That’s just how the educational data is reported in the U.S., because 55% U.S. K-12 students are part of subpopulations that are differently situated sociologically (e.g. recency of immigration, immigration filtering effects, etc). It usually misleading to look at the averages without accounting for that.