Tuesday, 13 January 2009

Golden Handcuffs: Teacher Recruitment and Retention

One of the key problems for poor neighbourhoods is the bad educational outcomes for children that live there. As I have argued before, the evidence suggests that traditional regeneration programmes (with a strong focus on the built environment) don't do much to address this problem. So will the government's proposal to pay teachers more for working in the most disadvantaged schools (themselves generally in the most disadvantaged areas) fair much better?

The answer is that we don't really know. There isn't much strong evidence of a link between teacher pay and student outcomes. But these "golden handcuffs" are targetting recruitment and retention rather than pay per se. Unfortunately, we know even less about the extent to which this might affect educational outcomes. Intutively, it feels like it should matter, but there is little, if any, evidence to back up this intuition.

Even if the policy does work, it will not do much to affect the spatial concentration of poor educational outcomes. This is because about 60-70% of the variation in educational outcomes is down to pupil background, with only about 10% attributable to the school and even less, if any, to the neighbourhood. The fact that we have evidence that school matters more than neighbourhood at least argues that this is a move in the right direction. But the overwhelming importance of family characteristics and the spatial concentration of poor families, means that the targetting of disadvantaged schools can only ever do so much to improve poor education outcomes in disadvantaged neighbourhoods.

1 comment:

Tim Leunig said...

This is an unnecessarily pessimistic post. We know that parental income is a fairly poor predictor of kid's ability at tests aged 2-3, but a very strong predictor beyond that. Well-educated parents are more likely to read to their kids, etc etc.

There is no reason to think that this cannot be corrected in the school system. In extremis, imagine that kids from backgrounds that are well-correlated with poor performance were taught in groups of 4 rather than 30, so that they got huge amounts of teacher time, and so that lessons were always tailored to their educational needs, rather than to those of the class as a whole, and so that the teacher was always teaching rather than maintaining order. Do we really think that such a system would lead to the current unequal educational outcomes?

Of course, it would be expensive. Very expensive. I might be an order of magnitude dearer. But spending £30k a year per person for a decade may well be a wise investment (in simple monetary terms) for currently middle class tax payers, if it means that these kids end up in decent jobs, and not on benefits.

And the moral case, if it works, is surely unanswerable.

Of course, none of this will regenerate geographical areas, since we know that people with good exam results are most likely to migrate to areas of higher economic opportunities. But it would transform the lives of many people who currently receive an ineffective education, and live in places with few opportunities.