To be less fair to the economist, "adjusted by poverty level" is a heck of a spin, we've had many generations as a developed nation now, your state poverty level is caused by your state education outcomes. And that's without even speculating about what "demographic factors" means or implies.

How should you measure an education system? Should you measure purely based on the student's performance? What if the students are just better at reading, independent of the school? It's not hard to imagine that even with identical teachers, that inner cities schools would have worse test scores than wealthy suburban schools, especially if the latter are rich enough to afford tutors, the family environment is more conductive to learning, etc. Recognizing this fact, it's fairly obvious that "you're either at a third grade level or not" is a terrible way of assessing how good of a job an educational system is doing.