The blog over at the LSE, British Politics and Policy at LSE, has an interesting post on the relationship between teacher pay and educational attainment outcomes. A fuller summary of the academic article is here, the article itself here. Both summaries are rather superficial; the meat of the methodology and econometric models can be found in the full paper.
It turns out that, holy crap!, pay teachers more, get better outcomes. The theoretical reasonings behind this simple hypothesis are compelling:
There are two potential explanations as to why teachers’ pay may be causally linked to pupil outcomes. The first is that higher pay will attract more able graduates into the profession. As the potential supply of teachers rises because of the higher pay on offer, entry into teaching as a profession will become more competitive. This in turn will mean that the average ability of those entering the job will rise. Once recruited, higher relative pay and/or more performance-related pay may provide teachers with stronger incentives to improve their pupils’ educational outcomes.
The second mechanism is more subtle – namely that improving teachers’ pay improves their standing in a country’s income distribution and hence the national status of teaching as a profession. As a result of this higher status, more young people will want to become teachers. This in turn makes teaching a more selective profession and hence facilitates the recruitment of more able individuals.
The models are a bit parsimonious for my liking given the topic (and I tend to favor succinct models in my own work where possible), which is possibly reflected in an R2 between .2 and .5 (which for aggregate data is somewhat middling) for the models which estimate the determinants of achievement. This level of model fit is — to my eye — consistent with the bivariate relationship illustrated by Figure 1 in the summary (to which no R2 data are available in any of the versions available that I could find).
The policy prescriptions are sound for the recruitment of new cohorts of teachers, but these are likewise obvious: pay teachers more. However, they flirt with controversy when entering the tricky realm of how to apply these findings as policy for the extant stock of teachers. Again, here, most are sound, but this is likely to cause concern: “One solution is to provide an incentive mechanism for existing teachers to improve quality by paying them according to the percentile performance (in value added terms) of their pupils.” Incentives for teachers to ‘teach to the exams’ already exist; this would simply compound that with a blatant self-interest motivation.
I read this thinking “this is so stunningly obvious, I should have thought of it”.
I’d love to see how Greater Wingnutia would spin this, but that would require the denizens of Greater Wingnutia to . . . well . . . gather information beyond their one source, and that would require . . . well . . . effort and curiosity.