ORDER
MYRON H. THOMPSON, Chief Judge.
Twice now, without making any reasonable effort at actual implementation, defendants
(the Alabama State Board of Education, its members, and the State Superintendent of Education) turned to this district court to attempt to get of out a settlement, embodied in a 1987 consent decree, governing teacher testing in the State of Alabama. The first time, this court relented and pitched the settlement,
Allen v. State Bd. of Educ.,
686 F.Supp. 64 (M.D.Ala.1986.) (Thompson, J.), but, on appeal, the Eleventh Circuit Court of Appeals disagreed and said that State must abide by its word.
Allen v. State Bd. of Educ.,
816 F.2d 575 (1987).
The second time, in compliance with the appellate mandate, this court issued a memorandum opinion and order on September 8, 1997, rejecting defendants’ request that the court vacate or modify the consent decree.
Allen v. State Bd. of Educ.,
976 F.Supp. 1410 (M.D.Ala. 1997) (Thompson, J.). The court stated that “defendants have not made a good-faith effort to develop a test that both meets the requirements of the consent decree and is psychometrically sound or even to find out whether such a test can be developed.”
Id.
at 1414-14.
This matter is again before the court, this time on motions for further relief and for a preliminary injunction filed by plaintiffs. Plaintiffs (a predominantly. African-American university and three African-American teachers, who represent a class of all black persons who have been or will be denied any level teacher certification because they failed to pass a test administered by the State of Alabama under the Alabama Initial Teacher Certification Test) contend that defendants have taken, and are taking, actions in conflict with the consent decree. Ironically, for essentially the same reasons the court gave in rejecting defendants’ motion to modify and vacate the consent decree, the court rejects plaintiffs’ motions.
I.
Plaintiffs point to recent State legislation which, they contend, conflicts with the consent decree and reflects that defendants are poised to engage in acts in conflict with the decree.
During the 1995 session, the Alabama Legislature enacted a law requiring that teacher candidates pass a “nationally normed” examination as a condition for graduation. The intent behind the legislation, according to its preamble, is “To require the State Board of Education to review requirements for teacher education; to cause colleges and universities which offer a teacher preparation program to require as a condition for graduation that their students preparing to be teachers pass a nationally normed teacher education test; to authorize the State Board of Education to approve the use and determine the level deemed to be a passing of the test; to require out-of-state graduates to pass the said nationally normed test before being certified.” The legislation, as later codified, provides:
“(a) The Legislature finds a compelling interest in adequately preparing teachers to teach. The two basic components of teaching consist of knowing what to teach (content) and knowing how to teach (methodology). To assure that teachers entering Alabama’s classrooms have been thoroughly prepared, the Legislature directs the State Board of Education to review the requirements of programs for teacher education and preparation and select a nationally normed teacher examination to be used. The nationally normed teacher examination may include the National Teacher Exam or any other national equivalent exam. Colleges and universities shall require each teacher candidate to pass the nationally normed examination as a condition of graduation. Notwithstanding any other provisions of this section, a person can, at any time, elect to not be a teacher candidate. The State Board of Education shall prescribe the manner in which the examination is to be administered and shall further determine the standard to be used for passing the examination.
“(b) Graduates of teacher education programs who apply for certification who have attended out-of-state colleges and universities must also pass the same nationally normed teacher examination as required of
students attending teacher preparation programs in Alabama. The examination shall be administered in a manner prescribed by the State Board of Education. The same standard used for passing the examination for students enrolled in teacher preparation programs at colleges and universities in Alabama shall be applicable. “(c) This Act shall become effective for teacher education candidates who enter their program effective August 1, 1995.”
1975 Ala.Code § 16-3-16.1. Part of the legislation, not codified, provides: “The provisions of this Act are severable. If any part of this Act is declared invalid or unconstitutional, such declaration shall not affect the part which remains.”
With their motions, plaintiffs ask that defendants be enjoined from implementing this law. Plaintiffs maintain that a “nationally normed” test — and, in particular, the National Teacher Exam, commonly referred to as the NTE — would conflict with the 1987 consent decree as interpreted by this court in its memorandum opinion of September 8, 1997. Plaintiffs read too confiningly the proscriptions of the consent decree, as interpreted in the September 8 opinion.
A.
It must be remembered that, in issuing the September 8 memorandum opinion, this court was confronted not with just a
court order
but rather with a
promise
memorialized in an order. This distinction has several important implications. The first is one whose significance, unfortunately, was only implicitly acknowledged in the September 8 opinion: that, at the time of the settlement that led to the consent decree, the parties had not developed an evidentiary record, and the court had no role in the development of that decree. Indeed, as stated, this court initially rejected, the consent decree, and the consent decree was later reinstated by the Eleventh Circuit. The decree was therefore not the product of a trial, that is, of a joint effort of the parties and the court. The consent decree is therefore not buttressed by detailed factual findings and observations. The enforcement provisions in the decree are prefaced by only the following:
“The individual plaintiffs and the class they represent challenge in this litigation the defendants’ implementation and use of the Alabama Initial Teacher Certification Tests. The class consists of all black persons who have been or will be denied any level teaching certificate because of their failure to pass the tests administered by the Alabama Initial Teacher Certification Testing Program. Plaintiff-intervenors are Eria P. Smith and the Board of Trustees for Alabama State University. In general, the plaintiffs and plaintiff-intervenors claim that: (1) the tests were implemented without adequate notice to the students and colleges; (2) the tests are unfair because they test students on matters which the students were not taught in the classroom; (3) the tests have an adverse and disproportionate impact on black colleges and black students; (4) the tests are culturally biased and bear no relationship to job performance as a teacher; (5) the tests’ disparate impact on blacks cannot be justified because said tests were not property validated; and (6) that the actions of the defendants violate statutory and constitutional rights of the plaintiffs and plaintiff-intervenors.
“The defendants deny these allegations and affirmatively assert that the process by which the tests were prepared did insure the tests to be valid. The defendants also assert that they are using the tests as a device to insure that teachers in the classrooms of Alabama will have the knowledge necessary to teach.
“The plaintiffs, plaintiff-intervenors and the defendants are desirous of implementing a solution to these issues without the further expense of litigation. The parties have waived further hearing and have agreed on the form of this decree. The defendants do not admit and expressly deny that they have acted in any manner to violate the Constitution, any statute of the, United States, or any other law or statute.”
Essentially, therefore, all that holds the consent decree together under the law is a promise by the State of Alabama to abide by
its terms, and an acceptance of, and reliance upon, that promise by plaintiffs.
This court, with the September 8 memorandum opinion, as it had been expressly mandated to do by the Eleventh Circuit, required that defendants live up to their promise and make good on their word. The lesson behind the memorandum opinion is a simple but fundamental one that every student, beginning in the earliest year, must learn: if you make a promise, you'must be prepared to live up to it.
Second, in the memorandum opinion, however, an important question still remained for the court: what exactly did the State defendants promise? Because there was no trial, the parties and the court did not have the benefit of court findings on the validity of plaintiffs’ claims, nor do they have an evidentiary record to give direction to, and a court determination of, what remedy 'would have been appropriate. All the court had were desires and expectations, that is, what “the parties desire[d] and expect[ed] [would] be reflected in, and be enforceable, as a judicial decree.”
Rufo v. Inmates of Suffolk County Jail,
502 U.S. 367, 378, 112 S.Ct. 748, 757, 116 L.Ed.2d 867 (1992). These desires and expectations, unfortunately for the court, were not for a specific test then on the market. Instead, what the parties agreed to, and what the court approved, was a
process
for the development of an acceptable test.
In addition, it must be emphasized that this agreed-upon process was, itself, untested; it was not the product of some empirical effort. In recognition of this fact, the parties built into the process certain measures to accommodate the real possibility that certain aspects of the decree might prove unworkable. For example, although the consent decree sets up restrictions on p-values— which are a basic measure of a question’s difficulty and simply reflect the proportion of test-takers that answer a question correctly — these restrictions are themselves flexible:
“[I]f 55% of the students answered an item correctly, then its p-value is .55. Under the decree, Type I items are ‘those for which the item difficulty (i.e. p-value) for blacks and whites differs by no more than 5%.’ Type II items are those ‘for which the item difficulty for blacks and whites differs by more than 5% points but not more than 10%.’ Type III items are ‘those for which the item difficulty (i.e. p-values) differ[s] by a percentage which is greater than 10% but not more than 15%.’ Under the consent decree, all test items would be field-tested before use on an actual test. On the actual test, Type I items would be used exclusively ‘so long as they are available in sufficient numbers to provide comprehensive coverage of the objectives sought be measured in the examination.’ Type II items could be used ‘only if Type I items are not available in sufficient numbers to provide comprehensive coverage of the objectives sought to be measured in the examination.’ The decree further provided that after the examination had been used a number of times, the test items would be reclassified to determine the actual difference in p-values among racial groups. If there were not a sufficient number of Type I and Type II items, then Type III items could be used on the test. Type III items could never comprise more than 10% of the total number of items on the test unless the monitoring panel agreed otherwise.”
Allen,
976 F.Supp. at 1410, 1416 (footnotes omitted): But most importantly, under the consent decree, even these flexible restrictions must ultimately give way to validity and psychometric soundness. “The plain language of [the decree] dictates that the test must be validated in accordance with the most recent standards in the field. [L]egal considerations dictate that validity trump the need to reduce disparate impact.”
Id.
at 1421.
At issue here, therefore, is not just a process but
a flexible process,
specifically geared to address the real possibility that certain provisions in the consent decree might not be workable or might not meet legal and test development standards.
In this sense, while the standards articulated by the United States Supreme Court in
Rufo,
502 U.S. at 383-86, 112 S.Ct. at 757, 760-61, governed this court’s September 8 memorandum opinion, the facts that gave rise to the
Rufo
standards differed dramatically in type from the facts presented to this
court. In
Rufo,
the court was confronted with a concrete, essentially unbendable court-ordered requirement: double-bunking of pretrial detainees in jail cells. And it was to this requirement that the Supreme Court sought to apply its newly articulated standards, the first of which is that there be “a significant change either in factual conditions or in law,”
id.
at 384, 112 S.Ct. at 760, such as, for example, “when a decree proves to be unworkable because of unforeseen obstacles,” or “when enforcement of the decree without modification would be detrimental to the public interest.”
Id.
Here, in contrast, this court was not presented with an unbendable court-ordered requirement but rather with a flexible process that specifically and most notably itself anticipated, and provided for, changed future circumstances. In other words, the consent decree itself takes into consideration “unforeseen obstacles” and the “public interest.” If, in implementing the consent decree, the parties confront obstacles whose elimination is in the public interest, then the decree must give way to that extent. As this court concluded in its September 8 memorandum opinion, “any test developed pursuant to the consent decree must be psychometrieally sound; it must be validated in accordance with the most recent standards in the testing field. Validity and soundness trump any need to redress discriminatory impact. The public deserves as much, and the court will approve nothing less.”
Allen,
976 F.Supp. at 1430-31.
As the parties recognized from the outset, the proof of the pudding (that is, the workability of the consent decree) is in the eating (that is, in the application of the decree). The answer to whether equitable relief under Rule 6(b)(5) is appropriate, is not a clear “yes” or “no,” but rather calls for a balancing, for an exercise of “sound judicial discretion” by this court as to whether to modify or vacate the decree.
Rufo,
502 U.S. at 380, 112 S.Ct. at 758 (quoting
System Federation No. 91, Railway Employees v. Wright,
364 U.S. 642, 647, 81 S.Ct. 368, 371-72, 5 L.Ed.2d 349 (1961));
see also id.
at 394, 112 S.Ct. at 765 (O’Connor, J., concurring);
Stovall v. City of Cocoa, Florida,
117 F.3d 1238, 1240 (1st Cir.1997). Because it could not be said what the outcome of the flexible process set forth in the consent decree would be, the court denied defendants’ motion to vacate or modify. The concerns raised by the defendants could very well be redressed through application of the consent decree iteelf, and thus until the consent decree was put into effect and tested, the court could not say whether the decree should be vacated or modified.
In other words, the parties could, within the boundaries of the decree conceivably, themselves reshape the decree to meet the concerns raised by defendants. Indeed, “[i]n this vein, the court suggested to the parties] that, as the experts proceed to develop a test, if a particular aspect of the decree is problematic, the parties should consider meeting with each other to see if they can agree on a modification to address the concern before returning to court. In addition, the defendants should also first explore the possibilities of reasonable modifications before concluding that the decree must be vacated.”
Allen,
976 F.Supp. at 1430-31 n. 115.
B.
These above reasons, which the court gave in rejecting defendants’ motion to modify or vacate the 1987 consent decree, lead the court to conclude as well that plaintiffs’ motions for further relief and for preliminary injunction should be rejected. First, it is worth restating that this court has not barred the State of Alabama from using teacher testing. To the contrary, the court has held that the State may pursue state
wide teacher testing. The court has simply-held that defendants must pursue this goal within the broad boundaries of the consent decree, unless it should appear that this is not achievable after a reasonable effort. With the September 8 memorandum opinion, the court faulted defendants for taking the position that it is impossible to meet the consent decree’s requirements and for seeking court relief, without having first, made this reasonable effort or otherwise presenting reliable and convincing evidence that to pursue compliance with the decree would be futile or unreasonable. Therefore, the command of the 1995 State legislation — that teacher candidates be tested — is entirely consistent with the 1987 consent.deeree. .
But second and more importantly, this court has not said that the State must adopt a test that has no disparate impact. Should it appear that'the consent decree or any parts of it are standing in the way of achieving a valid and sound test, the consent decree must give way to achieving this goal — á goal whose application is not new to this court. For example, in
United States v. City of Montgomery,
775 F.Supp. 1450 (M.D.Ala. 1991) (Thompson, J.), this court approved a test for selection of sergeants in the Montgomery City Police Department despite the fact that the test had a disparate impact on African-Americans. .The bottom line was that the test had to be, and was, validated. In addition and more to the point, in
Paradise v. Wells,
686 F.Supp. 1442 (M.D.Ala. 1988) (Thompson, J.), the court was confronted with a consent decree that required that any promotion procedure used by the Alabama Department of Public Safety meet the twin requirements of no adverse racial impact and validity under the Uniform Guidelines of Employee Selection Procedures; under the decree, therefore, a procedure could not be used if it had adverse impact. The court approved a modification that dramatically changed the decree to allow the department to implement a procedure with adverse racial impact, as long as the procedure has been validated and the adverse impact minimized.
Finally, this court disagrees with plaintiffs that it has barred the State from using a nationally normed test. The court cannot say at this time whether' the results of the process set up by the consent decree will prove acceptable under the law as well as under test-development standards; and, more specifically, the court cannot say what parts of the consent decree will ultimately survive this effort. Therefore, it is premature to conclude that a nationally normed test would fall outside the boundaries of the consent decree.
It is especially noteworthy that the 1995 legislation provides that the “State Board of Education shall prescribe the
manner
in which the examination is to be administered and shall further determine the
standard to be used for passing
the examination.” (Emphasis added). The State Board has yet to fashion these prescriptions, and when it does, the prescriptions may or may not meet the requirements of the consent decree. The State Board may administer the statute in a “manner” and may set up the “passing standard” so as to comply with the consent decree. For example, the legislation states that the nationally normed test is be used as a condition for “graduation” rather than for “certification.” Although the court cannot, without assistance, fully appreciate the significance of this distinction, the court is aware that defendants did, at one time, contemplate using the test for “remedial purposes” in teacher preparation programs. These remedial purposes — which may include improving the quality of teacher training programs and thereby teachers — may not be inconsistent with the spirit of 1987 consent decree (perhaps warranting some modification of the consent decree).
Admittedly, an expert for plaintiffs submitted an affidavit stating that a nationally normed test would “likely fail” to reflect the curriculum required in Alabama’s teacher training or preparation programs. Aside from the fact that the expert’s comments are overly conelusory and generalized, the expert overlooks two things. First, the statute provides that “the Legislature directs the State Board of Education to review the requirements of programs for teacher education and preparation and select a nationally normed teacher examination to be used.” Thus, because the Board’s selection of a nationally normed test is tied to the Board’s review of
the requirements for teacher training programs, it could be inferred from the statute that the test must mirror those requirements. But second and more importantly to the court, the solution to any lack of correlation between the test and any teacher-training requirements may be to change and improve the State’s teacher training programs (in particular, those programs in schools which the State finds to be producing teachers who cannot pass the nationally normed test) to better prepare teacher candidates to meet national standards.
The court must further admit, as plaintiffs argue, that the consent decree’s GPA provision (that, if a candidate does not achieve a passing score on the test, then a separate formula can be used under which the candidate’s college grade point average, commonly known as GPA, would count 50% and the test score would count 50%) appears at first blush to conflict with the 1995 statute’s requirement that a teacher candidate “pass” the nationally normed examination as a condition of graduation. However, as stated, the statute goes on to say that the State Board of Education shall “determine the standard to be used for passing the examination.” Whether this standard could incorporate a candidate’s GPA is a matter for the State Board.
But more importantly, even if the two provisions do conflict, the appropriate remedy is not apparent. The GPA provision in the consent decree obviously reflects a deep concern shared by plaintiffs and defendants that any state-wide teacher test adopted by the State might not correlate with, or mirror, what these candidates learned in the State’s teacher training institutions. This concern could be based on two possible scenarios: The first is that this lack of correlation manifests itself in most or all teacher training programs across the state; and the second is that it is more focused in a few, identifiable institutions. If the former scenario develops on a broad and racially indiscriminate scale, then the validity of the test may be drawn into question, and the State may be obligated to reconsider whether the test is testing what it should or whether the teacher training institutions are teaching what they should; and if on a small and racially indiscriminate scale, it may be seriously debatable whether the GPA requirement is still needed. On the other hand, if it should end up that this lack of correlation is substantial and is limited primarily to certain institutions, then the State may be called upon to explain why this is so. The GPA provision in the consent decree may simply be a way to bring this issue to the forefront, and place on the State the need to redress forthwith the conditions in those institutions which may be the cause of the scenario, with the goal of the eventual elimination of the GPA provision by the court. The adoption of a state-wide test without the GPA provision would not address these concerns. Because the GPA provision was carefully bargained for by the parties and because it may reflect an important dynamic between the parties not apparent from the face of the decree, the court should not: discard it too readily. In any event, and of course, if it should appear, despite these concerns, that the GPA provision is not in the public interest, then the court would be obligated to modify the consent decree to eliminate it.
The record before the court, though very limited, reflects that both the 1995 legislation and the 1987 consent decree may have good attributes. It is therefore premature to chose one and pitch the other in its entirety; rather, the parties should first make a reasonable and good faith effort to reconcile the two — if not in whole, then at least in substantial part. The 1995 statute and the 1987 consent decree are not all or nothing creatures. The consent decree is subject to modification to the extent it conflicts with the statute, and the statute is subject to sever-ability to the extent it conflicts with the consent decree. Arguably, therefore, each can be reshaped to form a logical and fitted piece of a unified whole.
II.
Finally, plaintiffs suggest that, even if the State defendants have not yet violated the consent decree and even if the 1995 legislation does not on its face violate the consent decree, the defendants may take action in the future that violates the decree. Plaintiffs ask that the court issue a broad injunction prohibiting defendants from taking any ac
tions that conflict with the consent decree. Plaintiffs acknowledge that the consent decree already prohibits defendants from taking any actions inconsistent with it, and that any additional prohibitory injunction would be redundant. The court, therefore, sees no need to issue an injunction on this basis.
III.
In conclusion, the court gives the same instruction to both plaintiffs and defendants: implement the consent decree, and let us see what happens. It may be that, after implementation, the parties will find that they are not as far apart as they think, or even that they have no real differences that they cannot resolve themselves to the benefit of all.
Accordingly, for the above reasons, it is ORDERED that plaintiffs’ renewed motion for preliminary injunction, filed on July 10, 1996, and plaintiffs’ motion for further relief, filed on August 9,1996, are denied.