Forecasting enrollments during court-ordered desegregation

Distinctive issues arise when a demographer must forecast enrollments in a context of court-ordered desegregation. The key issue studied here is whether magnet schools have strengthened a district's overall attractiveness to enrollees from outside or merely siphoned students away from other nonmagnet schools within the district, without any real districtwide enrollment gain. To clarify this issue, I analyze patterns of change in grade progression rates over several years as magnet schools were phased in at a large urban school district. Generally, magnet schools induced little actual gain, merely slowing the overall weakening of districtwide retention. These findings furnished an important 'reality check' on the judgment for crafting appropriate forecasting assumptions and the resulting forecast proved reasonably accurate.