Organizations representing different members of the Houston Police Department sought to intervene in a consent decree entered into by the City of Houston and a class of black and hispanic police officers. The district court denied intervention in the underlying case and also intervention for purposes of appeal. The organizations appeal the denial of these motions and also claim that the consent decree violates Title VII and the Equal Protection Clause of the Fourteenth Amendment.

We find that the district court appropriately denied the motions to intervene in the underlying case; accordingly, we DISMISS those appeals. We find, however, that the district court erred in denying the motions to intervene for purposes of appeal. Therefore, we REVERSE the district court in this matter. Finally, we find that the consent decree survives scrutiny under Title VII and the Equal Protection Clause of the Fourteenth Amendment. Therefore, we AFFIRM the district court's approval of the consent decree.

The original Complaint in this action was filed on August 19, 1992, under Title VII of the Civil Rights Act of 1964, 42 U.S.C. Secs. 2000e et seq., as amended by the Civil Rights Act of 1991 and the Equal Employment Opportunity Act of 1972. The plaintiffs in this action filed timely charges of racial discrimination with the EEOC, alleging that the City of Houston's promotional examinations for the ranks of Sergeant and Lieutenant in the Houston Police Department discriminated against African-Americans and Hispanic-Americans between August 29, 1991 and March 26, 1992.1 Efforts to obtain redress for the Police Department's allegedly discriminatory tests began in 1975 and 1976, when Kelley v. Hofheinz, C.A. No. H-75-1536, and Comeaux v. City of Houston, C.A. No. 76-H-1754, were filed.

Among other claims, the Kelley and Comeaux cases raised the claim that the promotional examinations of the Houston Police Department discriminated against African-American police officers based on their race in violation of Title VII. These challenges to the promotional tests were based upon 1975 and 1976 EEOC charges of racial discrimination in the promotional tests. In 1979, the Comeaux action was consolidated into Kelley. In 1983, there were unsuccessful settlement discussions between the Kelley plaintiffs and the City of Houston.

On April 16, 1992, the City of Houston refused to consent to the intervention in Kelley of the Afro-American Police Officers League, the Houston Police Organization of Spanish Speaking Officers, and a group of African-American and Hispanic-American police officers. On April 17, 1992 the above groups moved for leave to intervene in Kelley, alleging that they had been harmed by racially discriminatory promotional examinations for the ranks of Sergeant and Lieutenant in the Houston Police Department, that the disposition of Kelley could impair their interests, and that in light of the passage of time, their interests were not being represented effectively in Kelley. On the same day, the City of Houston moved to dismiss Kelley for want of prosecution.

A hearing was held on June 15, 1992, on the City of Houston's motion to dismiss Kelley, and on the motion for leave to intervene. The district court dismissed all claims in Kelley for want of prosecution except for test-promotion related claims after January 1, 1982, the district court also denied the application for leave to intervene, ordered the applicants for intervention to file a new lawsuit to be transferred to the same court, directed that the remainder of Kelley be consolidated into the new lawsuit, and ordered that the new plaintiff class consist of blacks and hispanics. This case was timely filed on August 19, 1992, after receipt of Notices of Right to Sue issued by the Attorney General of the U.S. The remnant of the Kelley case was then consolidated into this action.

The plaintiffs in this action alleged that the challenged examinations had the effect of disproportionately excluding African-Americans and Hispanic-Americans from promotion to Sergeant from 1982 to date, and of disproportionately excluding African-Americans from promotion to Lieutenant from 1982 to date. They further alleged that the examinations were not job-related or consistent with business necessity. The plaintiffs sued on their own behalf, on behalf of the African-American and Hispanic-American members of the Police Department who took a Sergeant examination from 1982 to date or who will compete for promotions to Sergeant in the future.

Settlement negotiations began in earnest between the plaintiffs and defendant City of Houston in the fall of 1992. The settlement negotiations resulted in a proposed Consent Decree which was submitted to the district court in final form on January 21, 1993. On February 3, 1993, the district court ordered that notice be given to all current and former Class A Peace Officers of the City of Houston whose rights and interests were affected by the Consent Decree tentatively approved by it on that date. The notice stated that a free copy of the Consent Decree could be acquired from the Legal Services Division of the Houston Police Department. The notice also stated that March 12, 1993 was the deadline for filing objections, and that a fairness hearing was scheduled for March 24, 1993.

Before the fairness hearing was conducted, representatives of the following groups moved to intervene in the main case: Houston Police Patrolmen's Union2 ("HPPU"); the original named plaintiffs in the Comeaux action (McLoy Medlock, et al.); Houston Police Officers Association3; Female Police Officers; Asian Police Officers; and Houston Parks Police Officers.

On March 17, 1993, the district court signed an order informing the parties that a hearing on the motions for intervention was scheduled for March 22, 1993. At the hearing, the Houston Airport Police Officers Association4 made an oral motion to intervene. It later filed a written motion on March 23, 1993. After the hearing, the district court denied all of the motions for intervention, including the Airport Police Officers oral motion.

On March 24, 1993, the district court held the fairness hearing. At the fairness hearing, the applicants for intervention were allowed to: (1) cross-examine witnesses, including the plaintiffs' statistical experts and the Chief of Police; (2) proffer evidence; and (3) raise any objections to the Consent Decree. The district court again denied the motions for intervention, but allowed the applicants for intervention to file motions to intervene for purposes of appeal.

On March 25, 1993, the district court certified the following class:

a. All African-Americans who are employed, or at any time since January 1, 1982 were employed, as Class A peace officers by the Houston Police Department and who took a promotional examination for the rank of Lieutenant or for the rank of Sergeant which was administered at any time from January 1, 1982 to the present, and those who will compete for such promotions in the future; and

b. All Hispanic-Americans who are employed, or at any time since January 1, 1982 were employed, as Class A peace officers by the Houston Police Department and who took a promotional examination for the rank of Sergeant which was administered at any time from January 1, 1982 to the present, and those who will compete for such promotions in the future.

On March 25, 1993, the district court made one modification to the Consent Decree and made the Consent Decree a final judgment.5 Under the terms of the Consent Decree, African-Americans and Hispanic-Americans who took an examination for Sergeant from January 1, 1982 to date, and who passed at least one examination for this rank, will receive a total of 96 remedial promotions; African-Americans and Hispanic-Americans who took an examination for Sergeant from January 1, 1982 to date, and who were promoted after a discriminatorily long waiting period which delayed their ability to compete for Lieutenant promotions will receive five remedial promotions to Lieutenant; and African-Americans who took an examination for Lieutenant from January 1, 1982 to date, and passed at least one examination for this rank will receive a total of five remedial promotions. The named plaintiffs in this lawsuit who meet the necessary conditions shall have priority. The remedial promotions will be made over a five-year period. The district court saw this as a concession of great magnitude by the plaintiffs. The Consent Decree does not provide for any back pay.

The Consent Decree seeks, during the next ten years, to reduce the amount of adverse impact against African-Americans and Hispanic-Americans taking exams for Sergeant and Lieutenant: (a) by striking "racially biased items"; and (b) by extending the life of promotional registers during this period of time to two years.

The Consent Decree supersedes some provisions of the Fire and Police Civil Service Act and the Texas Local Government Code chapter 143, as amended. Paragraph 55 provides a mechanism for striking from the test those items which the City identifies as biased items and which are not job-related. Paragraph 55(g) provides that the promotional registers of test passers remain in effect for two years rather than one. Finally, the need to perform statistical analysis of responses and to examine test items for bias necessitates a short postponement of release of the test results. Paragraph 55(f) requires that an eligibility list for promotions be posted as soon as possible.

On May 20, 1993 the district court denied motions to intervene for purposes of appeal by the following groups: Houston Police Patrolman's Union; Houston Airport Police Officers Association; Female Police Officers; Houston Police Officers Association; and the original named plaintiffs in the Comeaux action (McLoy Medlock, et al.).

The HPPU appellants and the Houston Airport Police Officers Association appeal the district court's denial of their motions to intervene in the underlying case. The HPPU appellants, the Houston Airport Police Officers Association, the Houston Police Officers Association, and the original named plaintiffs in the Comeaux action appeal the district court's denial of their motions to intervene for purposes of appeal. The HPPU appellants and the original named plaintiffs in the Comeaux action also appeal the district court's approval of the Consent Decree. Finally, the Houston Airport Police Officers Association appeals the district court's approval of paragraph 61 of the Consent Decree, entitled "Reclassification of Peace Officers."6

1. Did the district court improperly deny the appellants' motions to intervene in the underlying case?

The HPPU appellants and the Houston Airport Police Officers Association appeal the district court's denial of their motions to intervene in the underlying case. They claim that the district court improperly denied their motions to intervene. They claimed the right to intervene pursuant to Federal Rule of Civil Procedure 24(a)7 in part based upon the Civil Rights Act of 1991 provision which prohibits subsequent attacks on a Consent Decree entered in a Title VII proceeding, through separate litigation unless the objections and complaints are raised in the original proceeding. (42 U.S.C. Sec. 2000e-2(n)(1)(B)). They allege that by refusing to permit intervention, the district court substantially limited their ability to challenge the Consent Decree. They claim that they met all the requirements for intervention of right.

The HPPU appellants assert that the principal ground relied upon by the district court for denying their motion was timeliness. They claim that their motion for intervention meets this court's test for timeliness set out in Corley v. Jackson Police Dept., 755 F.2d 1207, 1209 (5th Cir.1985) (citing Stallworth v. Monsanto Co., 558 F.2d 257, 263 (5th Cir.1977)). They first argue that they met the district court's deadline by filing their motion on March 12, 1993. They further argue that their motion was filed just 37 days after the district court published notice of the terms of the Consent Decree and less than six months after the City of Houston filed its answer denying all of the plaintiff's allegations. They also argue that although rumors of a proposed consent decree began to circulate in late December 1992, they were not able to obtain, despite repeated efforts, any specific information regarding the Consent Decree until the February 3, 1993 notice was published. They claim that it was the appellees' own conduct, in refusing to disclose the terms of the decree sooner, that caused any delay. Finally, they argue that the appellees can hardly show prejudice from any delay, since the appellees were required to give notice of the decree so that objections could be filed.

The Airport Police Officers Association ("Airport Police") also claims that its application to intervene was timely. The Airport Police argue that they sought to intervene within a few weeks after obtaining a copy of the proposed Consent Decree. The Airport Police further argue that it is undisputed that they had no possible way of knowing prior to obtaining a copy of the proposed Consent Decree that it would contain a provision, such as paragraph 61, singling them out and seemingly intended to deny them the very transfer rights they won after lengthy litigation with the City of Houston.

The Airport Police also claim that they meet the test for timeliness set out in Corley v. Jackson Police Dept., supra. The Airport Police argue that the amount of time during which they could have known of their interest in the case before seeking to intervene was minimal. The Airport Police further argue that the appellees have not shown prejudice from any delay. Finally, the Airport Police argue that their prejudice from denial of intervention is substantial.

The HPPU appellants and the Airport Police alternatively sought permissive intervention under Federal Rule of Civil Procedure 24(b).

Denial of intervention of right is a question of law which we review de novo. Ceres Gulf v. Cooper, 957 F.2d 1199, 1202 (5th Cir.1992). Denial of permissive intervention is reviewed for clear abuse of discretion. Korioth v. Briscoe, 523 F.2d 1271, 1278 (5th Cir.1975).

At the hearing on the applications for intervention in the underlying case, the district judge stated that the appellants, "were entitled to ask to intervene earlier." He also stated that the appellants could not have more time, and that he was "not going to delay this [process]." Finally, the district judge stated, "I don't need an additional party's litigant at the last moment. Your clients have a copy--or had available to them a copy of the Police Chief's December 16th memorandum8 on the potential settlement." Based on these statements, we conclude that the district court denied the applications for intervention in the underlying case on the basis of untimeliness.

"Whether leave to intervene is sought under section (a) or (b) of Rule 24, the application must be timely." Stallworth v. Monsanto Co., 558 F.2d 257, 263 (5th Cir.1977).

"Timeliness," ... is not a word of exactitude or of precisely measurable dimensions. Rule 24 fails to define it, and the Advisory Committee Note furnishes no clarification. As a result, the question of whether an application for intervention is timely is largely committed to the discretion of the district court, and its determination will not be overturned on appeal unless an abuse of discretion is shown.

Id. (citing NAACP v. New York, 413 U.S. 345, 367, 93 S.Ct. 2591, 2603-04, 37 L.Ed.2d 648 (1973)).

In Stallworth, we developed four factors to be considered in determining whether a motion to intervene was timely:

Factor 1. The length of time during which the would-be intervenor actually knew or reasonably should have known of his interest in the case before he petitioned for leave to intervene....

Factor 2. The extent of the prejudice that the existing parties to the litigation may suffer as a result of the would-be intervenor's failure to apply for intervention as soon as he actually knew or reasonably should have known of his interest in the case....

Factor 3. The extent of the prejudice that the would-be intervenor may suffer if his petition for leave to intervene is denied....

Factor 4. The existence of unusual circumstances militating either for or against a determination that the application is timely....

Id. at 264-66. "Stallworth is not an algorithm, but a framework for analysis." Corley v. Jackson Police Dept., 755 F.2d 1207, 1209 (5th Cir.1985) (quoting Lelsz v. Kavanagh, 710 F.2d 1040, 1043 (5th Cir.1983)).

As the appellees point out, Chief of Police Nuchia met with the presidents of the Houston Police Officers Association and the Houston Police Patrolmans Union on November 4, 1992. At these meetings, Chief of Police Nuchia informed the representatives of those two groups that the City of Houston was undergoing settlement discussions with the plaintiffs. Moreover, on December 16, 1992, Chief of Police Nuchia issued a departmental circular to all police officers informing them of the potential settlement in this lawsuit.

Factor 1. The appellants erroneously contend that we should consider February 3, 1993, the date the district court sent out formal notice, as the date from which to consider timeliness. As Stallworth makes clear, timeliness is measured from the point in time at which the applicant for intervention actually knew or reasonably should have known of his interest in this case. Based on the fact that Chief of Police Nuchia notified the appellants not only personally, but through the form of a departmental circular, there is no question that the appellants actually knew or should have known of their interest in this case in November or December of 1992. Therefore, the appellants waited three and a half to four and a half months before applying for intervention. Standing alone, this would probably not merit a finding of untimeliness, however, it certainly does not bode in the appellants' favor.

Factor 2. The second factor relates to the prejudice to the existing parties if the appellants' intervention is allowed. In this case, what is of particular importance is not so much the length of the appellants' delay in filing for intervention, as what occurred during the period of that delay. During this period of delay, the proposed Consent Decree was put into final conceptual form. The interests of the City of Houston, of both groups in the plaintiff class, and of other employees were considered and meshed to the greatest extent possible; the number of remedial promotions was negotiated; and the seniority complications caused by the plaintiffs' concessions to the City of Houston and to non-class-members in stretching out the remedial promotions over five years were worked out. Therefore, as in Corley v. Jackson Police Dept., 755 F.2d 1207, 1210 (5th Cir.1985), the prejudice is apparent. "A negotiated settlement of a difficult problem is put at risk, to the disadvantage of the named parties, the class, the police department and the City." Id.

Factor 3. The third factor contemplates the extent of prejudice the applicant for intervention would suffer if his motion for intervention is denied. In this case, there is no prejudice to the appellants. At the fairness hearing, the appellants were allowed to present testimony and argument on their written objections to the Consent Decree. The appellants were also allowed to cross-examine Chief of Police Nuchia and the plaintiffs' statistical experts. The district court considered all of the appellants' objections to the Consent Decree. The district court, in the Findings of Fact and Conclusions of Law, answered each of the appellants' objections and stated why it was not persuaded by their arguments. In essence, the appellants were treated as if they were parties to the lawsuit. Thus, the appellants have already been afforded the substance of the benefits of intervention as to all their objections. The appellants have had their day in court.

Factor 4. There are no unusual circumstances militating either for or against a determination that the appellants' applications were timely.

From the perspective of the Stallworth analysis, we find that the district court did not abuse its discretion in rejecting the appellants' applications for intervention. Therefore, the appeals on the denial of intervention in the underlying case are dismissed.

2. Did the district court err in denying the appellants' applications to intervene for purposes of appeal?

All of the appellants claim that the district court erred in denying their motions to intervene for purposes of appeal. The appellants claim that they meet all of the requirements to intervene as of right pursuant to Federal Rule of Civil Procedure 24(a). They claim that this court established four requirements which must be met to demonstrate a right to intervene: (1) the application to intervene must be timely; (2) the applicant must have an interest relating to the property or transaction which is the subject of the action; (3) the applicant must be so situated that disposition of the action may, as a practical matter, impair or impede, his ability to protect that interest; and (4) the applicant's interest must be inadequately represented by existing parties. New Orleans Public Service v. United Gas Pipe Line Co., 732 F.2d 452, 463 (5th Cir.) (en banc), cert. denied sub nom., Morial v. United Gas Pipe Line Co., 469 U.S. 1019, 105 S.Ct. 434, 83 L.Ed.2d 360 (1984).

The appellants all claim that their motions to intervene for purposes of appeal were timely filed. They argue that their motions were filed by the district court's April 19, 1993 deadline.

The HPPU appellants also claim that the district court, by refusing to permit intervention, substantially limited the HPPU appellants' ability to challenge the Consent Decree. They argue that because the terms of the Consent Decree authorize the City of Houston to continue to impermissibly discriminate against them, and because the Consent Decree will remain in effect for ten years under the district court's jurisdiction, the district court's refusal to allow intervention has effectively denied them their right to participate in this and future proceedings affecting their promotional opportunities.

The Airport Police argue, with regard to the second factor of the New Orleans Public Service test, that if the City of Houston complies with paragraph 61 of the Consent Decree, they will be confined to the airport and will be unable to transfer elsewhere in the City for the next ten years of their careers. The Airport Police argue that their career paths and promotional opportunities will be sharply curtailed. The Airport Police conclude, therefore, that their interest in this transaction is "legally protectable."

The Airport Police further argue that if the City complies with paragraph 61, the City will violate the state court decree which directs the City to afford each Airport Officer the same transfer rights as are afforded to Officers of corresponding rank in the Patrol Divisions of the Houston Police Department.

The original plaintiffs in the Comeaux action argue, in reference to the third factor, that the Consent Decree impairs or impedes their ability to protect their interests. They argue that their claims for years prior to 1982 were literally wiped out. They concede that these claims were dismissed in 1992. However, they claim that this was in error and that they should have been afforded relief in the Consent Decree. With regard to the fourth factor, they claim that the plaintiffs had a clear interest in dismissing the pre-1982 claims, and that the City had an interest in dismissing those claims to avoid complying with any benefits or back pay awards.

The Houston Police Officers Association argues with regard to the fourth factor, that their interests in the underlying case were not the same and to a large extent, diametrically opposite from those of the plaintiffs and defendant City of Houston. Specifically, the Consent Decree was the end result of settlement negotiations between the plaintiffs and the City only. The interests of the Houston Police Officers Association was in no way factored into the settlement agreement. The Association further argues that the remedial promotions will cause persons with lower scores to be promoted over individuals who performed better on the promotional exams.

All of the appellants alternatively sought permissive intervention under Federal Rule of Civil Procedure 24(b). The appellants argue that since their claims all arose out of the proposed Consent Decree, there were common questions of law and fact.

As stated earlier, denial of intervention of right is a question of law which we review de novo. Ceres Gulf v. Cooper, 957 F.2d at 1202. In order to determine whether a party has demonstrated a right to intervene, we examine the four requirements set out in New Orleans Public Service, 732 F.2d at 463.

First, all of the appellants' motions to intervene were timely filed. The appellants filed their motions before the district court's deadline on filing the motions to intervene for purposes of appeal.

With regard to the second requirement, the critical question is what type of interest the appellants have in the promotional system. To demonstrate an interest in the transaction sufficient to support intervention as of right, an applicant "must demonstrate a 'direct, substantial and legally protectable' interest in the property or transaction that is the subject of the suit." League of United Latin American Citizens v. Clements, 884 F.2d 185, 187 (5th Cir.1989) (quoting New Orleans Public Service, 732 F.2d at 463). The appellants' interest is in making sure that the promotional system is not manipulated in such a manner, that it discriminates against them. This interest is sufficient to support intervention as of right.

To meet the third requirement, the applicant must be so situated that disposition of the action may, as a practical matter, impair or impede, his ability to protect that interest. New Orleans Public Service, 732 F.2d at 463. Although, the appellants were treated as if they were parties in the underlying action, the district court did not allow the appellants to benefit from appellate review of the Consent Decree. The appellants were able to protect their interests below because they were able to bring forth all of their objections and arguments. However, part of the ability to protect their interests is the ability to subjugate the district court's disposition of their case to appellate scrutiny. We stated earlier that the appellants have had their day in court. Concomitant with having one's day in court is appellate review of that day.

The fourth requirement inquires whether the applicant for intervention's interest is adequately represented by existing parties. Id. The appellants' interests were adequately protected in the district court because they were given the opportunity to represent their own interests. However, when the district court denied their applications for intervention for purposes of appeal, none of the existing parties adequately represented the appellants' interests. This is evidenced by the fact that neither the City of Houston nor the plaintiffs sought appellate review of the Consent Decree. Those two parties were content with their settlement, while the appellants obviously were not.

We find that the appellants meet all of the requirements to intervene as of right for purposes of appeal. Thus, it is unnecessary to determine whether the district court abused its discretion in denying permissive intervention.

We, therefore, reverse the district court's denial of the appellants' motions to intervene for purposes of appeal.

Having found that the district court erred in this manner, we now turn to the appellants' substantive arguments.

3. Did the district court err in approving the Consent Decree?

The HPPU appellants argue that this circuit requires a district court to become more involved in the settlement process when it is called upon to approve a Title VII consent decree. They argue that the district court failed to review the terms of the Consent Decree and the evidence in support of it with sufficient scrutiny. They also argue that the Eleventh Circuit has held that the operative law for judging a consent decree is the same as that for voluntary affirmative acts. In re Birmingham Reverse Discrimination Employment Litigation, 833 F.2d 1492, 1501 (11th Cir.1987). They further argue that the district court should have used a strict scrutiny standard because the Consent Decree on its face provides for both "racial quotas" and future racial classifications. See City of Richmond v. J.A. Croson Co., 488 U.S. 469, 493, 109 S.Ct. 706, 721, 102 L.Ed.2d 854 (1989). Finally, they argue that the district court's failure to take into account the special deference allowed to police departments in determining "job related" testing, constituted an abuse of its discretion. See Davis v. City of Dallas, 777 F.2d 205, 211 (5th Cir.1985), cert. denied, 476 U.S. 1116, 106 S.Ct. 1972, 90 L.Ed.2d 656 (1986).

The HPPU appellants also argue that in a case in which the Fourteenth Amendment is properly invoked, race conscious relief may only be employed to remedy identifiable past discrimination, and relief may only be granted to the victims of that discrimination.

[F]or the governmental interest in remedying past discrimination to be triggered "judicial, legislative, or administrative findings of constitutional or statutory violations" must be made. Only then does the government have a compelling interest in favoring one race over another.

Richmond v. Croson, 488 U.S. at 497, 109 S.Ct. at 723 (quoting California Regents v. Bakke 438 U.S. 265, 307, 98 S.Ct. 2733, 2757, 57 L.Ed.2d 750 (1978)). Furthermore, the governmental entity must have a "strong basis in evidence for its conclusion that remedial action was necessary." Id. at 500, 109 S.Ct. at 725 (quoting Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 277, 106 S.Ct. 1842, 1848-49, 90 L.Ed.2d 260 (1986)).

The HPPU appellants' argue that the district court failed to require such a strong basis in the evidence. The only evidence of past discrimination is the questionable statistical analysis performed by plaintiffs' counsel. They also contend that there was no attempt by the district court to determine whether the promotional exams measured performance of skills necessary for the jobs of Sergeant and Lieutenant of police within the Houston Police Department. They further contend that the City did not make any attempt to validate the contents of the exams.

The HPPU appellants argue that the law draws a distinction between relief that is intended to remedy the effect of past discrimination and relief which is designed to compensate an individual who has been the specific victim of past discrimination. See Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 578-79, 104 S.Ct. 2576, 2587-88, 81 L.Ed.2d 483 (1983). They contend that in the instant case there was no evidence to support any finding that any of the individuals who would be receiving the 106 remedial promotions were the actual victims of any discrimination. They further contend that the Consent Decree establishes an unconstitutional pool of positions available for the promotions of individuals within the Houston Police Department based solely upon their race. See generally Bakke; Wygant; Firefighters.

The HPPU appellants also contend that the Consent Decree fails to narrowly tailor the prospective relief provided to the disparity shown and unnecessarily trammels the interests of non-Black, non-Hispanic officers within the Police Department in a way that violates the standards applicable to even non-public employers. See United Steel Workers of America v. Weber, 443 U.S. 193, 99 S.Ct. 2721, 61 L.Ed.2d 480 (1979).

The HPPU appellants further contend that the Consent Decree was collusive and/or fraudulent and/or transparently invalid. They also contend that the Consent Decree contains several provisions which violate Title VII. For example, the Consent Decree divests them of vested promotional and seniority rights under a valid merit based promotional testing procedure. Paragraph 55 of the Consent Decree authorizes the Police Department to adjust future promotional test scores on the basis of race. The Consent Decree further provides for readjustment to scores in the event the "log linear" analysis results in failing an officer who otherwise passed the exam.

Finally, the HPPU appellants argue that the district court's refusal to allow them, and others, to present the testimony they had prepared to demonstrate the questionable validity of the Consent Decree, deprived them of a fair and equal opportunity to protect their rights under the law. Thus, their equal protection and due process rights were impaired.

The Airport Police Officers claim that the district court erred in approving paragraph 61 of the Consent Decree. They contend that the legitimate purposes of the Consent Decree are to cure the ethnic bias of the promotional exams and to improve the racial mix of the Sergeant and Lieutenant ranks of the Police Department. Paragraph 61 purports to further these goals by totally excluding them from even taking the tests. They argue that excluding them from taking the tests will not make the tests less biased. Likewise, their exclusion has no relationship to improving the racial mix of the Sergeant and Lieutenant ranks.

The original Comeaux plaintiffs argue that the Consent Decree is woefully unjust, unfair, unreasonable and inadequate. Under the Consent Decree some of the class members do not receive relief, including monetary relief; the relief for the appellee-plaintiffs is significantly greater than their relief; and some members of the certified class are treated differently than other class members. They also argue that the wrongful dismissal of their claims precluded them from being included in the negotiations leading to the Consent Decree. They further argue that even where there is a clear record of delay, the sanction of dismissal is justified only where a lesser sanction would not suffice.

Under the terms of the Consent Decree, African-Americans and Hispanic-Americans who took an examination for Sergeant from January 1, 1982 to date, and who passed at least one examination for this rank, will receive a total of 96 remedial promotions. Since African-Americans suffered 64.6% of the shortfall in Sergeant promotions, they will receive 64.6% of the remedial promotions to Sergeant, for a total of 62 remedial promotions to Sergeant. Hispanic-Americans suffered 35.4% of the shortfall in Sergeant promotions, hence, they will receive 35.4% of the remedial promotions to Sergeant, for a total of 34 for remedial promotions to Sergeant. African-Americans and Hispanic-Americans who took an examination for Sergeant from January 1, 1982 to date, and who were promoted after a discriminatorily long waiting period which delayed their ability to compete for Lieutenant promotions will receive five remedial promotions to Lieutenant. African-Americans will receive three of those promotions and Hispanic-Americans will receive two of those promotions. Finally, African-Americans who took an examination for Lieutenant from January 1, 1982 to date, and who passed at least one examination for this rank will receive a total of five remedial promotions.

The Consent Decree seeks, during the next ten years, to reduce the amount of adverse impact against black and hispanic police officers taking the promotional exams by: (a) striking items biased against any race; and (b) extending the life of the promotional registers during this period to two years.

The appellants attack the Consent Decree's validity under Title VII and the Equal Protection Clause of the Fourteenth Amendment. Although the obligations of a public employer under Title VII are similar to its obligations under the Federal Constitution, they are not the same.9 We, therefore, examine each of the appellants' arguments in turn.

"In Title VII litigation, this Court had held that the district court is entitled to a substantial measure of discretion in dealing with consent decrees, and that as a result 'on appeal, our duty is to ascertain whether or not the trial judge clearly abused his discretion....' " Williams v. City of New Orleans, 729 F.2d 1554, 1558 (5th Cir.1984) (en banc) (quoting Cotton v. Hinton, 559 F.2d 1326, 1331 (5th Cir.1977)).

As a preliminary matter, we note that petitioner bears the burden of establishing the invalidity of the [Consent Decree].... Once a plaintiff establishes a prima facie case that race or sex has been taken into account in an employer's employment decision, the burden shifts to the employer to articulate a nondiscriminatory rationale for its decision. The existence of an affirmative action plan provides such a rationale. If such a plan is articulated as the basis for the employer's decision, the burden shifts to the plaintiff to prove that the employer's justification is pretextual and the plan is invalid. As a practical matter, of course, an employer will generally seek to avoid a charge of pretext by presenting evidence in support of its plan. That does not mean, however, as petitioner suggests, that reliance on an affirmative action plan is to be treated as an affirmative defense requiring the employer to carry the burden of proving the validity of the plan. The burden of proving its invalidity remains on the plaintiff.

Johnson v. Transportation Agency, 480 U.S. 616, 626-27, 107 S.Ct. 1442, 1449, 94 L.Ed.2d 615 (1987).

There is no doubt that the Consent Decree's remedial promotions take race into account. Thus, the appellants have made their prima facie case. We now turn to the validity of the Consent Decree.

The Supreme Court in Johnson articulated a two-prong test for determining whether a race-conscious affirmative action plan comports with Title VII. "The first issue ... is whether the consideration of sex [or race]10 of applicants ... was justified by the existence of a 'manifest imbalance' that reflected underrepresentation of women [or minorities] in 'traditionally segregated job categories.' " Id. at 631, 107 S.Ct. at 1451-52 (citing Weber, 443 U.S. at 197, 99 S.Ct. at 2724). "We next consider whether the Agency Plan [race-conscious plan] unnecessarily trammeled the rights of male [or nonminority] employees or created an absolute bar to their advancement." Id. at 637-38, 107 S.Ct. at 1455.

In determining whether an imbalance exists that would justify taking race into account, a comparison of the percentage of minorities in the employer's workforce with the percentage in the appropriate labor market is required. See Id. at 631-32, 107 S.Ct. at 1451-52 (citing Teamsters v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977)). Where a job requires training, the comparison should be with those minorities in the labor force who possess the relevant qualifications. Id. (citing Hazlewood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977)). The manifest imbalance requirement is not the same as a prima facie case against an employer. "A manifest imbalance need not be such that it would support a prima facie case against the employer, as suggested in Justice O'Connor's concurrence, post, [480 U.S.] at 649 [107 S.Ct. at 1461], since we do not regard as identical the constraints of Title VII and the Federal Constitution on voluntarily adopted affirmative action plans." Id. at 632, 107 S.Ct. at 1452; see also, Id. at 632 n. 10, 107 S.Ct. at 1453 n. 10 (explaining the difference between "manifest imbalance" and "prima facie").

The plaintiffs' evidence contained statistical analysis on each of the Sergeant and Lieutenant examinations from 1982 to 1991. The plaintiffs provided the district court with statistical evidence on the number of total test takers for each exam, the number of test takers by racial or cultural classification, the total number of those promoted, the number of those promoted by classification, the percentage of those promoted by classification, the promotion rates of blacks and hispanics as a percentage of whites promoted, the number of expected promotions for blacks and hispanics, the difference between actual and expected promotions for blacks and hispanics, the number of standard deviations between expected and actual promotions, and racial disparities in mean and median written scores. The District Court found that no objector presented any credible evidence that any factor other than the challenged tests, accounted for any meaningful part of the racial disparities in the promotion rates between the different racial groups.

The relevant labor market that was used in this case was the number of minority police officers who took the promotional examinations. Undoubtedly, a comparison was made to the appropriate labor market.11 We find that the plaintiffs established a manifest imbalance in promotion rates for black and hispanic officers, and that they brought this to the City of Houston's attention.

Having found that the Consent Decree meets the first requirement of the Johnson analysis, we now examine whether the Consent Decree unnecessarily trammels the rights of nonminorities or creates an absolute bar to their advancement.

In this case, the Consent Decree does not provide any remedy for blacks or hispanics who failed the promotional tests at issue. Minorities who fail the promotional tests are not eligible for remedial promotions. See Johnson 480 U.S. at 636, 107 S.Ct. at 1454 (if plan fails to take distinctions in qualifications into account, it would dictate mere blind hiring by numbers). Furthermore, the remedial promotions are being granted on a one-time-only basis. The Consent Decree does not call for any remedial promotions after the five-year phase in period. With regard to the elimination of future test questions which are racially biased, questions biased against any race will be eliminated. This will provide benefits to minority and nonminority police officers. The elimination of racially biased questions will occur only over a ten-year period. After that period of time, the City will no longer face this constraint. Likewise, the extension of promotional lists from one to two years increases the possibility of promotion for all police officers.

The Consent Decree does not require the discharge of nonminority officers and their replacement with new minority officers. See United Steelworkers of America v. Weber, 443 U.S. 193, 208, 99 S.Ct. 2721, 2729, 61 L.Ed.2d 480 (1979). Furthermore, the Consent Decree does not create an absolute bar to the advancement of nonminority officers. Id. Although nonminority officers may not be promoted at exactly the same rate and numbers as before, they will continue to be promoted in substantial excess of their representation among test takers. Finally, the Consent Decree in this case is a temporary measure. It is not intended to maintain any specific racial balance, rather it attempts to alleviate a manifest racial imbalance.

Therefore, the Consent Decree does not unnecessarily trammel on the interests of nonminorities, nor create an absolute bar to their advancement.

Since the Consent Decree is justified by a manifest imbalance that reflected the underrepresentation of minority police officers in the Sergeant and Lieutenant ranks, and the Consent Decree does not unnecessarily trammel on the rights of nonminority officers, we conclude that the district court did not abuse its discretion in approving the Consent Decree under Title VII.

Having found that the Consent Decree is valid under Title VII, we now examine its validity under the Equal Protection Clause of the Fourteenth Amendment.12

"A district court evaluating a proposed Title VII consent decree must determine whether the decree will have an unreasonable or unlawful impact on third parties if approved." Black Fire Fighters Ass'n v. City of Dallas, 19 F.3d 992, 995 (citing Williams v. City of New Orleans, 729 F.2d 1554, 1559-60 (5th Cir.1984) (en banc)). Voluntary affirmative-action plans memorialized in a consent decree are considered equivalent to voluntarily adopted affirmative-action plans for purposes of equal protection analysis. Howard v. McLucas, 871 F.2d 1000, 1006 (11th Cir.1989), cert. denied sub nom., Poss v. Howard, 493 U.S. 1002, 110 S.Ct. 560, 107 L.Ed.2d 555 (1989); In re Birmingham Reverse Discrimination Employment Litigation, 833 F.2d 1492, 1501 (11th Cir.1987). Race-conscious remedial measures receive strict scrutiny under the Equal Protection Clause. Black Fire Fighters Ass'n, 19 F.3d at 995 (citing City of Richmond v. J.A. Croson Co., 488 U.S. 469, 492-97, 109 S.Ct. 706, 721-23, 102 L.Ed.2d 854 (1989) (4-Justice plurality); Wygant v. Jackson Bd. of Educ., 476 U.S. 267, 274, 106 S.Ct. 1842, 1847, 90 L.Ed.2d 260 (1986) (4-Justice plurality). "There are two prongs to this examination. First, any racial classification 'must be justified by a compelling governmental interest.' Second, the means chosen by the State to effectuate its purpose must be 'narrowly tailored to the achievement of that goal." Wygant, 476 U.S. at 274, 106 S.Ct. at 1847 (internal citations omitted).

The Supreme Court has "insisted upon some showing of prior discrimination by the governmental unit involved before allowing limited use of racial classifications in order to remedy such discrimination." Id. (citing Hazlewood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977)).

Evidentiary support for the conclusion that remedial action is warranted becomes crucial when the remedial program is challenged in court by nonminority employees.... In such a case, the trial court must make a factual determination that the employer had a strong basis in evidence for its conclusion that remedial action was necessary. The ultimate burden remains with the employees to demonstrate the unconstitutionality of an affirmative-action program. But unless such a determination is made, an appellate court reviewing a challenge by nonminority employees to remedial action cannot determine whether the race-based action is justified as a remedy for prior discrimination.

Id. at 277-78, 106 S.Ct. at 1849.

In this case the district court specifically found that:

Plaintiffs have proven the disparate impact of the challenged examinations. Their expert testimony and other evidence created substantial doubt as to the job-relatedness of the challenged tests. It is not necessary for the Court to resolve the question whether the challenged examinations were job-related. Plaintiffs have shown a sufficiently firm basis for the relief provided in the proposed Consent Decree.

As stated in the section discussing the validity of the Consent Decree in Title VII, the plaintiffs provided detailed statistics on the promotion rates of nonminorities, blacks and hispanics. The plaintiffs also identified numerous questions on the examinations which they alleged were extremely difficult to defend as job-related and consistent with business necessity. Clearly, the district court had ample evidence from which to conclude that the plaintiffs had proven disparate impact and that the City of Houston had justifiably concluded that it would be difficult to defend the job-relatedness of the questions on the promotional exams. More importantly, the City of Houston had a strong basis in evidence to conclude that remedial action was necessary.

Having found that the Consent Decree's racial classifications are justified by a compelling governmental interest, namely remedying prior discrimination, we must now decide whether the Consent Decree is narrowly tailored to the achievement of that goal.

The Supreme Court has focused on several factors in analyzing race-conscious remedial measures: the necessity for the relief and the efficacy of alternative remedies; the flexibility and duration of the relief, including the availability of waiver provisions; the relationship of the numerical goals to the relevant labor market; and the impact of the relief on the rights of third parties. United States v. Paradise, 480 U.S. 149, 171, 107 S.Ct. 1053, 1066, 94 L.Ed.2d 203 (1987) (4-Justice plurality).

In order to evaluate the district court's determination that the remedial promotions, the elimination of racially biased test questions on future exams, and the extension of the promotional lists from one to two years, were necessary, we must examine the purposes the relief was intended to serve. Id. The Consent Decree was entered into in this case to remedy past discrimination and to alleviate the adverse impact of the promotional exams in the future. The remedial promotions are only directed to those positions where the discrimination occurred. Moreover, those promotions are only for those who most likely suffered discrimination and in direct proportion to the amount of discrimination they suffered. As the district court pointed out, Asians and women were not allowed to become part of the plaintiff class because they could not show that they were discriminated against in the promotional exams.

To determine whether the Consent Decree meets the narrow tailoring requirement, consideration of the flexibility and duration of its proposed relief, must be undertaken.

The Consent Decree proposes remedial promotions which will be stretched over a five year period, the elimination of questions biased against any race over a ten year period, and the extension of promotional lists from one to two years. The relief which the Consent Decree proposes does not continue indefinitely. Rather, it is of a temporary nature. The district court found that the plaintiffs' agreement to have the remedial promotions phased in over five years was a concession of great magnitude. The remedial promotions will be phased in over a great period of time solely to make sure that the disruption within the police department is minimized. The fact that racially biased questions will be eliminated from exams for ten years does not mar the Consent Decree, because questions biased against all races will be eliminated. Thus, all police officers will benefit. Likewise, the extension of the promotional lists to two years will help the promotional opportunities of all police officers.

With regard to flexibility, the Consent Decree does not impose rigid quotas against the police department. The Consent Decree requires only the remedial promotion of qualified blacks and hispanics. Only members of the plaintiff class who have passed the appropriate promotional exam are eligible for promotion. See United States v. Paradise, 480 U.S. 149, 177, 107 S.Ct. 1053, 1069, 94 L.Ed.2d 203 (1987) (fact that requirement to promote blacks on a one for one ratio with whites could be waived if no qualified black candidates were available, weighed in favor of the flexibility of an affirmative action plan under equal protection analysis). Hence, the police department is never required to promote an unqualified minority in preference of a qualified nonminority.

More importantly, the Consent Decree is not designed to maintain any particular racial balance. The police department is not required to increase the number of minority promotions simply because it happens to promote more nonminorities in the future.

Another method of determining whether the proposed relief is narrowly tailored, is to examine the relationship of the numerical goals to the relevant labor market. This is one of the strengths of the Consent Decree. The number of remedial promotions exactly matches the number of promotions lost by black and hispanic police officers. For example, on the September 23, 1982 Sergeant examination, the statistical analysis showed that hispanics did not suffer any adverse impact. Consequently, the hispanics were not given any remedial promotions based on this exam.

As previously stated, the relevant labor market that was used in this case was the number of minority police officers who took the promotional examinations. Unlike, other race-conscious plans which have not passed constitutional muster because they inappropriately compare the number of minorities in a specific position, regardless of the level of skill required for that position, with the number of minorities in the general labor force, the Consent Decree makes its comparison solely with the number of minorities in the police department. Clearly, the number of minority police officers who took the exam is the relevant labor market.13 Once the remedial promotions are made, the promotional relief ends, regardless of the percentage of blacks and hispanics in the Sergeant and Lieutenant ranks.

The impact upon third parties is a major aspect of this Consent Decree. A state or local government may constitutionally require innocent nonminorities to share the burden of remedying the effects of past identified discrimination. See Fullilove v. Klutznick, 448 U.S. 448, 484, 100 S.Ct. 2758, 2777, 65 L.Ed.2d 902 (1980).

It is evidently clear that the impact upon the nonminorities is negligible. The Consent Decree does not impose an absolute bar to the promotion of nonminorities. See Paradise, 480 U.S. at 184, 107 S.Ct. at 1073. More importantly, all the remedial promotions will be to newly created positions by the City of Houston. In no way are the nonminorities hurt by this Consent Decree. The nonminorities will be able to compete for exactly the same number of promotions that would exist in the absence of the Consent Decree. Their expectations and promotional opportunities are left fully intact.

As the district court stated in its opinion on denial of intervention for appeal:The remedial promotions will be to positions that the city would not have otherwise created, leaving the non-class officers the same number of promotions that they would have had in the absence of the decree. The movants could have claimed, but did not, that they should be allowed to compete as equals for every position the city creates whether it would have otherwise under current city policy. The city could not create an additional set of officer positions and gratuitously reserve them for blacks, women, or East Europeans. The positions created by the consent decree and the partial reservation are predicated on an established history of abuse, not on a claim to quotas or sensitivity to ethnicity in gross.

As to the HPPU appellants' fears that people who scored lower on the promotional exams will be promoted before their members with higher scores:

Their objections reflect confusion, at best. All of the examinations will be re-scored. The union members must compete fairly for promotions; they do not have a vested interest in continuing to receive the benefits of past discrimination.

The Airport police also cannot claim that the Consent Decree harms their ability to be considered Class A officers or to be merged with the Class A officers of the Houston Police Department. The District Court of Harris County, Texas has specifically ordered:

IT IS, THEREFORE, ORDERED ADJUDGED AND DECREED that members of the Houston Police Department Specialized Police Division, Park Police, of the City of Houston, Texas be denied entry into the uniformed and Detective Class, (Class "A") of the Houston Police Department except by application, qualification, and entry at an entry level position.

Based on the factors outlined by the Supreme Court in Paradise, we find that the Consent Decree is narrowly tailored.

Therefore, we conclude that the Consent Decree is valid under the Equal Protection Clause of the Fourteenth Amendment.

Finally, we must answer the appellants objections which did not fall under traditional Title VII or Equal Protection Clause analysis. The Consent Decree's use of "log-linear" analysis to eliminate racially biased test items does not violate Sec. 703(l ) of the Civil Rights Act of 1964, as amended by Sec. 106 of the Civil Rights Act of 1991. That section was intended to prevent the manipulation of test scores on valid employment-related tests. This provision is not applicable to the promotional tests at issue, because there is substantial doubt as to their job-relatedness.

In response to the Comeaux appellants' objections, the district court found that it would be unreasonable to delay or cancel the proposed relief for persons who actively sought to press their rights, in favor of persons who did not press their rights. As the appellees point out, the Comeaux appellants have not proffered any evidence that the failure of prosecution was exclusively the fault of their counsel.

The district court aptly stated that there can be little doubt that the federal courts have the power to enter an order overriding provisions of State or local law, where necessary, to provide an appropriate remedy in a settlement of a case in which the plaintiffs are alleging a violation of Federal law. However, the district court must take care that the provisions of State or local law are overridden only insofar as it is reasonable and appropriate to the remedy in question. United States v. City of Chicago, 549 F.2d 415, 437-38 (7th Cir.), cert. denied sub nom., Arado v. U.S., 434 U.S. 875, 98 S.Ct. 225, 54 L.Ed.2d 155 (1977). We agree with the district court that the Consent Decree impinged as little as possible upon the provisions of State law.

We conclude that the Consent Decree is valid under Title VII, the Equal Protection Clause of the Fourteenth Amendment, and all of the appellants' objections.

Therefore, we affirm the district court's approval of the Consent Decree.III. CONCLUSION

We find that the district court appropriately denied the appellants' motions to intervene in the underlying case on the basis of untimeliness. Accordingly, we DISMISS those appeals. We REVERSE the district court's denial of the appellants' motions to intervene for purposes of appeal. Since we reach the appellants' arguments as to the validity of the Consent Decree, and AFFIRM the district court's approval of the Consent Decree, the district court's denial of the motions for purposes of appeal is harmless error. Furthermore, since we have affirmed the district court in the entering of the Consent Decree and have held that the district court's denial of the motions to intervene for purposes of appeal is harmless error, the court below is AFFIRMED in all respects.14 ,15

PARKER, Circuit Judge, specially concurring.

While concurring with the majority opinion, I write separately to address some of the concerns expressed in the dissent.

The dissent professes surprise that the City would buy into the fundamental premises of the Plaintiffs' case and is deeply troubled by what the dissent characterizes as a decision of the City to use a consent decree from the federal court as a crutch to achieve what it could not accomplish as a matter of its own independent decision making. Whether we, as judges, favor these types of controversies or not, whether we might have made different decisions if we had been a party, whether or not we lament the political process that on occasion produces consent decrees in civil rights cases is of no consequence. The fact is that the parties decided to settle this case.

This nineteen year old litigation was settled in a manner that provided a limited and narrowly tailored remedy that responded to the wrongs associated with the testing provisions for advancement within the Houston Police Department. The testing provisions effectively limited advancement opportunities for blacks and hispanics. The consent decree provides remedial positions available to the group of officers adversely affected by the old testing procedures. The narrowness of the settlement is further evidenced by the fact that the consent decree takes no jobs away from non-class members and imposes no duties or responsibilities on them. The consent decree does exclude non-class members from competition for 106 remedial promotions with remedial seniority, but the remedial promotions will not prevent non-class members from being promoted on regular promotion schedules, and there will be no demotions or layoffs in order to make remedial promotions possible. Based on the total aggregate promotions to sergeant and lieutenant from 1982-91, remedial promotions would only account for about 28% of the total promotions available over the five year period of remedial promotions, or 14% of the total promotions over the life of the consent decree. While these remedial promotions do have a limited effect on promotional opportunities of non-class members, it is difficult to imagine any remedial scheme that would not have some effect. The leveling of the playing field for black and hispanic officers does not unduly trammel the interests of officers who are not members of the plaintiff class.

Our role is not to delve into the propriety of the decisions that resulted in the settlement but to address the legal issues presented in this appeal. The issues in this case are straightforward. They are (1) whether there is some procedural or jurisdictional impediment to our reaching the merits, and, if not (2) whether the district court erred in approving the consent decree.

It is entirely proper for this Court to consider the merits of the consent decree once it finds that a motion to intervene was improperly denied. Under the "anomalous rule" governing the appeal of orders denying intervention in this Circuit, the Court has provisional jurisdiction to determine whether the district court erroneously concluded that Appellants were not entitled to intervene, and if it determines that the district court properly denied intervention, appellate jurisdiction evaporates and the case is dismissed; however, if the district court was mistaken, the appellate court retains jurisdiction and reverses. Stallworth v. Monsanto Co., 558 F.2d 257, 263 (5th Cir.1977). But our reversal of the district court's denial of a motion to intervene does not end our review. If intervention was improperly denied, it is incumbent upon us to determine whether that denial was harmless error. See, e.g., Meek v. Metropolitan Dade County, 985 F.2d 1471 (11th Cir.1993) (affirming on the merits after reversing the district court denial intervention); In re Grand Jury Proceedings in Matter of Freemen, 708 F.2d 1571, 1575 (11th Cir.1983) (per curiam) (holding that failure to permit intervention was harmless where the appellate court "considered [the defendants'] claim ... as if the district court had allowed them to intervene"); Washington State Bldg. & Constr. Trades Council v. Spellman, 684 F.2d 627, 630 (9th Cir.1982) (holding that improper denial of motion to intervene did not require a new trial where proposed intervenor was permitted "to participate in the argument on the appeal from the order granting summary judgment, and its conditions were duly considered"), cert. denied 461 U.S. 913, 103 S.Ct. 1891, 77 L.Ed.2d 282 (1983); Halderman v. Pennhurst State School & Hospital, 612 F.2d 131, 134 (3rd Cir.1979) (applying harmless-error analysis to denial of motion to intervene for the purpose of appealing).

Assuming the district court acted within its discretion in denying as untimely Appellants' initial motions to intervene,1 Appellants still have a sufficient interest to have a right to appeal the decision of the district court, and it was an abuse of discretion to deny intervention for purpose of appeal. The district court's denial of Appellants' motions to intervene for the purpose of appeal cannot be sustained on the grounds of untimeliness. First, as noted in the majority opinion, the motions were filed at the invitation of the district court within the time frame for such motions set up by the district court. Second, although the initial motions to intervene were untimely under Stallworth, the Stallworth analysis does not lead to the same result with respect to the denial of the motions to intervene for appeal.

The second Stallworth factor requires that we consider the prejudice to existing parties if the appellants intervention is allowed. See Stallworth, 558 F.2d at 265. In the context of a Rule 24(a) intervention as of right, the only prejudice that we consider is that prejudice caused by a would-be intervenor's delay in filing its motion to intervene.2 Since Appellants moved to intervene for the purpose of appeal only and did not raise any issues that were not presented to the district court, it does not appear that the delay in filing their motions could have prejudiced the existing parties. Cf. McDonald v. E.J. Lavino Company, 430 F.2d 1065 (5th Cir.1970) (holding that, although post-judgment motions to intervene are normally viewed with a "jaundiced eye," the timing of post-judgment motion to intervene for the limited purpose of staking a claim to the proceeds of suit could not prejudice existing parties). The existing parties would have been in exactly the same situation if Appellants had filed their motions to intervene for appeal only on the day this case was filed.

Perhaps more importantly, 42 U.S.C. Sec. 2000e-2(n) creates an "unusual circumstance" under the fourth factor of the Stallworth analysis. Under 42 U.S.C. Sec. 2000e-2(n), if Appellants were denied entirely the right to intervene in this action, they would be precluded from raising any challenge to the consent decree. The fact that the district court initially denied Appellants motions to intervene without prejudice to refiling for the purpose of appeal removed the effect of 42 U.S.C. 2000e-2(n) from the "unusual circumstances" in factor four of the Stallworth test. Cf. NAACP v. New York, 413 U.S. 345, 368, 93 S.Ct. 2591, 2605, 37 L.Ed.2d 648 (1973) (absence of special circumstances warranting intervention illustrated by the fact that appellants were free to renew their motion to intervene following the entry of summary judgment). However, when the district court denied Appellants' motion to intervene for appeal, it foreclosed the possibility of any further challenge or appellate review of the consent decree. This altered the balance of the Stallworth factors to the extent that it was an abuse of discretion to deny Appellants motions to intervene for the purpose of appeal.

If, as the dissent contends, intervention for the purpose of appeal were not permitted by the Rules,3 it would have been error for the district court to deny Appellants' initial motions to intervene because of 42 U.S.C. Sec. 2000e-2(n)'s effect on the fourth Stallworth factor. We would still reach the merits of Appellants' claims.

The district courts' denial of Appellants' motions to intervene for the purpose of appeal had no practical effect on their ability to present their positions on appeal. Although Appellants' motions to intervene for the purpose of appeal were denied, Appellants did present their objections to the consent decree to this Court. Though not "allowed" to participate in the appeal by the district court, Appellants did participate in the argument on appeal and presented their objections to the consent decree to this Court, and their arguments were duly considered.

Appellants' situation with respect to the proceedings before the district court was similar. Appellants had their day in court; they had the amount of process they were due. The dissent's view of what constitutes one's "day in court" would preclude consent decrees and would apparently mandate that these type cases all be tried. The dissent's position appears to be that if one attains the status of intervenor, then he can effectively thwart any consent decree by being entitled to discovery, the presentation of evidence and witnesses, the right to a decision by a judge or jury, and the right to appeal. Such a narrow view of fairness hearings in the context of consent decrees is without discernable authority.

Appellants have no substantial grounds to complain regarding their participation in the fairness hearing. If Appellants had been granted intervenor status they would have been entitled to present evidence and have their objections heard at the hearings on whether to approve a consent decree, but they would not have the power to block the decree merely by withholding their consent. Local Number 93, International Asso. of Firefighters, etc. v. Cleveland, 478 U.S. 501, 529-30, 106 S.Ct. 3063, 3079, 92 L.Ed.2d 405 (1986). Although Appellants were not technically permitted to intervene in the district court action, they were allowed to introduce evidence and cross-examine witnesses at the fairness hearing. Appellants were permitted to air their objections to the reasonableness of the decree and to introduce relevant evidence; the district court considered these objections and explained why it was rejecting them. See Cleveland, 478 U.S. at 529, 106 S.Ct. at 3079. Neither intervenors nor objectors are entitled to hold consent decrees hostage and require a full-blown trial in lieu of a fairness hearing. The issue here is how much process appellants were due. Even if appellants had been designated as intervenors, they still got the process they were due.

The district court afforded Appellants de facto intervenor status at the fairness hearing and found their objections unconvincing. Appellants have not shown that the district court was wrong or that they would have been any more convincing at the fairness hearing if they had been designated intervenors rather than objectors. Appellants suffered no prejudice by the district court's failure to designate them intervenors apart from denying them their right of appeal on the merits. We have remedied that error.

For the reasons set out in the majority opinion and in this special concurrence, I agree with the majority that we have jurisdiction to review the merits of Appellants' claims. After conducting such a review, I agree that any error in denying Appellants' motions to intervene was harmless. Although I agree that the district court acted within its discretion in denying Appellants' initial motions to intervene, I do not believe that that determination is critical to the outcome of this case; if the initial motions to intervene were improperly denied, I would reach the same result.

With some of the trepidation that I suspect was in the mind of David when he stepped forward to face Goliath, I write to express my dissent and disagreement with the analysis and conclusions reached by my distinguished colleagues in the foregoing opinion.

I disagree with the panel majority's extrapolation from the record that the district judge denied the motions for intervention because of "untimeliness." In my view, if the district judge articulated any reason for his denial of the motions for intervention, it was because he felt that the right to object at the fairness hearing was all he was obligated to afford to the would-be intervenors.

I dissent and disagree with the majority panel conclusion that the district judge's denial of the motion for intervention because of "untimeliness" was not clearly erroneous. In my view, such a ruling is clearly inconsistent with and not supported by any prior case law nor by an appropriate construction of the newly adopted amendment to the Civil Rights Act which speaks to this question.

I dissent and disagree with the panel majority's conclusion that denial of a motion to intervene "for appeal only" is a final judgment which vests this court with appellate jurisdiction, when a prior motion to intervene in the main case has been denied and sustained on appeal by the court. In my view, there is nothing in the Federal Rules of Civil Procedure which contemplates a motion at the trial court level to intervene for appeal purposes only and there is nothing in the newly enacted amendment to the Civil Rights Act which evidences an intention to create an appeal from the fairness hearing.

Some history is necessary to put this case and my dissent in perspective. Prior to the Supreme Court's decision in Martin v. Wilks, 490 U.S. 755, 109 S.Ct. 2180, 104 L.Ed.2d 835 (1989), most federal courts of appeal had adopted a very restrictive rule precluding all challenges or collateral attacks on a Title VII consent decree once it had been entered by a court.1 These courts concluded (i) that the effectiveness of such decrees as a mechanism to settle claims short of or early in the litigation process would be substantially undermined, if not eliminated, if the decrees were subject to perpetual challenge; and (ii) that such decrees have played a beneficial role in providing relief to thousands of victims of systemic employment discrimination.2 However, in Wilks, the Supreme Court, in rejecting the "impermissible collateral attack doctrine" of the circuit courts, found that (i) one of the clear principles of Anglo-American jurisprudence is "that one is not bound by a judgment in personam in a litigation in which he is not designated as a party or to which he has not been made a party by service of process;" (ii) that one of the rules deeply rooted in our historic tradition is "that everyone should have his own day in court;" and (iii) that the law does not impose on anyone the burden of voluntary intervention in a suit to which he is a stranger; and that "[u]nless duly summoned to appear in a legal proceeding, a person not a privy may rest assured that a judgment recovered therein will not affect his legal rights." Wilks, 490 U.S. at 762, 109 S.Ct. at 2184-85. As the Supreme Court recognized, both FRCP 24(a) covering intervention as a matter of right, and 24(b) covering permissive intervention, are cast in "permissive terms."3 Concerns for finality and completeness of judgments are "better [served] by [the] mandatory joinder procedures" set forth in Rule 19(a). Id. Finally, "the linchpin of the 'impermissible collateral attack doctrine'--the attribution of preclusive effect to a failure to intervene--is therefore quite inconsistent with Rule 19 and Rule 24." Id. at 764, 109 S.Ct. at 2186.

The U.S. Congress, however, concluded that the Supreme Court had gone too far in Wilks, which in effect required mandatory joinder of all parties whose interests could potentially be affected, and the Congress therefore set upon the task of devising a remedial scheme which would "balance the rights of non-litigants against the need for finality of judgments and prompt relief for discrimination." H.R.Rep. No. 40(I), 102d Cong. (1991), reprinted in, 137 Cong.Rec. 588 (1991), U.S.Code Cong. & Admin.News 1991, 549, 588. That balance is best achieved when all interested parties are allowed to participate in a single proceeding. As the Judiciary Committee stated:

[T]he Wilks decision imposes inefficient and inequitable rules on Title VII litigation. Once an employment dispute has reached the courts, the parties, all non-litigants with a stake in its outcome, and the public have a strong interest in bringing the litigation to an expeditious end. Thus, all related interests and claims should ordinarily be adjudicated in one proceeding. (emphasis added). H.R.Rep. No. 40(I), 102d Cong., reprinted in 137 Cong.Rec. 591 (1991), U.S.Code Cong. & Admin.News 1991, 591.

To accomplish this goal Congress passed Public Law 102-166, Section 108, which is now codified at 42 U.S.C. Sec. 2000e-2(n) and became effective January 21, 1991.

Subsection n(1) defines the circumstances under which subsequent challenges to a consent decree will be precluded. It is similar in content to the pre-Wilks case law, although it imposes some new requirements (i.e., "actual notice of the proposed judgment") which were not requirements of that prior law.

Subsection n(2) sets forth the various circumstances in which subsequent attack on a Title VII consent decree will not be precluded. Congress expressly stated its intent to include all relevant parties in Subsection n(2)(A) which provides that nothing in the Subsection is intended to "alter the standards for intervention under Rule 24." Although the majority opinion makes no mention of the effect these amendments, I believe that we should reconcile and address their effect on this suit.

With this general legal background in mind, I need to add some comments about this particular lawsuit. It is a "disparate impact case" and not an "intentional discrimination case." It was filed August 19, 1992. As the majority opinion indicates, two previously existing lawsuits, Kelly and Comeaux were consolidated into this case. Kelly, filed in 1975, and Comeaux, filed in 1976, both included assertions of disparate impact similar to those in the current case. The City of Houston filed an answer in this case on September 30, 1992, and denied generally the allegations of disparate impact in this case, just as it had done in the prior cases of Kelly and Comeaux. The disparate impact is alleged to occur "in racially discriminatory promotional examinations for the ranks of sergeant and lieutenant in the Houston Police Department." The plaintiffs sought certification by the trial court of two classes of plaintiffs, African American and Hispanic American members of the police department who took examinations for sergeant and lieutenant during the years in question. As I understand the record, the process for promotion in the Houston Police Department consists of the following steps:

1. The police officers who have served a defined number of years in a lower grade are eligible to take the examination for promotion to a higher grade.

2. The officers who take and pass the promotional examination with at least the established passing grade are eligible for promotion.

3. All eligible officers are then placed in rank order on a list of those eligible for promotion in accordance with the sum of:

a. their actual score on that examination, plus

b. additional points, based on their years of service with the police department.

4. As vacancies occur (resulting from death, withdrawal, retirement, resignation, or from expansion of the force by action of the City Council) these vacancies are filled from the top of the list of eligibles downward until all vacancies are filled.

This system produces the following general results:

1. Individual police officers who make the highest scores on the examination and have the largest number of points for years of service are the first ones to be promoted; and

2. The number of promotions actually made is a function of the number of vacancies and not the number of eligibles on the list.

The plaintiffs tendered various statistical studies which showed that comparing "test takers" and "numbers of promotions" for each of the categories of whites, blacks, and hispanics, there was a variation in promotion rates for blacks and hispanics sufficiently large to rule out chance as the determining factor, and therefore there was an adverse discriminatory impact. This adverse impact was attributed by plaintiffs to certain questions (not identified either by number or by content) on the examinations, which blacks and hispanics were alleged to have answered incorrectly more frequently than whites. However, in my judgment, the theory upon which the plaintiffs plead and supported their claims (as indicated by paragraphs 24-39 of the findings of fact and conclusions of law which the plaintiffs prepared and persuaded the trial judge to sign in this case) suffer two fundamental conceptual breakdowns which render their claims questionable:

A. First of all, to prove "disparate impact" in the giving and taking of an examination, one must compare "test takers" and "test passers." Not only did the plaintiffs not proffer any statistical testimony based on this comparison, they alibied out of the task (and got the district court to sign a finding) on the grounds that the expense of making such a determination was prohibitive.

B. Secondly, to prove "disparate impact" in the promotion system used by the city, you must compare the number of individuals eligible for promotion with the number of individuals actually promoted in each of the racial categories.

As I explained earlier, to be eligible for promotion, each police officer, of all racial categories, must achieve the same minimum score; and it is important to note in this case that there is no contention that the minimum passing score for blacks and/or hispanics was any different from that for whites. But in their statistical data, for purposes of comparison with the actual promotions in each minority category, the plaintiffs utilize a concept of "availability" (the percentage of blacks and hispanics of all test takers ) as the starting point for determining a concept of "expected promotions" (the number of minorities who would be promoted if you applied the "availability" percentage to the total number of actual promotions of all categories.) Clearly, in my judgment, the factor of "available percentage" should have been determined by determining the percentage of each minority category of all "test passers" (not test takers ) and the number of "expected promotions" would then be the percentage attributable to each minority group which were eligible for promotions as applied to the total number of promotions actually made.

If the statistical data incorporated in the findings of fact was the same data presented to the City of Houston in the settlement negotiations, I am genuinely surprised that the city would buy into the fundamental premises of the plaintiffs' case. Obviously, the city had faced similar "disparate impact" contentions in the Kelly and Comeaux cases for more than ten years, and had consistently declined to recognize the validity of such contentions. The net results of the plaintiffs' claims and the city's capitulation thereto, is a determination of a "short fall" in promotions to each of the minority groups and a determination to award "remedial promotions" to each of these minority groups on the basis of such "short fall." There is, however, no evidence or findings which can determine any individual as to which these remedial promotions should be granted. In fact, under the consent decree, these promotions will be awarded to whomever the city chooses from among the two minority groups. I am deeply troubled by the decision of the city to use a consent decree from the federal court as a crutch to achieve that which it could not accomplish as a matter of its own independent decision making, i.e., expand the composition of the police department by 96 new sergeants slots and award those slots on the basis of race.

At about this point, various other groups of police officers sought to intervene in this case. Some of these groups are representative of other minorities, i.e., women and Asians. Others of these groups represent associations and unions of police officers. Some of these groups sought to intervene as plaintiffs, alleging that they too had been victims of "disparate impact" in the promotion process and that they should be entitled to share in the assignment of the new sergeant and lieutenant slots authorized by the city council. Other groups sought to intervene, in effect on the side of the City of Houston, as defendants, asserting that the examinations given for promotions were legitimate "job related" examinations, and consistent with the "business necessity" and professional desire to see that the most qualified individuals were promoted to positions of leadership. Some of these latter groups plead and proffered testimony that would show that the statistical data upon which plaintiffs contended "disparate impact" had occurred was flawed in concept or factually insupportable. And these contentions bring us to what I consider to be the "gut" issue in this lawsuit, i.e., "Did the trial court properly dispose of these motions to intervene?"

The trial court set a hearing to consider the motions to intervene. This hearing began at 9:30 a.m., continued without a recess, and concluded after generating only 65 pages of transcribed testimony. Approximately the first 30 pages of this transcription set forth a dialogue between the trial judge and plaintiffs' counsel regarding the nature and content of plaintiffs' asserted discrimination claims and the remedy provisions of some of the aspects of the proposed consent decree. About mid-way through the hearing, counsel for one of the would-be intervenors (the Houston Police Patrolmen's Union) asked the court to direct its attention to the issue of intervention, which was the purpose of the hearing, and twice advised the court regarding the applicability of 42 U.S.C. Sec. 2000e-2(n) as it relates to the subject of intervention. The trial court, however, ignored such references to subsection (n). Furthermore, at this hearing, supposedly on the subject of intervention, the court made no reference to Fed.R.Civ.P. 24(a), 24(b) or to any of the factors relating to the subject of intervention under those rules. There was no mention of the Stallworth case, nor the four factors for determining a question of timeliness thereunder. At the conclusion of the hearing, the court announced its ruling from the bench as follows:

The motions to intervene are denied without prejudice to them being reasserted after the objections are heard for purposes of appealing the decision itself as opposed to any ruling of the preparation of this issue. [sic]

Three days after the intervention hearing the district judge filed a "order on intervention" which read, in its entirety, as follows:

1. The motions to intervene in this case as full active parties representative by women, Asians, the airport police and park police, and by Houston Police Patrolmen's Union and Park Police Association are denied.

2. The court will consider motions to intervene for purposes of appeal. These motions must be filed by April 19, 1993.

The district judge did not file any me