This is not a response...

This is not a response to Aaron's response, but really the second part of the post I posted for Matt C yesterday.

I'm enjoying being an observer to the dialogue!


Matt Chudek said:

Here are a couple of extra thoughts about education, since I've already started.

One of the measures for improving the quality of education mentioned in the article, aside from a nationalised curriculum, is "improving teacher quality through merit pay".

This is an interesting idea. It seems to stem from the economic notion of letting the market optimise efficiency of service provision by competition for the best rewards. I'm sure we're all familiar with the economic version: there's money to be made selling bread, several bread-makers compete, the ones which can make the best bread for the least cost get the most customers because of their cheap price for a better product, the others go out of business and only the most efficient producers are left.

I could be getting it wrong, going on just the one brief quote above, but it sounds like they're trying to achieve something similar in education: post a big reward and let competition improve efficiency. I suspect though, and this is just my own humble opinion, that the two situations may be strongly disanalogous and this might not work out so well.

A key feature of free-markets that let them optimise production is selection pressure, poor producers are simply eliminated, good producers take their place. They expand their business to fill the void. This is not the case in education. There's a shortage of teachers, let alone good teachers. Even when you reward the best teachers, the worst are not eliminated from the market and replaced, they're still desperately needed and still teaching, but now having resources channeled away from them towards the best teachers.

Even the best teachers can only teach a few kids. In fact, trying to identify the best teachers and by rewarding them hoping that they'll somehow "replace" the worst is crazy, increase a teacher's class size over a certain threshold kills their ability to teach well, and in fact reducing the number of kids in their class increases the quality of education they can provide to each. Let this play out and you have the already good teachers with small classes getting extra money, and thus needing to teach fewer pupils to sustain their income and being able to reduce their class sizes further; while the worse teachers are left with ever increasing class sizes, and those students get ever worse education.

In short, rewards for good teachers necessarily entail a basic cost for other teachers, the opportunity to have invested that money in improving all teachers, but could also have greater costs for poor teachers: demotivation since rewards are seemingly unattainable and increased class sizes.

I doubt that's the motivation for "merit pay", I suspect it's more likely to be motivated by this sort of argument:

Every teacher wants more money. They can get more money by becoming a better teacher, so they'll work hard to become a better teacher.

Now if it were the case that every teacher could make a choice to magically work harder and become better, that could work. But teaching isn't like working on a production line, where all you have to do is sew faster and take shorter lunch breaks to increase your productivity. It's a complex art that deals with a very, very sensitive complex medium: human minds. Just like making a race out of building rocket ships tends to produce rocket ships that explode, good teaching is more likely to result from investment in on-going training and support than a scramble to win the prizes.

What's more, if the prizes are awarded the basis of student performance on national tests, teachers have a very strong incentive to teach students "for the test", rather than actually aiming to produce intelligent, well-rounded students. Anyone who's ever studied for an exam, passed it and forgotten everything a week later (but is still talking about the few arts lectures they went to without even being enrolled in), knows that teaching-for-assessment is the anathema of good education. Fostering an environment where students' test performance is a teachers' income, isn't likely to put students in a situation where the significant adult relationship in their daily lives is a very healthy one.

Still, there is a class of teachers for whom "merit pay" could produce a significant improvement, and they're worth mentioning. Imagine again a bell curve, this time describing quality of teachers: few awful teachers on the left, and few prodigious pedagogs on the right and most teachers somewhere in the middle. Those teachers just up from the middle, who could be excellent teachers but just lack the motivation, might get just kick they need from the promise of extra money and really fulfil their potential.

I wonder though if this would be more than compensated for by the already excellent teachers who really inspire their students and produce world-leaders and geniuses, but instead are forced to start teaching-for-tests in order to earn their money.


Rather than forcing the misplaced metaphor of market-competition on this rather disanalogous situation, I reckon we should try to figure out how best to improve a field where you can't "out-compete" the worst, but rather need to improve everyone's skills, given a whole spectrum of prior teaching-ability and potential. I reckon programs of constant and on-going teacher training and peer-support would be a much better idea.

I actually did a dip-ed at unimelb (withdrawn: term 3), and was astounded at how flimsy the training was. The uni-based lectures taught entirely detached theory (which I happen to love, but helped about butt-kiss when I actually got in a class room), and the practical placements involved watching one other teacher teach just a couple of classes and then jumping in the deep end and just making it up as I went (which I also enjoy, but is really a terrible way to induct new teachers, especially if you don't want every single teacher re-inventing the wheel). I made a point of going to see as many different teachers teach as I could during my placements. Most were happy to have me there, but mentioned that this was very unusual, and certainly not required. It was completely unheard of for established teachers to go to each other's classes and learn from each other, even though this would obviously be the cheapest, simplest and most effective way to drastically improve everyone's skills.

I've been reading a book recently on deep cultural differences in worldviews, especially between the "East" and "West", and their consequences for our everyday behaviour. It's called "The Geography of Thought" by Richard Nisbett. I recommend it. These two paragraphs on education caught my attention and I thought I'd quote them for you:

p.55 "Japanese schoolchildren are taught how to practice self-criticism both in order to improve their relations with others and to become more skilled in solving problems. This stance of perfectionism through self-criticism continues throughout life. Sushi chefs and math teachers are not regarded as coming into their own until they've been at their jobs for a decade. Throughout their careers, in fact, Japanese teachers are observed and helped by their peers to become better at their jobs. Contrast this with the American practice of putting teachers' college graduates into the classroom after a few months of training and then leaving them alone to succeed or not, to the good or ill fortune of a generation of students.

An experiment by Steven Heine and his colleagues captures the differences between the Western push to feel good about the self and the Asian drive for self-improvement. The experimenters asked Canadian and Japanese students to take a bogus "creativity" test and then gave the students "feedback" indicating that they had done very well or very badly. The experimenters then secretly observed how long the participants worked on a similar task. The Canadians worked longer on the task if they had succeeded; the Japanese worked longer if they failed. The Japanese weren't being masochistic. They simply saw an opportunity for self-improvement and took it. The study has intriguing implications for skill development in both East and West. Westerners are very likely to get good at a few things they start out dong well to begin with. Easterners seem more likely to become Jacks and Jills of all trades."

p.189 "Asian math education is better and Asian students work harder. Teacher training in the East continues throughout the teacher's career; teachers have to spend much less time teaching than their American counter-parts; and the techniques in common use are superior to those found in America...Both in America and in Asia, children of East Asian background work much harder on math and science than European Americans. The difference in how hard children work at math is likely due at least in part to the greater Western tendency to believe that behavior is the result of fixed traits. Americans are inclined to believe that skills are qualities you do or don't have, so there's not much point in trying to make a silk purse out of a sow's ear. Asians tend to believe that everyone, under the right circumstances and with enough hard work, can learn to do math."



Afterthought:

Aside from competition not being able to eliminate bad teachers (and possibly worsening them), there's a second disanalogy between “merit pay” for teachers and market competition. Students often have little choice in who their teacher is, many parents have little choice about which school they send their kids to. Consumer choice is very limited in the education market, but is essential if market competition is going to be able to optimise teacher quality. Without it you're just channeling more resources to those who need them least.