Getting back into writing regularly is sort of like getting back into regular exercise. The first few attempts are usually painful and half-hearted! But if you keep putting off getting started, then you don't ever get back into shape.
So here is a brief thought and some links to other stuff going on around the webs.
Andrew Gelman asks for the definition of a professional career. The main target appears to be teachers -- especially public, primary and secondary school teachers. Gelman is arguing with
Jonathan Chait about whether "accountability" is constitutive of professional status. Chait thinks that to have professional status, involves the possibility of quick upward advancement for good performance and also quick discharge for poor performance. Among other things, he thinks that professionals have performance-based income inequality.
Figuring out the nature of professional work is an interesting problem. It looks a lot like traditional philosophical problems (like, what is the nature of courage)! But I was somewhat surprised that Gelman didn't take on the underlying causal question in his post. Instead, he ends his post with a sort of punt: "If you think that it would be better for teacher salaries to be more unequal, that’s a position to take–but I don’t think that either salary inequality or lack of protection from firing are at all essential to the idea of a professional career."
Okay, it's a position one could take. But is it a reasonable position to take? Is it supported by the data? Do we know enough about the causal structure here to say whether performance-differentiated pay would make education better?
Most researchers seem to think that performance-based pay for teachers is a bad idea. For example, in this fascinating RSAnimate project adapted from a talk by Dan Pink, we learn that performance pay is great for algorithmic work (like making widgets) but not very good for creative work (like ... I don't know ... maybe teaching). And Steven Levitt of Freakonomics fame says something very similar.
For scholarly work directly on point, see this excellent review article by James Perry, et al. about merit pay in education, and this paper by David Marsden, on merit pay more generally. However, there is some dissent. For example, this study of the impact of performance-related pay for teachers in England by Adele Atkinson, et al. and this observational report using international data by Ludger Woessmann.
Anyway, I would have appreciated Andrew weighing in on the causal question: would merit pay for teachers improve educational outcomes?
My guess is that merit pay for teachers will have only a very small impact. We would, I think, be better served by improving teacher autonomy and helping teachers to master their disciplines.
On a related note, there is this interesting post about making mathematics education better from the guys at The Eternal Universe.