Yes, exactly! And yet, the schools are still really, really good!

Partially it's because the analysis is at the county level, and not at the district level. But I'd argue that using "averages" in the first place is extremely lossy and a classic Stats 101 mistake.