“…Black enrollment share did not impede "token" desegregation, but was an important predictor of both resistance to more intensive desegregation and being supervised by a court in later 4 Giles (1975aGiles ( , 1975b examines heterogeneity in desegregation patterns, but limits his analysis to 1968 and 1970 and uses fewer explanatory variables than considered here. Several other papers attempt do the same for earlier years, but these are either limited to districts (or counties) in only a few states (Pettigrew, 1957;Pettigrew and Cramer, 1959) or use state-level aggregate data (Harris, 1968;Vanfossen, 1968). Each of these papers covers fewer years than the present study, and none measure "explanatory" district characteristics before they may have been affected by the desegregation process.…”