Another thing the Swedes do better

"Sweden does not allow schools to disaggregate data for ethnicity or race. Instead, Swedish schools report achievement based on whether the student is a first- or second-generation immigrant and by the education level of his or her parents." (Quote from "Swedish For-Profit Chain to Run Charter School in N.Y.C." by Sarah D. Sparks in Education Week, February 9, 2011)

In addition to race, the US uses eligibility for Free or Reduced Lunch, an imperfect proxy for "socio-economic status" in part because of how variable reporting is. But we've always known that education level of the parents is the better predictor and therefore the better "denominator" of student learning to use in tracking the effectiveness of schooling. I don't know how the Swedes gather the data, but if done well it makes much better sense.