Bill Gates & Obama favor teacher pay based on student performance - How in the world can this work?

I’m sorry to depress…

You are a teacher. You are in the front-line trenches of academia. You are responsible for taking the youth of this nation and giving them skills and (at least some) academic rigor…making it so the youth have as good of minds as possible instead of mush.

It is a tough, stressful job (no cite but it is usually in the top 5 of all stressful job lists). It requires many different skills to do well. A good teacher, IMO, is really something to behold…the equivalent of a good lawyer.

God-damn-it…you SHOULD have respect. Respect includes pay. Respect means that young, smart, energetic college students aspire for it as a career. Respect certainly means that nearly all of the population SHOULDN’T think they should make more than you. Respect means that you DO NOT overhear a conversation between 2 waitresses where one is lamenting the fact that her daughter is dating a teacher and she (the mother) disapproves because she wants her daughter to marry and have kids with a successful man and not someone with whom she would struggle financially and would prefer her to marry someone who actually works*.

No disrespect to you - bt you are part of the problem. With respect comes all the things you long for. In addition, YOU may be fine with your position but, believe me, the best and brightest are not…and these people are not ‘knobheads’.

*True story. I overheard this conversation when I was a teacher.

Umm… no. With MONEY comes the things I long for. Huge, sustained investment in education. And yes, anyone who needs the ‘respect’ of people who judge others solely or even largely by their profession certainly matches my definition of a knobhead.

I don’t know - it’s such a dumb, reductive view of teaching and learning and development. If we up the wages a bit, or make it merit-based, suddenly a slew of new, better-qualified, exciting, creative graduates will enter the profession and everything will get better. All those people who made the banking world the model of probity and functionality that it is, I guess. What will actually happen is what ALWAYS happens when people try to apply deterministic, linear, business-based models to education; lots of fuss, lots of bluster, and no actual change. But, hell, we’ve tried everything else apart from massive, sustained investment. Might as well try this as well I suppose.

But look, merit pay isn’t just about pay. It’s also about measuring success. It’s a system that will bring forth a lot of data. Data that will prove you right. Inarguable data that will demonstrate the merit of smaller class sizes, will find most effective strategies for dealing with problem children, will prove whether computers are useful or not, etc. And all the things will be implemented. They’re not implemented now not because we all know they’re true but don’t care. They’re not implemented now because we don’t really know they’re true, and it’s easy for penny-pinching bureaucrats believe they aren’t. The data itself will do amazing things to illuminate education.

Well, that’s because the majority of teachers are bad at their jobs. People will have immense respect for the ones that are verifiably good.

I don’t know how familiar you are with educational research or how revealing such data is in education, but I have to disagree. Ed data always shows that some things work for some teachers with some kids at some times. Depending on the prevailing attitudes of the people in charge at the time, this might be a good thing or a bad thing, but the definitions of ‘good’ and ‘bad’ will always change enormously with a change in government. And teachers never have any say whatsoever. I worked for many years in the UK, a system which is drowning in educational data and has been for many years. And the quality of education there is generally regarded as being pretty dire.

The things we ‘all know’ to be true actually AREN’T true in a lot of cases; that’s part of the problem. And they aren’t implemented now because they cost a lot of money. After their value has been ‘proved’ and lots of teachers have been judged to be failing and replaced by these mythical creative, results-orientated new graduates, these things will STILL cost a lot of money and they STILL won’t get done. Then, we’ll all be on this thread in 25 years bitching about the short-sightedness of people who, instead of looking at complete systemic overhaul, decided to take the dumb, easy route by saying idiotic, pretend-radical things like ‘the majority of teachers are bad at their jobs’ and blaming teachers for the failings of an entire system and society.

I don’t have much faith in traditional “research.” It is too much money spent gathering too little data that is interpreted poorly, often to push some PC worldview (social sciences are plagued by this).

The difference between the data that will come from the merit pay system and the data from formal research will be night and day.

Merit pay is the “dumb, easy” route? It is short-sighted?

It is a dynamical process that will set off a whole set of other “systemic” changes. (As long as we succeed in solving the core problem of effectively, non-trivially, holistically evaluating a teacher’s performance.*)

*Phrasing it like that almost seems to take us to the start. And maybe it does. Maybe the point argued is that in trying to solve the concrete, real-world problem of merit pay will focus us to produce a truly effective algorithm of evaluation. And it is that which will propel us greatly, and discover more than all the formal, disorganized “research” ever did.

Could you give me an example of the type of thing you might expect to find out? I’ve taught in private schools around the world where merit pay is very much the order of the day - I’m in one at the moment, in fact - and I’d be interested in what ‘knowledge’ might accrue as a result of implementing such a system.

Oh, I meant to say - evaluating teacher performance is an enormous problem. It’s generally reduced to exam results (fine, but prepare yourself for some HORRIBLE classroom practice, UK teaching) and / or headteacher / administrator recommendation (cue all teachers staying in post for years because headteachers know that swapping teachers around is pretty much a superficial nonsense.) Again, it’s a problem that people have been discussing for decades. I could, I suppose, post the enormous matrix my organisation uses to assess performance and thus pay. It’s incredibly complex and extremely stupid. I assume that there is a mythical ‘correct’ algorithm out there, possibly being designed as we speak by all the mythical bright, creative, caring, results-orientated but ethically-minded new graduates who are going to be tempted into teaching by having their pay decided by factors more or less outside of their control…

I’ll come out in support of merit pay as well, but I don’t have the perfect algorithm either. I don’t think that any private industry has one. Either you can game the system or it is so horribly complex that you might as well flip a coin to determine merit.

However, I do know that from my school days and from seeing my daughter’s teachers, it is very clear to me. There are some very good hardworking teachers out there that need to be compensated more. Everything we can do to keep them happy should be done.

There are some teachers out there that shouldn’t be working at Taco Bell. It would be an insult to Taco Bell employees to have some of these teachers working with them. How they got a college degree amazes me. They shouldn’t be around kids at all. I don’t know how that they dress themselves in the morning.

Those teachers should be fired. It is a crime that the good teachers get the same pay at the lousy ones…

I think an ideal algorithm will have two features. One, it will have a lot of controls (ie, tries to correlate how each student was performing previously, how he performs currently in other subjects, etc.) Included here would also be any other mathematical tricks. This basically doesn’t cost anything, although it has to be cleverly figured out.

But the second feature is to make the tests more sophisticated. No more simple computer-graded multiple choice. Not exclusively, at least. And, after all, the teachers already do such labor-intensive evaluations in the day-to-day schooling, but it isn’t done for “standardised” tests. It has to be. To make the evaluations consistent, you have to spread out the grading among several evaluators, and probably again apply some math to correct for differences among evaluators further. And why stop at exams. Some homework may also be subject to such treatment.

With modern technology such a scheme would not be impractical, as would have been the case 15 years ago.

If there isn’t enough money to pay teachers well now, where will the money come from to pay them extra for their outstanding merit?

And what person enters teaching because of the money?

The point is that we want better people to enter the profession, and money is a meaningful incentive: I didn’t enter teaching for the money, per se, but if it paid a third less than it does, I wouldn’t have become a teacher.

You know what I would do? I would 1) invest money in finding ways to control student behavior more effectively. In many schools, the children are out of control and teachers are left to basically wrestle students into behaving 7 hours a day. It’s demoralizing, difficult, and unrewarding work, and classroom management, more than anything else, drives good teachers of the classroom. As the case stands today, at many schools, classroom management is seen as totally a teacher’s problem, and requests for help will be turned into discussions of what you are doing wrong. While a skilled teacher can take steps to make a classroom more orderly, there are enough places where this is such a systemic problem that I really feel like there has to be some sort of systemic way to control it.

  1. In addition to systemic student behavior reform, I’d give principals the ability to negotiate salaries within a 10-15% range. Every middle manager in America is given this discretion, so we should be able to trust principals on average. That variation in salary could be used to retain the best teachers and recruit better ones. Principals would have an upper limit on total salaries for the building, so they couldn’t just give everyone the top of the range. Some principals would probably just give their favorites the 15%, but those principals would lose their best teachers. Principals with a weak department would have a tool to go recruit strong teachers in that area.

Teachers aren’t salespeople, and merit-pay systems don’t seem like the best model. But some sort of discretion in salaries is needed.

Oh, I knew why, and I should have stated that. But you explained it really well and did my homework for me :slight_smile:

Ack…I would NEVER want my principal to have the ability to negotiate salaries. They say those can, do. Those who can’t, teach. And those who can’t teach…become principals. To be fair, my principal now is fantastic. He’s what every principal should be. But the other seven? Awful. They had no business being principals, much less in charge of my salary.

(And principals aren’t middle managers in the corporate sense. School Districts have a finite amount of money to work with. We can’t make more and better widgets and make more money. In many ways, of course, that can be an advantage. My salary stays the same no matter what the economy does. In general, I won’t make less money, and my job security is amazing compared to corporate America. However, I also don’t make nearly what someone with my educational/experience level could in another job. I also don’t want those other jobs.)

I like your idea #1. Every time there’s a “bash the teachers” thread, I’m always ready to roll up my sleeves and defend the profession…but there are many, many shitty teachers out there. Maybe most teachers don’t sign up for the money…but I have no idea why most teachers sign up at all. I shudder to think that perhaps all professions are like that…what if doctors, lawyers, fire fighters, police, etc. all have the same percentages of lousy practitioners as you’d find in teaching???

I need to go rest…

Well, it seems like a principal’s judgment is at least as likely to know who their good teachers are as is some super secret formula. But I agree that we really need better principals overall. Weaknesses there are both more pervasive and more damaging than among the teaching staff, IME.

Here’s the thing, though. If my doctor is lousy, I can go to a new doctor. I can also hire a different lawyer. Same with an auto mechanic, a dentist, a barber, a preacher, plumber, bartender, contractor, and a/c repair guy. There are market forces here that tend to put the bad apples out of business.

But, if my daughter has a bad teacher, and I don’t have money for private school, she is stuck with that teacher. And what amazes me is that even the good teachers tend to get a union mentality and rally around the bad teacher like they are in a brotherhood or something.

Just an aside, though about the criticisms of “teaching to the test”. If the test is a good indicator of knowledge, shouldn’t teachers do this? If they teach to the test, and the kids score well on the test, isn’t that an indication that they have learned the necessary material?

If no, doesn’t that mean that it is the test at fault? In other words, that the test is not as good an indicator as thought?

Moving thread from IMHO to Great Debates.

Multiple choice tests are far from perfect. But more than that, they test the lowest-common-denominator. You are supposed to learn a certain amount, and the test is supposed to check if you learned the important stuff (a subset). If you teach for the test, you’re not teaching everything you should be. If you change the test to include everything, kids will be pissed that it’s “too hard.”

Actually, I think the latter isn’t much of a problem, and curved “hard tests” that test “trivia” are necessary.

That’s one point out of several. But yeah. We need to rethink our tests so that “teaching to them” becomes much better. The tests can even evaluate more complex things than pure book knowledge if they include written sections.

Actually, those market forces are somewhat ineffective, and the bad apples continue being in business. (Well, for doctors, a/c repair guys, and possibly preachers. Not so much for barbers or bartenders.) Evaluation of doctors, and publicising those evaluations, should be a major point of healthcare reform.

But, yeah, it’s messed up that you can’t change teachers even if you want to.

There is no question that merit pay is necessary for teachers. The problem is that the school system is essentially socialized, so there are no larger incentives for schools or districts that would lead them to figure out how to determine merit.

You know, if software companies were all run by the government, and everyone earned fixed salaries based on union contracts, and someone decided that merit pay is necessary to attract better programmers, can you hear the arguments?

“How are you going to measure it? If you just go by lines of code, the programmers who write tight code will be punished! There will just be an incentive to pad code.”

“What about the programmers who are assigned really hard problems? They won’t produce as much code as someone who gets the easy tasks, will they?”

“If you grade them on bug count, there will be an incentive to avoid volunteering for the really complex jobs that tend to be more buggy.”

And you could go on and on. And yet, we pay pogrammers on merit today, and don’t really have difficulty doing so. How come? Because the competitive forces of the market have forced us into figuring it out. Merit isn’t any one thing. It’s not a single measure. It could be everything from attitude with co-workers to willingness to take on tough tasks, to being on time with promises, to being creative and punctual, or whatever. In different shops it may mean different things. It may even mean different things from person to person. And yet, we figure it out.

We do so because the manager of the project is held responsible for quality and budget. And his manager is also held accountable. Ultimately, the ones who are good at figuring out merit wind up with the better teams, and rise in power where they have more of an effect. They teach their skills to their replacements when they get promoted. And so it goes. It becomes yet another job skill that people learn.

The best way to figure out teacher merit is to put the principal on the hook for the performance of his school, and then to let him figure out what teachers he wants to reward in order to meet his goals. And the best way to do that is to turn parents and students into customers. To do that, you need competition. So make the public schools compete on an even footing with private schools.

A big problem in education is also parental disengagement. And that happens because there is no incentive for them to engage. We’ve turned education into a state-operated entitlement, and the parents dump their kids with the state for a few hours a day and expect them to be educated. We have to break that pattern. And to do that, you need a reward/punishment system.

Perhaps with more school choice, with educators free to kick kids out of the school if they don’t measure up, we could create a system were if your kid gets kicked out of the top tier schools, the next tier down requires the parents to volunteer for a couple of hours of week as a condition of enrollment. There are many other ways to do this. If the kid is disruptive, the school could charge a premium for the added attention needed - said premium being paid for by parents, either in cash or volunteer time. On the other hand, there might be a discount over the voucher amount for a kid who improves a GPA by a certain amount, with the parents getting to keep the difference.

These are not specific proposals, so don’t bother trying to nitpick them. I can already spot flaws in these specific ideas. But the broader point is that if we really want to improve education, we need to find ways to bring in the equivalent of market forces to modify the behavior of schools, teachers, children, and their parents. And we need to do this while making sure that disadvantaged kids get the opportunities they need and that everyone who desires an education can get one.

What’s all this algorithm crap? Use ratings of students from previous year, fellow teachers, parents, and administrators. Everyone know who the sucky teachers are, it’s just that there is nothing you can do about it currently.

I am really confused, don’t most of you work at companies that have some sort of merit-based pay? I worked at 4 different companies over a 25 year career and they all used merit pay. It wasn’t perfect, but what is?

I generally agree with what you are saying, except that you have to be careful not to put responsibility too far from where it is needed. In theory you can just go after the principal and expect it all to trickle down, but in practice that won’t work as effectively as going after the teachers directly. Similarly, while you can and should incentivise parents (with forced labor >=), you should also go directly after kids. Punish/reward them.

Yeah, I mean it doesn’t have to be perfect. Even a little will go a long way. But more will be better. “Yearly evaluations” and the threat of being fired keeps people on their toes in the commercial sector, but i think it doesn’t push them anywhere close to their full potential. Better, more fine-grained algorithms will do that (and employees who are driven well by their bosses succeed because the bosses run such algorithms in their heads–but we want it to be more formal and automated).