Justice Souter, concurring....
We recognize that although demographic changes influencing the composition of a school’s student population may well have no causal link to prior de jure segregation, judicial control of student assignments may still be necessary to remedy persisting vestiges of the unconstitutional dual system .... This is, however, only one of several possible causal relationships between or among unconstitutional acts of school segregation and various Green-type factors. I think it is worth mentioning at least two others: the dual school system itself as a cause of the demographic shifts with which the district court is faced when considering a partial relinquishment of supervision, and a Green-type factor other than student assignments as a possible cause of imbalanced student-assignment patterns in the future.
The first would occur when demographic change toward segregated residential patterns is itself caused by past school segregation and the patterns of thinking that segregation creates. Such demographic change is not an independent, supervening cause of racial imbalance in the student body, and we have said before that when demographic change is not independent of efforts to segregate, the causal relationship may be considered in fashioning a school-desegregation remedy. Racial imbalance in student assignments caused by demographic change is not insulated from federal judicial oversight where the demographic change is itself caused in this way, and before deciding to relinquish supervision and control over student assignments, a district court should make findings on the presence or absence of this relationship.
The second and related causal relationship would occur after the district court has relinquished supervision over a remedied aspect of the school system, when future imbalance in that remedied Green-type factor (here, student assignments) would be caused by remaining vestiges of the dual system. Even after attaining compliance as to student composition, other factors such as racial composition of the faculty, quality of the physical plant, or per-pupil expenditures may leave schools racially identifiable. (In this very case, for example, there is a correlation in particular schools of overrepresentation of black principals and administrators, lower per-pupil expenditures, and high percentages of black students. Moreover, the schools in the predominantly black southern section of the school district are the only ones that use “portable classrooms,’' i.e., trailers.) If such other factors leave a school identifiable as “black,’' as soon as the district court stops supervising student assignments, nearby white parents may move in the direction of racially identifiable “white’’ schools, or may simply move their children into these schools. In such a case, the vestige of discrimination in one factor will act as an incubator for resegregation in others. Before a district court ends its supervision of student assignments, then, it should make a finding that there is no immediate threat of unremedied Green-type factors causing population or student-enrollment changes that in turn may imbalance student composition in this way. And, because the district court retains jurisdiction over the case, it should of course reassert control over student assignments if it finds that this does happen.
