Lots of strongly held views on this thread but also some very interesting hard data. One aspect not developed has been that around persistent regional differences.
As previously highlighted, the NE of England often underperforms in measures around secondary progress, HE participation etc. There will be other parts of the country with similar issues, though often disguised because they sit within a wider, generally more affluent region.
Although teacher salaries go further in such places, recruitment and retention can be difficult, local graduate level employment opportunities are not abundant and easily accessible in the same way as in the big centres like London, Birmingham and Manchester. These factors may well combine to limit average attainment and aspirations.
The fact that more students from educationally disadvantaged backgrounds are getting into elite universities and subsequently entering teaching, perhaps in their home region, may be an important accelerant in encouraging and assisting future cohorts of students to follow the same path and become teachers (and so repeat) or provide a pool of high calibre graduate labour in areas outside the 'defaults' of a small number of major conurbations and so lessen the current in-built geographic advantage experienced by those in the SE of England in particular.
This might also lead to a reappraisal of which universities are 'good' and which less so. Average salaries for Exeter graduates are typically ahead of those of Lancaster graduates but, accounting for the high proportion of SE of England students at the former, the Exeter economic 'valued added' might be marginal in absolute terms and poor relative to Lancaster. Equally Durham is a very strong, research-led university but it is really doing enough to help address the lower levels of secondary attainment and HE participation on its very doorstep?