We like TES editor Gerard Kelly's weekly provocations: they often hit nails in the region of the heads. Take last week ("Rigid positions see the system stall at foreplay"): of course schools do better if they are autonomous. Up to a point.
We all remember the fiasco when the Government trumpeted that it was going to give good schools the power to innovate. It turned out that virtually all the innovations proposed were already permitted. And the headteacher bonus scandal shows quite clearly that autonomy has to be tempered with proper accountability.
The trick is getting the balance, and the areas of autonomy, right. The Association of Teachers and Lecturers believes in professional autonomy in pedagogy.
However, our main issue with last week's editorial is its errors of interpretation of research. In simple terms, yes, academies do tend to outperform the schools they replace. But the performance of secondary schools as a whole tends to improve, and, crucially, the question of possible changes in intake after the makeover is ignored.
Which brings us to the second error. A huge range of research - including, interestingly, the school effectiveness tradition - confirms that about 85 per cent of the variation in achievement is due to factors external to the school. This is why intake is the vital context to schools' results.
Mr Kelly is right if he believes that the teacher effect is responsible for virtually all the remaining 15 per cent, and ATL supports his conclusions about valuing teachers. Well, we would, wouldn't we? But it is not as much as the "40-50 per cent" he cites.
Child poverty policy is the major route to better educational performance in the long run.
Martin Johnson, Deputy general secretary, Association of Teachers and Lecturers.