I first encountered the problem as a graduate student and assistant. The professor assigned essay-type reports and tests to his undergraduates, first and second year students, but gave them to the assistants to read; he said the poor writing ability of the students made him ill.
Over the three years I read these reports, other assistants and I noticed that student writing fell into three groups: well-written, few spelling or grammatical errors, thoughts clearly expressed; thoughts and conclusions understandable, but many 'little' errors; thoughts and conclusions muddy or confused, bad spelling and grammar.
The first group was small and the others were not clearly separable.
We (the assistants and some professors with whom we discussed the writing) became interested and looked into the student records. The good papers were by graduates of (only) four high schools in the state. As some might expect, the worst mostly were from small schools in poorer areas of the state (Texas). The large group in the middle showed no discernable pattern. Our approach was superficial and composed of personal opinions, so no significance should be attached, with the exception perhaps of our agreement on the four schools whose students were at the top on writing ability. Two of these were Sunset (I am unclear on details, too long ago), I believe in Dallas, and Masonic Home(?) of, I think, Fort Worth. I have forgotten the other two.
It was obvious that language ability received much attention at the four schools, and, as Bruce suggested, all of the instructors cooperated in teaching it.

SRS- any comments? It was over 50 years ago and the situation has undoubtedly changed.