Rebecca Allen and Simon Burgess report that it is very hard to predict who will be a good teacher. I wonder: is this because researchers are looking in the wrong place? They are looking at individual attributes of teachers, rather than matches between teachers and students or schools.
Three things make me ask this.
1. We know from other occupations, from doctors to TV presenters, that what looks like individual skill is in truth a mix of individual and organizational factors. Why should teaching be different?
2. Most readers of this blog will have had one or two great teachers. But were they great for all pupils, or just for a few of you? Conversely, did your contemporaries have what they thought were great teachers, but whom you could never understand?
3. This new paper finds that:
A student's achievement in a subject in which the teacher shares the same gender, caste and religion as the child is, on average, nearly a quarter of a standard deviation higher than the same child's achievement in a subject taught by a teacher who does not share the child's gender, caste or religion.
Now, that comes from India. But there’s evidence from the UK of systematic ethnic differences in teachers’ perceptions of pupils’ ability, and the stereotype threat suggests that such perceptions have a nasty habit of becoming self-fulfilling.
Could it be, then, that this is another area in which we should pay less attention to individual skill, and more to organizational factors or just dumb luck?