bona fide seniority system, thus insulating it in large part from an attack by plaintiffs. Invocation of the concept of a bona fide seniority system cannot and should not become a talismanic substitute for analysis. The Court finds no merit to defendants’ reasoning post hoe, ergo propter hoc that the time-in-grade eligibility requirement is a bona fide seniority system. First, not every time-in-grade eligibility requirement is transformed into a seniority system merely by virtue of the fact that time on the job is involved. Nowhere have the defendants identified any written seniority policy, procedure or system which incorporates this time-in-grade requirement. Finally, the evidence does suggest that if in fact a seniority system existed at all, it manifested itself when seniority was taken into account by adding seniority points to the many successful applicants for purposes of ranking those applicants. The fact that seniority points were being used in the promotional process casts serious doubt on defendants’ contention that the time-in-grade eligibility requirement was also a bona fide seniority system. In sum, the Court finds no merit to defendants’ attempt to justify the time-in-grade eligibility requirement as a bona fide seniority system despite that requirement’s adverse impact. Next, the Court turns to ah analysis of the sergeants promotional examinations for the years 1976, 1978, 1980 and 1982. Neither Dr. Cranny nor any other witness testified that the 1980 sergeants examination had an adverse impact, and the Court believes that the statistical evidence concerning that year’s examination fails to establish any discriminatory impact. Turning to the examinations for the remaining years, Dr. Cranny arrived at his opinion of adverse impact by using three statistical methods of analyzing the relevant data. The data used was derived largely from Jt. Ex. 87, 91, 92, 93. First Dr. Cranny analyzed the data using the 80% or 4/5’s rule of the Uniform Guidelines on Employee Selection Procedures (hereinafter “Uniform Guidelines”). When using the 80% rule, which is the Uniform Guidelines “rule of thumb concerning adverse impact,” one attempts to determine adverse impact of selection procedures by comparing the selection ratios of the two groups in question — in this case black officers and white officers. Stated rather simply, if the selection ratio of blacks divided by the selection ratio of whites is less than 80%, then the 80% or 4/5’s rule is violated and there is adverse impact. In this case, Dr. Cranny concluded that the 80% or 4/5’s rule was violated and that the 1976, 1978 and 1982 sergeants examinations had an adverse impact on black officers. While as has already been noted, the 80% rule is merely a “rule of thumb,” some courts suggest that violation of that rule alone is sufficient to establish a prima facie case of discrimination. Williams v. Vukovich, 720 F.2d 909, 926 (6th Cir.1983); Guardians Ass’n. v. Civil Service Commission, 630 F.2d 79 (2d Cir.1980). The Court need not decide the issue of whether violation of the 80% rule alone is sufficient to establish a prima facie case of discrimination since there is additional statistical and other evidence in this case to warrant that conclusion. In addition to the foregoing statistical analysis, Dr. Cranny performed a Chi-square analysis in an effort to determine the relationship between race and passing the sergeants promotional examinations administered in the years 1976, 1978 and 1982. Dr. Cranny also analyzed the difference between the mean scores of blacks and whites for the sergeants exams for 1976, 1978 and 1982 in further effort to determine the adverse impact of those exams. For each of the years in question — 1976, 1978 and 1982, Dr. Cranny concluded that the differences in mean scores of blacks and whites was statistically significant. Using what is known in the jargon as a t (letter “t”) statistic, Dr. Cranny stated, to paraphrase him in rather simplistic terms, that the mean scores of blacks who took the exam were lower than the mean scores of whites who took the exam thus giving the overall impression that blacks did not score as well as, and found the exam more difficult than, their white counterparts. (Tr. 3/19 at 78-80; 154-160.) Dr. Cranny’s analysis of the relative passing rates of blacks and whites also suggests that the exam in question had a disparate impact on blacks. Using rather traditional two-by-two Chi-Square tables, Dr. Cranny calculated the probability of obtaining the difference in selection or passing rates between blacks and whites by chance. For the years 1976 and 1978 Dr. Cranny found that the probability of obtaining the difference between black and white passing rates by chance was less than .05. The probability of obtaining the difference in passing rates by chance on the 1982 examination was .08, a little more than the .05 level of significance but less than the .10 level of significance. The Court pauses to note at this point that there are no hard and fast rules among statisticians as to what constitutes an absolute level of significance. Some statisticians performing Chi-square tests set .05 as the level of significance. See Smith and Abram, Quantitative Analysis and Proof of Employment Discrimination, 1981 U.Ill.L.Rev. 33 at 43. Others, however, are critical of using .05 or any other level as an absolute standard of statistical significance. Harper, Statistics as Evidence in Age Discrimination, 32 Hast. L.J. 1347, 1354 (1981); Baldus and Cole, Statistical Proof of Discrimination, § 9.2 at 101-103 (1983 Cumulative Supp.). The Court believes, without attempting to resolve the controversial debate raging among those much more knowledgeable than this Court about such matters, that generally .05 is an appropriate level at which to judge statistical significance. The Court candidly admits that its conclusion in this regard is somewhat arbitrary; nonetheless, given the lack of serious disagreement between the parties concerning this matter, the choice of a .05 level of statistical significance seems both fair and rational. With respect to the 1982 sergeants exam, the Court has noted a level of statistical significance or a probability of .08. This fact does not alter the Court’s conclusion that this examination had an adverse impact on blacks. While the probability of obtaining a difference in passing rates was slightly greater than .05, Dr. Cranny concluded, and this Court finds, that this examination had an adverse impact using the 80% rule. Moreover, the difference in mean scores between blacks and whites was statistically significant. In sum, the Court believes there is sufficient evidence of a prima facie case of discrimination with respect to the 1982 sergeants promotional examination despite a Chi-square probability for that examination which exceeds .05. The Court also pauses to note at this point that Dr. Cranny’s probability calculations are one-tail probabilities. Using a two-tail test, the probabilities discussed in the foregoing paragraphs would be approximately double those reported by Dr. Granny. In the years 1976 and 1978, even using a two-tail test of probabilities, the likelihood of obtaining the difference in passing rates by chance was still under .05. (Tr. 3/19 at 79, 144-148.) There was substantial testimony and discussion during the course of trial concerning the propriety of using a one-tail as opposed to a two-tail test of probabilities. At times the debate over this issue was reminiscent of a classic battle of the experts. Some statisticians use a one-tail test while others use a two-tail test in formulating an opinion about statistical significance. Friedman, Introduction to Statistics, 146 (Random House 1972). Statisticians themselves candidly admit that the choice is often arbitrary. W.B. Connally & D.W. Peterson, Use of Statistics in Equal Employment Opportunity Litigation (1980). Social scientists seem slightly more prone than mathematicians to the use of a one-tail test, although this is not universally true. Baldus & Cole, supra, § 9.2 at 101-103 (1983 Supp.); D.H. Kaye, The Numbers Game: Statistical Inference in Discrimination Cases, 80 Mich.L.Rev. 833, 841 (1982); H.M. Blalock, Social Statistics (McGraw Hill, 2d ed. 1972). Frankly the Court is somewhat at a loss to understand the nature of this dispute in the context of this case. Common sense would seem to dictate use of a one-tail test since the question one is attempting to answer is one directional (i.e., is there evidence that blacks are being treated less favorably than whites at a statistically significant rate?). (Tr. 3/19 at 71-75.) As the Court has already noted in the years 1976 and 1978 even using a two-tail test of probabilities, the likelihood of obtaining the difference in passing rates by chance was still less than .05. The Court has already discussed at some length, and will not repeat the discussion concerning, probabilities with respect to the 1982 examination. Despite the foregoing, defendants argue that use of the one-tail test favors plaintiffs’ viewpoint and is a method of manipulating data to prove a desired result. The Court does not find it necessary to decide whether use of a one-tail test unjustly facilitates a finding of significant results. While the Court makes reference in the pages which follow to the one-tail test performed by plaintiffs’ experts, the Court believes that substantially similar conclusions about the data can be reached by using a two-tail test. Similarly, the Court does not find it necessary to reach any conclusions concerning defendants’ argument that prior Supreme Court cases involving the use of statistics in EEO matters consistently use two-tail analysis, and that therefore two-tail tests alone are appropriate. See Hazel-wood School District v. United States, 433 U.S. 299, 308-312, 97 S.Ct. 2736, 2741-44, 53 L.Ed.2d 768 (1977); Castaneda v. Partida, 430 U.S. 482, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977). The Court finds no evidence in either of the above Supreme Court opinions suggesting that in EEO matters a two-tail analysis should be used exclusively. The Court finds substantial evidence of a scholarly or academic nature suggesting the appropriateness of using a one-tail test in circumstances • similar to those presented herein. (Tr. 3/22 at 22-25; Tr. 4/9 at 34, 204-207.) In addition to the foregoing tests of statistical significance, Dr. Cranny attempted to garner knowledge concerning the level of black representation in supervisory positions by performing a Hazelwood-type analysis. In this regard, Dr. Cranny compared the proportion of black supervisors to the proportion of blacks in Metropolitan Columbus (see Pl.Ex. 27, 72) and concluded based upon the comparison that the probability of having so few black sergeants was less than 1 chance in 10,000. (Tr. 3/19 at 168-171.) Before moving on to a discussion of the criticisms of Dr. Cranny’s analysis, it should be noted that Dr. Cranny was not alone in reaching the conclusions he reached. With respect to the 1978 examination, Dr. Kriska, chief of testing for the City of Columbus, also concluded that the difference in the mean scores and the passing rates of black and white officers was statistically significant using a two-tail test. (Tr. 3/15 at 118-121.) In fact, Dr. Kriska wrote a letter suggesting that further consideration be given to the sergeants examination since he believed the examinations “had an adverse impact on blacks.” (Jt. Ex. 90.) Dr. Wayne Villemez, plaintiffs’ other statistics expert, also examined the promotional data. Dr. Villemez, a professor at the University of Illinois, also has substantial experience in use of statistics in the EEO area. The results of Dr. Cranny’s analysis of promotional data were reaffirmed in part by Dr. Villemez’s binomial probability calculations. Dr. Villemez conducted a binomial probability analysis to determine whether, in the context of promotion decisions, there was any difference in treatment between black officers and white officers. Dr. Villemez began with the assumption that if there is equal treatment of black and white officers, one would expect that the rates of promotion for both groups would also be equal. See Davis v. Califano, 613 F.2d 957 (D.C.Cir.1979). Dr. Villemez first calculated the white promotions rate and then calculated how many blacks one would expect to be promoted if, in fact, blacks were being promoted at the same rate as whites. Dr. Villemez next compared the difference between the expected number of promotions and the actual number of promotions and calculated the probability of this difference occurring by chance, “that is, how likely is it that we could get a difference this large if, in fact the black rate is the same as the white rate.” (Tr. 3/22 at 12.) If the number of standard deviations from what one would normally expect was greater than 2 or if the probability of this difference occurring by chance reached the .05 level of significance, then Dr. Villemez considered the results or probability to be statistically significant. Stated another way, Dr. Villemez attempted to determine whether the assumption of equal treatment should be rejected because of the substantial unlikelihood of finding so great a difference in the actual number of black promotions as compared to the expected number of black promotions by chance. Based upon his calculations, Dr. Villemez rejected the assumption of equal treatment, and concluded that from 1976 until present, there was a substantial statistical disparity between the actual and expected number of blacks promoted. Defendants contend that the statistical disparity calculated by Dr. Villemez cannot be deemed to be the result of discrimination but rather is due to the large influx of black officers as mandated by this Court’s order in Haynie v. Chupka. (Pl. Ex. 72.) The defendants’ proffered justification for the statistical disparity in the rate of black as compared to the rate of white promotions may well be correct or at least a partial explanation for the disparity. Defendants argue that historically, the average time-in-grade for an officer promoted to sergeant is ten years or more. Therefore, defendants continue, since a large number of blacks did not enter the Division until 1975, less than ten years ago, one would not expect as many black sergeants. While this argument has a certain validity, it does not alone require this Court to conclude that blacks were not discriminated against in promotions. It is only one of many factors the Court has considered concerning the statistical analysis of promotions. First, assuming there is some merit to defendants’ contention in this regard, the Court remains doubtful that the influx of blacks into the department in 1975 and 1976, standing alone, would provide an adequate justification for evidence of significant statistical differences mentioned above between black and white officers. Moreover, the fact that present discrimination in the area of promotions may be due in part to prior discrimination in the area of hiring does not render the former non-actionable. In fact the present underrepresentation of black supervisors bolsters plaintiffs’ contention of unlawful discrimination even if that underrepresentation is due in part to prior discriminatory hiring practices. After being ordered to remedy discriminatory hiring practices, defendants should have been more sensitive to the need to, and made an effort to, alleviate discrimination in other aspects of employment with the Division. The conclusions reached by Drs. Villemez and Cranny with respect to the defendants’ promotional procedures were attacked directly and collaterally by Dr. Milligan, defendants’ statistics expert. Dr. Milligan, a professor of statistics at Ohio State University, clearly qualified as a statistics expert, although his experience with EEO matters was somewhat less than that of plaintiffs’ experts. Dr. Milligan, like Dr. Cranny, performed a Chi-square analysis of the promotional data except that, unlike Dr. Cranny, Dr. Milligan used two-by-five or two-by-six Chi-square tables, reasoning that factors other than race should be taken into account in calculating the Chi-square. (Df. Ex. H-M.) Dr. Milligan claimed that in performing his statistical analysis he attempted to determine whether the entire or overall promotional process operated to adversely affect black officers. In this regard, Dr. Milligan grouped together various components of the promotional process as indicated in defendants’ exhibit H-M. This Court has no way of knowing how or why Dr. Milligan selected the various components of the promotion process to be included in his analysis. For example, included in his analysis as a separate component were calculations concerning the number of blacks who failed to show up to take the sergeants exams. We have no way of knowing why that component was selected or why it was deemed an important consideration. Nor does this Court believe that this or certain other components considered by Dr. Milligan rise to the same level of importance as, for example, the passing rates oh the examinations. In sum, the Court has substantial difficulty in adopting Dr. Milligan’s analysis concerning the adverse impact of the promotional process. The Court has other difficulties with Dr. Milligan’s analysis. By grouping various component parts of the promotions process together, Dr. Milligan violated what statisticians refer to as the “assumption of independence.” The assumption of independence requires that each of the component parts grouped together for the purpose of calculating probabilities be independent of every other component part. In other words, there should not be any overlap in the column or component headings in the two-by-five and two-by-six Chi-square tables. In this case the assumption was violated since persons in certain columns, for example the column headed “# passed exam,” may also be counted again by virtue of their appearance in another column, for example, the column headed “# promoted.” In large part the Court believes that Dr. Cranny correctly identified the type of analysis which should be performed when attempting to determine the disparate impact of a promotions procedure. The Court believes that the more appropriate question to ask in cases like the one at bar is whether any of the separate components of the promotional process have an adverse impact. Since discrimination may be proven even though the “bottom line” shows no statistically significant proof of adverse treatment, Connecticut v. Teal, 457 U.S. 440, 445-456, 102 S.Ct. 2525, 2529-2535, 73 L.Ed.2d 130 (1982), the more relevant analysis in this case is whether specific component parts of the promotional process adversely impacted a protected group. Despite this Court’s conclusions concerning the need to analyze component parts of the promotional process, Dr. Milligan maintains that even if he violated the assumption of independence, his calculations represent an accurate view of the treatment of black and white officers. Dr. Milligan testified that he has proven that the violation of the assumption of independence is irrelevant because of the use of what he refers to as the “Monte Carlo technique.” Dr. Milligan was unaware, as is this Court, of any EEO case or other authority which indicates that the assumption of independence may be violated and then justified by use of the Monte Carlo technique. It is extremely difficult to understand how the Monte Carlo technique was used in this case. In sum, the Court is unpersuaded that the subsequent use of the Monte Carlo technique serves to justify the violation of the assumption of independence. The Court finds that Dr. Milligan’s results and violations of the assumption of independence may have affected the matters which he analyzed and that as a result meritorious objections have been raised concerning his statistical analysis. The Court finds that Dr. Milligan failed in his attempt to impeach or cast doubt upon the statistical analysis performed by plaintiffs’ experts in the area of promotions. Finally, the Court finds that the statistical analysis of plaintiffs’ experts is bolstered by evidence concerning individual black officers in the department. Specifically, in the last 10 years, only 4 blacks have been promoted to sergeant. Thosé blacks who received promotions to sergeant include: John Rippey (1974); Tom Hawkins (1976); David Crawford (1980); and Fred Robinson (1981). No black has been promoted to sergeant since May of 1981. Of the 43 officers promoted to lieutenant between 1974 and 1984 only one was black — Tom Hawkins, who was promoted to lieutenant in 1980. In the last 10 years the only other black promoted in the upper ranks was Deputy Chief Jackson who was promoted from captain to deputy chief. Less than 3% of those promoted in the last decade have been black. 2. Job Analyses and Validation Studies As will become apparent in the legal discussion which follows, once a court concludes that there is sufficient evidence to indicate, prima facie, that an examination or other procedure had an adverse impact on blacks, it is then incumbent upon defendants to produce sufficient evidence showing that the exam or procedure is job related. Before addressing the job analyses performed with respect to promotional exams, the Court believes a brief discussion of the alleged job relatedness of the three-year, formerly five-year, time-in-grade eligibility r