If the market is heartless and government is incompetent, then the government should decide on goals and pay actors in the private sector for implementing them.
Want to figure out which teachers are the best ones? Pay for student performance. The market will sort everything else out. You don't have to publish standards of what makes a good teacher, heck you don't even have to make pursuit of good teachers a defined goal, any more than you need to pass a law saying "Make sure you hire those programmers who are ten times as effective as the other ones."
(I know there is one glaring problem here, namely that a politically powerful lobby considers that solution anathema, but it would work.)
More than one, really. State schools in the UK are incentivised in a way similar to this.
What it results in is gaming of the performance measures.
For example, pupils are steered into softer subjects where results are easier to attain. Pupils below-but-close-to a grade increment are given more intensive teaching at the expense of those who will attain the relevant grade left to their own devices.
Push that down to the per-teacher level and good teachers would not want to work with difficult pupils, since they will get poor results and less pay. "So make the pay conditional on improvement rather than absolute attainment". OK, in which case you've just made an incentive for children to be labeled as behaviourally difficult and academically poor so that they can be improved more.
Want to get better teachers? Pump more damn money in. Get good managers and get good quality people teaching. Make it a profession people aspire to work in once more.
The thing is, you'll have to keep it up for about a generation (with not much to show for it), to flush the system through. Good luck with that politically.
Markets are not a panacea. Particularly so with subjective measures.
Education (in the broader sense of what we want to achieve) is subjective - even academic results are more subjective than one might think - how many mathematics qualifications is a history qualification of the same level 'worth'?
I'm not really sure what I believe here, but I'm going to make the argument of true believer in the powers of market forces. The purist objection here would be that you need to follow through; get out of the market's way further down the chain.
If you choose to reward achievement, you have to measure achievement. If achievement is measured by standardized testing, then people will game the metrics. However, if the market chooses the testing procedures, it will reward metrics which are harder to game. This would take several iterations (read: many years) to stabilize, but it should converge on some sort of solution.
I think the fundamental problem here is forcing the market to not become complacent. This sort of thing needs to be done like a genetic algorithm. Every state tries a different testing procedure and then tests cross-pollinate and evolve under the pressure of market forces. There would need to be some way to reward experimentation as well, so that we don't wind up with one homogenized solution that stops progressing; there would need to be "random mutations".
Market forces aren't infallible. They're compromised by a profit motive, and that's not always a good thing.
If you want to see a market in action, look at TV networks. The big networks move farther and farther away from actual reporting. They go for lowest-common-denominator. The channels that do produce great pieces of TV either do it only sporadically, or they're pay-for channels that attract a more sophisticated audience. Even then, the show that critics call the greatest show in history - The Wire - had to fight to get its last season aired. It had an extremely small viewing audience.
However, if the market chooses the testing procedures, it will reward metrics which are harder to game.
Which metrics possibly exist that can't be gamed? The minute you have any metric, you're creating a "good" and a "bad" scale, and that means that people can cram towards the good as much as possible.
> However, if the market chooses the testing procedures, it will reward metrics which are harder to game.
How would you achieve this?
> This would take several iterations (read: many years) to stabilize, but it should converge on some sort of solution.
What makes you think this? Markets aren't immune to fashion and fad. We don't see stable financial markets, why should we see stable markets in education?
"Quality", "design", and "beauty" are all very subjective things and the market regularly gives options incorporating all of them. I see no reason why "educational attainment" would be any harder. (Imagine a future in which two schools got into a Mac vs. PC war: "Sure, you could go to Boring Prep. And you'll learn to read and do calculus. Yeah yeah, boring. Come to Trendy Fashionable High, where we nurture your mind holistically... and also have mandatory gym class so you look like this svelte model." Seeing that ad would give me about as much joy as a staunch Boring Prep grad is capable of feeling in his shrunken, blackened heart.)
>>
Pupils below-but-close-to a grade increment are given more intensive teaching at the expense of those who will attain the relevant grade left to their own devices.
>>
I could think of worse outcomes than public schools discovering triage. Academic triage, like medical triage, would probably result in more survivors. I mean, we're certainly not effectively educating the highest risk students now: if making that explicit saves education for the average folks who share the same classrooms, well, that is a price I'm willing to pay.
Here's the problem: how do you define student performance? Are we going based on grades? Because that leads to a system that's easily gamed without significant results. I was a B-and-C student in high school because I thought a lot of the stuff I was learning was crap. In the classes I did like, I'd deliberately take risks with what I did and that kept my grade from staying an A+. Is that a failure? The best teachers I had very often didn't give me As, because they were the teachers I'd really strain myself for - and for me, straining was trying to solve answers in a unique way rather than churning out generic A crap.
The other option is that you monitor students for achievement. Again, it's hard. Is that based on money? Personal student happiness? Again, depending on what the metric is, teachers will game it. Go for money and English teachers will start encouraging students to do something more worthwhile than write and study English for a living. If you make it "how many students get a college degree" then you're diluting the college pool further and you're neglecting students who're interested in vocational school and in working on their own rather than continuing education.
The best solution is to remove metric and go by the opinions of experienced people. Good administrators know good teachers when they see them. Good teachers always go by their own formula. If the system effectively removes restriction amongst teachers, then you'll see an increase in student performance, and suddenly paying teachers more is worth it. The problem is that teaching is an act of creativity; it's an art as much as anything is. Formula kills teaching. The minute you make something mandatory it's dead to kids. The problem is, formula feels safer, but it never works.
Want to figure out which teachers are the best ones? Pay for student performance. The market will sort everything else out. You don't have to publish standards of what makes a good teacher, heck you don't even have to make pursuit of good teachers a defined goal, any more than you need to pass a law saying "Make sure you hire those programmers who are ten times as effective as the other ones."
(I know there is one glaring problem here, namely that a politically powerful lobby considers that solution anathema, but it would work.)