The effort to establish a voluntary interdistrict transfer plan as part of the remedy to desegregate the Kansas City, Missouri School District is before us again. The Jenkins Class filed a motion with the district court for approval of voluntary interdistrict transfer plans for the North Kansas City and Independence School Districts. The State, KCMSD, and the American Federation of Teachers filed their opposition to these plans. The district court recognized that numerous significant issues had been brought to the court's attention regarding possible problems and complications with the proposed plans. The court enumerated a number of such problems and stated that they had raised issues too important to ignore, despite the desirability of implementing transfer plans generally. The court denied approval of the plans, but encouraged the Jenkins Class and the two suburban school districts to submit revised proposed transfer plans addressing the opposing parties' objections. Order of May 24, 1991. The Jenkins Class renewed its motion, again proposing the same plans, and the district court once again denied the motion. Order of January 24, 1992. The Jenkins Class appeals. Because the two orders of the district court do not squarely decide the many objections raised by the State, KCMSD and AFT, we remand for further consideration of these plans, and for renewed exploration of the issue of voluntary interdistrict transfer plans.

The district court held that the State should pay for transportation and tuition costs associated with a voluntary interdistrict transfer plan in its first remedy order. Jenkins v. Missouri, 639 F. Supp. 19, 38-39 (W.D.Mo.1985), and ordered the State to "actively seek the cooperation of each school district in the Kansas City, Missouri metropolitan area" in such a program. Id. at 38. The district court pointed to the success of the voluntary interdistrict transfer plan which was a part of the settlement between the suburban school districts and the State in the St. Louis area. Id. at 38-39. In affirming the desegregation remedies in Jenkins v. Missouri, 807 F.2d 657, 683-84 (8th Cir.1986), cert. denied, 484 U.S. 816, 108 S. Ct. 70, 98 L. Ed. 2d 34 (1987) (Jenkins I ), this court en banc affirmed the district court's provisions for a voluntary interdistrict transfer plan, directing the court to make certain modifications and holding that the finding of intra-district constitutional violation was sufficient grounds under Liddell v. Missouri, 731 F.2d 1294, 1303 (8th Cir.) (en banc), cert. denied, 469 U.S. 816, 105 S. Ct. 82, 83 L. Ed. 2d 30 (1984), to make the State pay for the interdistrict program. We specifically commented that a voluntary interdistrict program had great potential for improving the racial balance of schools in the Kansas City area. Jenkins I, 807 F.2d at 683 n. 30. Judge Ross concurred specially to underscore the importance of this program. 807 F.2d at 687. Judge Ross said that "failure to organize and implement [such a] program would be a very significant factor in determining discriminatory intent in ... future litigation." Id. All of the judges participating in the en banc consideration concurred in the desirability of a voluntary interdistrict plan. However, actually designing and implementing such a plan has been a long, laborious process.

In 1986 the district court ordered the State to renew contacts with the Boards of Education of the suburban school districts in the metropolitan Kansas City area and to report to it. The district court announced that if the State did not demonstrate its commitment to obtaining the cooperation of the suburban districts, the court would seek other methods of accomplishing the task at the State's expense. 639 F. Supp. at 51.

Meanwhile, the Jenkins Class tried to enroll black KCMSD students as non-resident, tuition paid students in the North Kansas City, Independence, and Lee's Summit school districts. Those districts refused to admit the students, and the students sued them in the Naylor litigation, in which the district court granted summary judgment in favor of all defendants, Naylor v. Lee's Summit Reorganized School District R-VII, 703 F. Supp. 803 (W.D.Mo.1989). We reversed and remanded this case with orders that it be transferred to Judge Clark. Jenkins v. Missouri, 904 F.2d 415 (8th Cir.), cert. denied, --- U.S. ----, 111 S. Ct. 346, 112 L. Ed. 2d 311 (1990).

Thereafter, the Missouri City School District agreed to accept ten voluntary interdistrict transfer students from KCMSD. The second year funding of this plan was before this court in Jenkins v. Missouri, 965 F.2d 654 (8th Cir.1992).

Our remand in the Naylor case led to negotiations between the Jenkins Class and Independence and North Kansas City School Districts. Eventually, each of the two districts proposed a separate voluntary interdistrict transfer plan and the plaintiffs dismissed the Naylor suit with prejudice. The Independence School District plan would admit 420 KCMSD students, beginning with thirty-five seventh grade students in 1992 and an additional thirty-five seventh graders the next year. After the first two years the program would be centered in a collaborative school that would require construction of a new building at an estimated cost of $4,402,341 to ultimately house 480 students, half of whom would be minority students from KCMSD and the remainder students from the Independence District. The North Kansas City plan would accept sixty kindergartners and sixty first graders in the program's first year, and thereafter sixty kindergarten students a year would be admitted so that a maximum of 780 KCMSD students might ultimately attend school in North Kansas City. The KCMSD transfer students would be dispersed throughout the North Kansas City school system. The North Kansas City plan also envisioned construction of additional facilities at a cost of $4,386,600 plus additional amounts for architect's fees, land acquisition, furniture and equipment, and attorneys' fees. Both plans would admit minority KCMSD students with no history of serious discipline problems. A representative of the Jenkins Class was to select the students to participate in the transfer program. Both plans would require the expenditure of funds for tuition, transportation and incremental costs such as salaries for additional certified classroom teachers, counselors, school nurses, and administrators required by the program, as well as costs of staff development programs to prepare the staff, resident and transfer students and parents for successful operation of the plan. Both districts state that their school system is now operating at or near capacity.

The Jenkins Class moved in the district court for approval of these plans. The other parties to the suit, KCMSD, the State and the American Federation of Teachers, all articulated support for the general concept of voluntary interdistrict transfer plans, but objected to the particular plans before the district court. The State objected to the lack of uniformity in the plans, which would make the plans difficult to administer. Order of May 24, 1991, slip op. at 2. The district court catalogued a number of the State's other objections, including complaints that: the plans are proposed in addition to the existing long range magnet and capital improvements plan which should therefore be cut back as quid pro quo for the new programs; the tuition figures for the state to pay are excessive; the definition of minority students to include students other than blacks is over inclusive; the forty-five minute transportation time limits provided in the plans are unreasonable; and the Independence plan to create a predominantly minority school in a predominantly white school district is of questionable constitutionality. Order of Mar. 24, 1991, slip op. at 2-3.

The court also pointed to KCMSD's concerns about whether additional capacity is actually necessary to accommodate transfer students in Independence; whether a magnet school is necessary there to attract minority transfers from KCMSD; and whether it would be permissible for Independence to operate a predominantly minority school in an overwhelmingly non-minority school district. Slip op. at 3. KCMSD also expressed concern about whether there is any need for the requested capital improvements in North Kansas City. Id.

The court recognized the teachers' concerns that the plans do not adequately address insuring a racially and ethnically diverse work force or provide appropriate protection against job displacement for KCMSD employees. Slip op. at 4.

The court further acknowledged the truth of the Jenkins Class' response that because the suburban school districts are not parties to the case or adjudged constitutional violators, they can walk away from the plan if the parties do not agree to it exactly as proposed. Id. It concluded, however, that the issues raised were too important to ignore for the sake of quickly implementing the two plans. Id. at 4-5. It encouraged the Jenkins Class and the SSD's to submit revised proposed transfer plans and respond to the concerns set forth by the State, KCMSD, and the teachers. Id.

Thereafter, the Jenkins Class informs us that it held further discussions with Independence and North Kansas City, and the Class renewed its motion for approval of the voluntary interdistrict plans. Again, KCMSD, the State and the Teachers filed responses in opposition. The district court recognized that the plans were identical to those submitted earlier, and acknowledged the Jenkins Class' statement that the objections and reservations had been considered and could not be accommodated without losing the voluntary participation of the two suburban school districts. The court then denied the renewed motion for approval of the plans "for the reasons set forth in this court's order of May 24, 1991." Order of Jan. 24, 1992.

Although the parties suggested the need for a hearing the court did not conduct a hearing on either of the motions.I.

This case stands in a somewhat unique posture as the May 24, 1991 order catalogs issues too important to ignore, without deciding them. When the parties resubmitted the identical plans, the court denied them without articulating its reasons, other than by reference to its first order, or conducting a hearing. Order of January 24, 1992. Thus, while recognizing the issues, the district court did not finally decide them and no opportunity was given to explore modifications. There has been no clear articulation of the reasons for the decision.

We must first address two procedural questions before considering the merits of the Jenkins Class' motion.

First, the nub of the Jenkins Class' argument is that because the denial of approval by the district court was not based upon factual findings, we must conduct a de novo review. The Jenkins Class then argues that since it presented a settlement with two suburban districts, the court must assume that the plans as a matter of fact will accomplish the interdistrict transfer of KCMSD minority students consistent with the voluntary interdistrict transfer component of the remedy. They further argue that unless this court concludes that these plans call for practices which have been ruled in prior judicial decisions to be per se illegal or unconstitutional, the district court's orders must be reversed and the plans approved.

KCMSD attacks the Jenkins Class' argument that its "settlement must be presumed valid," since there has been no settlement between the parties, but only between one of the parties and two nonparties. KCMSD's response has a significant weakness: as a voluntary interdistrict transfer plan of necessity must involve suburban school districts that are not parties to this litigation, it would be essential that the settlement include non-parties in order for there to be any plan at all. KCMSD and the State as constitutional violators, are subject to an order to participate in such plans even though they may not have participated in the settlement discussions or agreed to the settlement. See Liddell, 731 F.2d at 1311. In view of the hesitancy or timidity, if not tacit resistance, of the three opposing parties to this settlement, it is not surprising that the Jenkins Class representative would directly approach the suburban school districts to try to achieve a settlement. While the settlement before the court in Little Rock School District v. Pulaski County Special School District No. 1, 921 F.2d 1371, 1384-85 (8th Cir.1990), was between all parties to the litigation, under the unique circumstances before us, we believe that a settlement between one of the parties and two of the only kind of entities with whom a settlement is possible, is at least deserving of careful scrutiny by the court and an effort to determine if the plan may be effectuated.

Second, the State argues that this case is moot as both Independence and North Kansas City have withdrawn their agreement to participate in the plans. The record indeed contains correspondence from the attorney for the two districts indicating that the districts' willingness to participate in the proposed plan would expire if Judge Clark ruled the plans unacceptable, and that neither district was willing to " 'negotiate' any sort of participation in a voluntary transfer plan." The State argues that the Jenkins Class dismissed its action against the suburban school districts in Naylor with prejudice in order to get the two suburban districts to participate in the plans presented to Judge Clark and containing the expiry provision. Therefore, once Judge Clark failed to order the plans, the Jenkins Class lost its bargaining power. This, says the State, was a gamble that the Jenkins Class took and lost.

We cannot accept the State's argument that the appeal should be dismissed as moot. Since 1985 the district court, with the approval and urging of this court, has been attempting to include as part of the desegregation remedy a voluntary interdistrict transfer plan. Thus far, this has resulted in the implementation of a program involving exactly ten students in Missouri City. In view of the earlier statements of the district court and this court that such plans should be implemented, we refuse to terminate further exploration of voluntary interdistrict plans on the basis of an argument of mootness.

Further, as the armamentarium of the lawyer must include intractability, we hope that on further examination the court may find that the school districts have some flexibility consistent with their commendable attitude in initially agreeing to the plans.

With these preliminary comments in mind, we turn to the specifics of the plans before the district court, and whether the plans contain any constitutional or legal infirmities.

With respect to Independence, three infirmities present themselves. First, all of the Jenkins Class students participating in the voluntary plan in Independence would be transferred to the newly constructed magnet or collaborative school as a single minority outpost in an overwhelmingly non-minority school district. KCMSD and the State argue that the plan is defective for this reason. We entertain no doubt that this plan is constitutionally suspect on its face.

A second deficiency, also of constitutional dimension, is apparent in the Independence plan. The plan requires construction, at a cost in excess of $4,000,000, of a magnet school in a predominantly non-minority area that would exist to serve a school population one-half of which would be non-minority students from that district. The State and KCMSD argue that this would create a windfall for Independence in building additional space to house Independence' own students. Such a remedy would not be tailored to remedy a constitutional violation, as required by Milliken v. Bradley, 433 U.S. 267, 282, 97 S. Ct. 2749, 2758, 53 L. Ed. 2d 745 (1977) (Milliken II ); Board of Education v. Dowell, 498 U.S. 237, ----, 111 S. Ct. 630, 637, 112 L. Ed. 2d 715 (1991). This is particularly true in light of the claims made by the State, which have not been subjected to any verification by factfinding, that there is sufficient classroom space available in the Independence district to house transfer students participating in an interdistrict plan.

Finally, KCMSD and the State make the same argument with respect to the provision in the Independence plan that there be one teacher for every 15 students in the new magnet school. This is a substantially lower teacher to student ratio than has been approved by the district court for the schools in KCMSD. The desegregation remedy provided for a ratio of 22:1 in grades K-3 and 27:1 in grades 4-5. It is difficult to see how a student-teacher ratio found adequate to remedy constitutional violations within KCMSD is inadequate, without some record support and factual findings, for the same students once they transfer to a suburban school district.

Thus, there are particular constitutional infirmities in the Independence plan that prevent approval of such a plan by the district court.

In the North Kansas City plan, the transferring students were to be dispersed throughout the district. However, the plan requires construction of additional classroom facilities, also at a cost of in excess of $4,000,000, without any determination by the district court as to the need for such facilities and with some dispute between the parties as to whether sufficient classroom facilities already exist. Without a demonstration of need for the facilities to house the transferring students, this plan also runs afoul of the principles of Milliken II, as reiterated in Dowell, 498 U.S. at ----, 111 S. Ct. at 637.

The State argues that the net effect of the capital improvement aspects and, to a degree, the program costs of both plans, is to require the State to pay for a desegregation remedy twice, first within KCMSD and then in a suburban district participating in a voluntary interdistrict transfer plan. We see this as producing no more than the possibility of such a duplication, as the KCMSD intra-district remedy is still in the implementation phase (even though it may be well along at this point). Should participation of KCMSD students in voluntary interdistrict transfer plans result in lessened need for facilities within KCMSD, adjustments can be made. However, the spectre of duplication does not foreclose construction of new facilities that the district court determines to be necessary to accommodate transfer students in a suburban district, as the court can offset such increases with reductions in later construction within the district, if required. Needless to say, there should be no reduction in KCMSD construction for this reason until it is evident that the voluntary interdistrict transfer plans are having a real impact on enrollment within KCMSD. Avoidance of duplication is for the district court to consider in ordering new expenditures.

Many of the more technical complaints, particularly the State's objections concerning transportation times, tuition rates, and attorneys' fees, are issues that require further factual development by the district court. The fact that the State has proposed a plan demonstrates that many of these issues are matters of degree that may be resolved by negotiation and further investigation by the district court.

Both the State and KCMSD urge that the indemnity agreements that are a part of the two plans present potential constitutional or legal deficiencies, but we think that this is a sufficiently fact-bound inquiry that it should be first developed by the district court.

The teachers argue that the plans do not provide appropriate protection against displacement of Kansas City School District teachers, as a reduction of students in the district would result in a reduction in teachers and support staff in the district, and the agreements as proposed do not insure an ethnically diverse work force in the suburban districts. We do not consider the plans legally deficient in these respects. Certainly, the hiring practices in all school districts must comply with laws prohibiting discrimination. On the other hand, to mandate further protections in the voluntary interdistrict transfer plans adds an additional obstacle to the adoption of such plans.

The numerous objections to the two plans demonstrate the difficulties of achieving an interdistrict plan. Years have passed, however, and such plans have not come into existence, with the exception of the one noteworthy plan, which insofar as we are able to ascertain is working successfully. It is evident that the efforts to this point have not been successful in expanding the voluntary transfer remedy beyond this small scope. The desegregation monitoring committee, while serving an admirable function in other respects, has not succeeded in developing voluntary interdistrict transfer plans.

It is important for the district court to give full attention to development of an interdistrict plan. An evidentiary hearing is necessary for this purpose and that the court should give detailed consideration to the plan prepared by the State, as a starting point. The suburban school districts should be invited to participate in this evidentiary hearing, and the district court should make clear to the suburban schools that if they participate, such participation will not be used in any way to affect whatever legal rights they have. Any plan developed should have sufficient flexibility so that it may be tailored to each district. The beginning of this entire effort, however, should be a hearing by the district court to develop an overall master plan as a basis for discussions with the suburban districts. The time has come for affirmative efforts to push such plans to completion. Circulation of plans with a cover letter and the efforts of some committees have simply proved to be inadequate. Positive action is required.

The development of the plan should have as its cornerstone a truly voluntary selection of students, in which the students make application and indicate their choice of schools and willingness to attend. Once accepted into those schools, the students should be subject to the administrative and disciplinary procedures of the host district.

The funding provisions imposed upon the State must be sufficient to cover the reasonable costs for the plans so that the participating suburban districts are fully reimbursed for their expenditures. While the State makes much of avoiding windfalls to participating districts, it is evident that suburban districts participating in such plans must have the costs of such plans paid by the State.

As pointed out by KCMSD, the State, through its legislative and administrative bodies, has authority to effectively establish an interdistrict plan if it chooses, and it may require suburban districts to accept minority transfer students from KCMSD. The State could establish such a plan and implement it without the constraints that limit the courts. As the Supreme Court made clear in Freeman v. Pitts, --- U.S. ----, ----, 112 S. Ct. 1430, 1445, 118 L. Ed. 2d 108 (1992):

[I]t must be acknowledged that the potential for discrimination and racial hostility is still present in our country, and its manifestations may emerge in new and subtle forms after the effects of de jure segregation have been eliminated. It is the duty of the State and its subdivisions to ensure that such forces do not shape or control the policies of its school systems. Where control lies, so too does responsibility.

Indeed, meaningful and effective action of the State in establishing such plans would demonstrate a commencement of responsible assumption of local control of educational policies for desegregation purposes as referred to in Dowell and Freeman.

KCMSD proposes that the district court consider the possibility of enlisting the assistance of an independent, well-respected individual to attempt to bring the parties and suburban school districts together in achieving a plan. We consider this a sensible suggestion. There are a number of individuals in the Kansas City area with not only local, but also state-wide and national standing, reputation and abilities to make them valuable in this respect. Possibly someone from outside the area would give fresh insight to this effort. We urge that the district court consider the desirability of such a step. What is of utmost importance is that a major effort be made to achieve what the district court and this court have held to be a desirable program in orders going back more than seven years. The time has come for effective action to develop a voluntary interdistrict plan that "promises realistically to work now." Green v. New Kent County Sch. Bd., 391 U.S. 430, 439, 88 S. Ct. 1689, 1694, 20 L. Ed. 2d 716 (1968), quoted in Freeman v. Pitts, --- U.S. at ----, 112 S. Ct. at 1436.

We remand for further consideration the two plans now before us and for renewed efforts to effectuate a voluntary interdistrict plan in the Kansas City area.