Teacher Pay

by Harry on December 15, 2004

The latest issue of Education Next has three interesting articles on teacher pay in the US. All three articles attack the uniform salary schedule that is standard in union contracts. Teachers are normally compensated according to three indicators: years of service, numbers of university credits earned, and the welath of their district’s tax base. This means teachers who are better, or in shortage subjects, or work in schools for which it is more difficult to recruit teachers, are not paid more. Gym teachers get paid the same as Math teachers, despite the fact that it is much more difficult to recruit qualified Math teachers; inner citiy teachers get paid less than suburban teahcers even though it is more difficult to recruit to inner city schools.

Al three articles suggest alternatives to the current arrangements, and the first, by Brad Jupp, a Denver union leader, describes the real alternative they have established in Denver. He reports the interesting finding that his own members strongly supported merit pay:

bq. Though Denver had a typical salary schedule (see Figure 1) our data overthrow many of the preconceived notions held by teacher unions, school administrators, policy leaders, and opinion makers about how teachers perceive compensation systems. Since 1998 our union has asked its members what they thought about incentives for “teaching at schools with the highest percentage of high-need students.” By 2003, when the last available survey was conducted, the number of people favoring these incentives had reached 89 percent. The percentage of teachers who favor incentives for “teaching in content areas of short supply” is only slightly less, at 82 percent.

So, to put my cards on the table, I’m completely in favour of paying Math teachers more than Gym teachers, and English teachers more than Guidance Counsellors (not only because I’m married to one, either). I’m also strongly in favour of paying inner city schoolteachers more than suburban teachers. But I am very sceptical, not on principled but on practical grounds, of proposals for paying better teachers more than worse teachers. Here’s why:

There are three ways of deciding who the better teachers are. The first is by using objective indiciators like numbers of college credits. This is already built into the salary schedule and is, I think, less-than-ideal: it provides a nice income for Ed Schools, but teachers can take academically questionable courses, and get bare passes. There will be major resistance to requiring teachers to take worthwhile classes in order to get the salary bumps.

The second way is the increasingly popular ‘value-added’ method (which Kerry, for example, implied in his education platform). The idea is that teachers will be judged better if their students improve more than the students of other teachers, over time. This is, in principle, a great idea, and in practice an incredibly bad idea. If a student takes more than one class in the course of the year it is going to be impossible to know whether the improvements he makes are attributable to one class rather than another. Suppose the English department has several good, very hard working, teachers, who improve their students writing in the course of 10th Grade. Those students also improve dramatically in their Social Studies classes, because they are learning a lot of relevant history and improving their writing in their English classes. Their Math skills stagnate because their Math teachers are as incompetent as their Social Studies teachers, and their English teachers are not improving their Math skills at all. The Social Studies teachers are no better than their Math teachers, but score well on the value-added scale because their coleagues in English make them an inadvertant gift.

On top of that, teachers typically do not have enough students over the course of the year to get statistically significant results: and nor should they. Over the course of an entire career they probably do have enough students; but we are not going to defer raises till retirement. Finally, we do not know enough about the effects of several variables over which teachers have no control (like student mobility, the effects of classroom interruptions, etc) to control for them.

My favourite method of evaluating teachers is by having principals make actual judgements about quality. The principals would know something about what made a teacher a good teacher, and would visit classrooms, and observe teachers, and promote good teachers but not bad ones….Well, you can already see the problems with this one. Unions are very resistant to contracts that would allow this mechanism to work, and I ahve a great deal of sympathy for them. The transition problem would be huge. We currently have principals who have very little experience and understanding of teaching, and who would predictably behave in arbitrary and/or ill-informed ways. Unions fear that giving these people the kind of power I think Principals should have would be a recipe for cronyism, unfairness, and divisiveness. And from my limited expereince of actual schools they are right. Principals would not know how to use, and do not want, this power. There’s a kind of deadlock, and its hard to see how to overcome it. Value-Added ‘objective’ measures of quality are almost certainly NOT the right way forward, though.



Stephen Karlson 12.15.04 at 4:03 pm

Once upon a time, principals were promoted from the ranks of the teachers, often after doing a stint as disciplinarian or as department head. What has changed?


BigMacAttack 12.15.04 at 4:28 pm

‘This is, in principle, a great idea, and in practice an incredibly bad idea. If a student takes more than one class in the course of the year it is going to be impossible to know whether the improvements he makes are attributable to one class rather than another.’

That is why you combine it with some form of human qualitative review. You combine one and two. It could be a principal or it could be a separate position. And if the principal’s pay depends on the school’s overall performance as compared to similar schools they will get the expertise and desire quick enough.

‘On top of that, teachers typically do not have enough students over the course of the year to get statistically significant results: and nor should they.’

Sorry, what? 5 or 6 classes with as few as 12 students would be 72 students. We are probably talking in excess of a 100 students a year.


Jake McGuire 12.15.04 at 4:33 pm

Possibly more appreciation of the difference between teaching and administrative/management skills? I know that good engineers often make lousy managers, and would imagine something similar holds for teachers.

There are many more cynical explanations out there.

Asking the question “how can this proposed system be gamed, abused, or cheated around” is an extraordinarily good idea, and I think that Harry does a great job of it here.


bull 12.15.04 at 4:49 pm

I will freely admit that I know nothing about teaching beyond having spent many years of my life being taught. However, in a normal business the better employees are paid better, as they should be. If they’re not paid better, they go to a company that will pay them better, and the company they left suffers the consequences. I suspect schools are the same, except to the extent they rely on (exploit?) the altruism of teachers who are in it for more than the money.

In addition, one of my favorite explanations is, “sure, sure, economic theory may work fine in theory, but my business is different; you just don’t understand; it’s complicated you see.” Usually this phrase is used to justify subsidies or protectionism. Here, it’s used to justify inefficient personnel compensation schemes.


Cranky Observer 12.15.04 at 4:53 pm

> That is why you combine it with
> some form of human qualitative
> review. You combine one and two. It > could be a principal or it could be
> a separate position

I realize that US teachers unions are considered hidebound and an obstacle to progress today, but it wasn’t so long ago that the opposite was true. You have to read the history of school boards and the New York and Chicago teachers unions in the period 1930-1960 and then the 1970s to understand why this proposal is anathama to them.



nihil obstet 12.15.04 at 4:55 pm

And do we know what good teaching is? The testing mania currently underway in the U.S. assumes that a good teacher produces students who on average score well on standardized tests. Between student mastery of a body of facts, ability to apply necessary skills, independence and creativity of thought — there’s enough complexity there. Add to it the fact that students are very different and what works with some will leave obstacles in place for others.

Discussions of teacher merit outside consensus on the purpose and results of education are at best close to useless.


catfish 12.15.04 at 5:03 pm

Bulls comment does nothing to answer Harry’s specific points. Instead, he simply restates the theory that merit pay will work because it does in other industries. It seems to me that proponents of merit pay must answer these particular objects. Even if it is conceded that competitive markets will work in every situation imaginable, there is still a need to create a mechanism to make sure that this new market does not degenerate into petty cronyism and petty backbiting.


Ben Stanfield 12.15.04 at 5:04 pm

This is why I’m happy to work for a non profit educational organization that’s working to strengthen the principal’s role as an instructional leader. Only by ensuring that principals are well versed in the best practices of teaching can your third system work well. But it does require a fundamental change in how we train and think of school principals.

I don’t want to sound like a blatant advertisement, but for those interested, check out The National Institute for School Leadership program to see how we’re working to ensure principals have the tools they need to become effective instructional leaders. Disclaimer: I work for NISL’s parent organization.


David 12.15.04 at 5:10 pm

“However, in a normal business the better employees are paid better, as they should be. If they’re not paid better, they go to a company that will pay them better, and the company they left suffers the consequences”

That’s a lovely theory. It is also, many cases, utterly untrue. Witness the compensation of many CEOs.


Randy 12.15.04 at 5:10 pm

I hope people understand that Education Next is the educational equivalent of The National Review or The Weekly Standard – only not so august. It’s a conservative rag. They are always looking for ways to weaken teachers’ unions or public education.

That being said, I’d be very much in favor of teachers in city schools getting more than teachers in suburban schools. Right now the opposite is true. I have known lots of teachers in NYC who could not say no to suburban districts offering them $20,000 more in salary. To reverse that would take lots of money, would require the end of local property-tax-based funding of schools, and would therefore require some mitigation of the US tradition of “local control.” With No Child Left Behind, local control is almost a joke now anyway. But the federal govt is only paying about 5% of the cost of education. So approaching the idea from the starting place of what teachers should get paid is not to get very close to the Rubik’s cube solution to the whole problem.

As for differential pay by subject. It may end up being necessary. Schools are nowhere near being able to get the expertise they need in technology or even math and science. The example you use of English teachers doesn’t exactly work though. I’m an English teacher, and unfortunately for us, English is never a high need area. It’s important, sure. But there aren’t that many jobs for English majors, so there’s usually not a shortage of English teachers. That means, if you differentiate pay, you’ll have math teachers making lots more than English teachers. In most school buildings, that’s not going to seem very fair.


catfish 12.15.04 at 5:32 pm

I would like to modify my earlier comment, slightly. I think that Harry is right to argue that competition is good, but we must look at the transition cost. Will Baude has a post up a Crescentia Sentenia (sic) in which he argues against a radical market reform of law school instruction on Hayekian grounds. In some ways, I think that this applies to public education. It is absolutely necessary to increase the pay of Math, Science, and inner city teachers. These are competative reforms that do not disturb the essential workings of public education. They should be implemented way before merit pay is considered.


Laura 12.15.04 at 5:51 pm

National Board Certification is a great, non-biased way to evaluate high-performing teachers for the purpose of a pay raise.

What if the gym teacher is the best teacher in the school? He incorporates math and science into his lessons by having students measure their sprint time and distance and calculate rates. He teaches editing techniques as students rewrite drafts of their “Healthy Lifestyles” reports…Gym teachers should not necessarily be paid less because they are gym teachers. And guidance counselors – they’re some of the most important people in the school! (This from a former science teacher.)

Paying inner city teachers more would go a long way toward improving those schools. But many of the problems of teacher retention everywhere will not be addressed until Americans are serious about education funding and reform. Passing legislation requiring testing to make sure schools are up to par is stupid unless you provide the tools necessary to make real change…innovative leaders, more behavioral support staff (like guidance counselors), more pay for ALL teachers, alternatives to academic courses for the non-college bound.


Dubious 12.15.04 at 6:16 pm

Merit based teaching rewards, as for instance in rewarding those who teach in schools with the ‘highest percentage of highest need students’ seem to focus on rewarding the moral goodness of teachers. We should not seek to reward merit, but rather to reward performance.

Under that specific proposal, won’t teachers who want to maximize their pay gravitate towards difficult districts, but have no incentive to improve their performance?

I think that instead of concentrating on merit, we need to concentrate on performance and productivity. Yes, there are many important things to be taught that are hard to measure on tests. But I think a value-added approach is really the way to go.

Sample size problems? Public sector wages are subject to very low variance. If 10% of a teacher’s total compensation were performance-based, a sample size of 20 students would probably introduce a variance in pay of what, like 1 or 2%?

Having principals assess productivity… I don’t think that’s ever going to happen. Many districts have extremely low levels trust between administration and the faculty, and having the principals judging performance would hardly help.

Higher pay for Math teachers and thus lower pay for English/History teachers seems like a no-brainer to me.


Russell L. Carter 12.15.04 at 6:35 pm

“There are three ways of deciding who the better teachers are.”

Nope, there’s at least four: have the teachers rank each other, and then total ’em up. That’s the way they do it at Semco, the worker’s paradise that also happens to be highly profitable.


Sebastian Holsclaw 12.15.04 at 6:41 pm

“Even if it is conceded that competitive markets will work in every situation imaginable, there is still a need to create a mechanism to make sure that this new market does not degenerate into petty cronyism and petty backbiting.”

In normal markets that doesn’t happen to a crippling degree because good employees who aren’t getting paid move to other companies which then outperform the crony-comapanies. That doesn’t work as well in jobs with high levels of public control. We should also note that publically controlled programs as currently run (and public schools in particular) are anything but immune to the dangers of cronyism and back-biting.

As for CEO pay, we can note the problem and still realize that CEOs do not represent the way a typical market for workers operates.


Cranky Observer 12.15.04 at 6:50 pm

> That doesn’t work as well in jobs
> with high levels of public control.

It also takes a really good teacher 3-5 years to get in tune with the rythyms if the school and community. So although the /could/ pick up and move, they would prefer not to, as it would be damaging to the community.



Laura Mck 12.15.04 at 7:00 pm

harry, as we’ve discussed, I think that better teachers should get paid more. Evaluations are imperfect and subjective, but they work. I’ve really learned a lot about how it’s done after my husband went into the private sector.

If teachers can figure out a way to grade students, then can’t we figure out a way to grade them?

Evaluate teachers based on a portfolio assessment. Administrators, fellow teachers, and parents could all weigh in. Actual student performance could figure in, but only as one factor among many. Teachers could submit one stellar class plan.

Obstacles: 1) the principals don’t want that responsibility. Well, let’s retrain them. Make recruitment and training of better administrators a higher priority.

2)The unions. My first thought was “fuck ’em.” But that isn’t polite or sensible. I actually think that the union power is weakening. The fact that Kerry put forward your plan #2 is an indication of their weakness. The unions also might be amenable to this plan if everyone got an across the board pay raise, but the good ones got more.


BigMacAttack 12.15.04 at 7:01 pm

I did answer those objections. You just aren’t listening.

There should be enough students, say about a 100, for a statistically valid sample.

Combine options one and two.

Human qualitative performance appraisal would help eliminate cross variable problems. In Harry’s example the performing English teachers high qualitative assessment would help boost her pay even higher. While the poor Social Studies and Math teachers would have their pay decreased to average or maybe slightly below average by their poor qualitative assessment. More importantly it also works in reverse. The poor teacher dragging down the good teacher. The poor teacher is doubly punished while the good teacher would still receive at least average or maybe better than average, depending on the weighting, pay. The poor teachers will quickly be gone.

Basing the principals pay on the schools success versus similar schools would provide incentive to acquire the needed expertise. Assuming the principal was the one evaluating performances.

The objective measurements would prevent cronyism from occurring except at the edges. Glaring and consistent discrepancies over time would stand out like a sore thumb. Basing the principal’s pay on the schools success versus similar schools would be an even stronger incentive. The principals would have a string financial incentive to refrain from cronyism.

No, not perfect. But a very workable approach and one that answers Harry’s objections.

But of course as any teacher will tell you their performance is immeasurable by any combination of qualitative and objective analysis and their contributions are severely undervalued at any price.


Chad Orzel 12.15.04 at 7:17 pm

My favourite method of evaluating teachers is by having principals make actual judgements about quality. The principals would know something about what made a teacher a good teacher, and would visit classrooms, and observe teachers, and promote good teachers but not bad ones….Well, you can already see the problems with this one. Unions are very resistant to contracts that would allow this mechanism to work, and I ahve a great deal of sympathy for them.

It should be noted that, anecdotally at least (based on my father’s stories from thirty-odd years of public school teaching), the principals aren’t really wild about this, either. At least, they’re not big on the “visit classrooms and observe teachers” part of the process– they’d be happy to have more salary control, but not to do the work to actually justify their decsions.


catfish 12.15.04 at 7:22 pm

But would these pay for performance schemes be worth the effort? Would the differences in salary be enough of an incentive for bad or mediocre teachers to work harder? Couldn’t the time spent compiling portfolios or gaming the system be better spent in other ways? Is it worth the acrimony (public salaries are easily available to anyone who wants to see them)? Wouldn’t it be more cost effective to simply make it easier to get rid of underperforming teachers? You could put this reform in place by coupling it with big across the board pay increases.


Kimmitt 12.15.04 at 7:32 pm

At the magnet high school I attended (IMSA), the teachers were on 1-3 (and occasionally 5) year contracts, based on things like student evaluations, administrative evaluations, and the like. I’m not sure if this was successful because the implicit competition of the State’s public school system existed or not. But I had the pleasure of really excellent teaching, and the teachers held a few votes to unionize and never did, apparently believing that unionization would be more expensive than useful.

It would seem to me that a system that might work for high schools might not work for elementary or middle schools. We should probably keep that in mind, as well.


Marcus Stanley 12.15.04 at 9:26 pm

How about evaluating *principals* based on the value-added for all the grades in their school, and then giving principals more authority over hiring and paying teachers? This addresses the problems you cited about sample sizes and spillover effects, while simultaneously lowering incentives for cronyism (because the principal suffers the effects when they put an incompetent pal into a teaching position).

Several cities are moving toward this already.


Observer 12.15.04 at 9:33 pm

One factor not explored in the comments above is the “years of experience”. It’s not uncommon for teachers with 15 years experience to earn more than double the salary of a first year teacher.

While this structure is fine for 23 year olds right out of college, it presents a severe barrier to career changes. Imagine a journalist who after 20 years decided to take a pay cut and teach English. This person’s real life experience would probably be a real boon for his/her students. Or ditto for an engineer who decided to teach math after 20 years. Alas, such situations are unusual in large part because those potential teachers are expected to start at $30k or so instead of the $60k or so they would earn if their job experience were factored into the salary equation.


oldtom 12.15.04 at 10:08 pm

I recognize the complexity of the many factors that influence student performance that are beyond the control of a teacher. I also recognize that a common theme seems to be that teachers shouldn’t be rated and compenstaed on subjective factors, that wouldn’t be “fair”.

What is wrong with you people? We have to achieve utopia before we can distinguish between better and worse? Outside of academia and the civil service, most people are rated, compensated, promoted or fired based on subjective judgements (however they may be quantified). Same same for principals.

I bet that teachers know who the good teachers are, who the good administrators are and vice versa. Why don’t you take the responsibility for what you can control?

I’ve always voted for bond issues over the past 20 years. Results? SAT scores dumbed down, international ranking down. At least some of that must be due towhat is happening in the schools. And what do I read? Don’t hold teachers and adnimistrators accountable.

What am I missing, please?


kaw 12.15.04 at 10:30 pm

Pay-for-performance would only be equitable (to teachers) if we randomly assigned students to teachers. Otherwise, there is virtually no way to tease out “treatment” effects of good teaching from selection effects. Bigger sample sizes won’t do it, entering controls for family background and school and neighborhood context won’t do it, and obtaining longitudinal data on student performance won’t do it. And I just don’t see much political support for educational reforms that entail randomly assigning Junior and Juniorette to a school or teacher any time soon…


Jackmormon 12.15.04 at 11:41 pm

Kimmit’s example of the one or three-year contract teachers who never unionized kinda scares me: how old were these teachers, and did long most of them expect to stay on in the job?

One of the questions that should be asked here is for how long do we want people in stay in the teaching profession? A lot of the seniority protections are now up in the air. Some of the restrictions against people entering into teaching are being lowered, understandably, as there’s a huge need for teachers particularly in poorer areas.

But we seem to be moving in the direction of enabling a temporary teaching workforce. I’m not really sure that’s in the public interest, as it simply takes a little while for even the best qualified people to settle into teachers. Career-teachers are finding themselves up against continually changing standards and programs. Without job protection of some sort, how can we expect anyone to take teaching on as a career?

And to answer your question oldtom, I suspect that one of the reasons schools are getting worse is that the teacher has become a political target and that parents are no longer supporting teachers as authority figures. The teacher is now the ideological enemy to be undermined, the unsympathetic obstacle to little Johnny’s Ivy League admission, the service-provided to be tipped with expensive gifts. And on the other hand, the teacher is now the psychologist, medicine-distributer, and bouncer.

I’m just praying I can get into a tenure-track position. Of course, the parents have started gunning for the teachers there, too.


Jared 12.15.04 at 11:42 pm

The idea of having teachers rate each other is another which won’t work in practice.

First, teachers rarely sit in on each others classes. At best, they can assess how dedicated another teacher is by how they relate to student outside the classroom. Inside the classroom it’s an entirely different story.

Second, departments are always small enough that personal factors and department politics will play too large a role in peer-evaluations.

Oldtom, I think we all agree that subjective evaluations are the way to go. We’re just trying to find a way that works.


harry 12.15.04 at 11:51 pm

100 students is a paltry size, given the pretty small variation in improvement that we can expect to see. Studies of school effectiveness in the UK, using much more sophisticated techniques than have been developed for the US, and much richer data (and much bigger samples — we’re talking 800 or so kids a year) cannot distinguish within the middle 80% of schools — the top 10% and bottom 10% perform differently than the others, but within the vast majority we can’t get good comparisons. Too bad, but there it is. US SCHOOLS, being much bigger, would give us better results, but not individual teachers. SO Marcus’s point has some validity. I warn, though, that constructing good value added measures is much harder than people seem to think. The Brits have worked hard on this, and I’m astonished to see these debates which have been thrashed out for years in the UK proceeding as if it is all brand new. If I were more technically competent I’d point you to Harvey Goldstein’s website at the Institute of Education — In fact I’ll do a post telling people to read his stuff in a couple of days.

I’m not opposed to holding teachers and administrators accountable at all. I think the ‘objective standards’ is a fake for technical reasons but is an inevitable response to Union power. Its also true that Laura’s bomb-throwing attitude would, I suspect, yeild good results EVENTUALLY; but eventually we’re all dead and there are many costs to be borne. Predictably by the least advantaged kids.

I’m not being quiestistic — just trying to provoke thought!


BigMacAttack 12.16.04 at 2:28 am

‘100 students is a paltry size, given the pretty small variation in improvement that we can expect to see. Studies of school effectiveness in the UK, using much more sophisticated techniques than have been developed for the US, and much richer data (and much bigger samples — we’re talking 800 or so kids a year) cannot distinguish within the middle 80% of schools — the top 10% and bottom 10% perform differently than the others, but within the vast majority we can’t get good comparisons.’

What makes you sure there is a difference?

No significant differences between 80% of sample sizes of a 100 or 800 doesn’t mean that 100 or 800 are paltry sample sizes. Unless we assume differences exist.

It does not mean that ‘On top of that, teachers typically do not have enough students over the course of the year to get statistically significant results: and nor should they.’, unless we assume that there should have been statistically significant differences. (Dear heaven semantics.) I say statistical insignificance can be significant.

Reflexively 10/80/10 seems about right. And could be quite helpful. Eliminating the bottom 10%. A conservative estimate would be that the top and bottom 15 – 20% would be encouraged to perform better by 10/80/10 results.

A substantial improvement.


Russell L. Carter 12.16.04 at 3:37 am

Teachers at least from middle school on in my experience get to deal with kids on an in/formal basis from the set of all the teachers the kids have in their schedule.

The kids provide a big part of the diffusion of knowledge of competency; the watercooler does the rest. This is arguably better than in commercial ventures. You have the benefit of adult vs. child viewpoints.

Henry, have you spent any time at all looking at Semco?


Russell L. Carter 12.16.04 at 3:45 am

Oh geez, end of a long hard happy day. Henry->Harry


benton 12.16.04 at 4:01 am

Jared –

Peer review for purposes of determining tenured status is something that is being done. Its done well in Toledo and COlumbus Ohio. Not necessarily done so well in California, but its there too. The stats from Toledo imply that teachers are much harder on eachother than administrators were.

Glenn/Bigmac I’m with you at least in theory on the 10/80/10 but I don’t know how you get the bottom ten up to the middle so easily. That’s like saying that the Mets could improve if they could just get their worst 3 hitters to hit as well as their best three. I think they are trying to accomplish this by getting Manny Ramirez. When I was teaching in NYC we pretty much has ten percent of the slots always open and were always rotating in people who were not destined to be successful teachers and trying to rotate them out when that became apparent. Not enough Manny Ramirezes. Not enough positive karma/good working conditions/money to convince people to try to become Manny Ramirezes.

There are people who blame this on the certification rules and talk about how Bill Gates couldn’t get certified in NYC etc. Actually that may be true now under the NCLBA. But when I was there, I routinely was in a staffroom filled with uncertified teachers. In fact I was one of them. Bill Gates would have fit right in if he’d felt like it. Somehow he thought being a philanthropist was a better move than coming to Hollis everyday.


Russell L. Carter 12.16.04 at 4:22 am

I see google is not forthcoming. Here’s a start:


Only part there important to the present discussion is the centrality of peer review: teachers should review each other.


Randy 12.16.04 at 7:07 am

In the interest of accuracy — SAT scores are not down, and many, many, many more students are taking the SAT test. That’s not a failure. International comparison rankings are not down. There haven’t been many, and the US has never been especially high. Given what we spend per pupil on education, we’re above where we should be. Check out The Manufactured Crisis by Berliner and Biddle.


Harald Korneliussen 12.16.04 at 7:46 am

It’s conventional business wisdom that not all sorts of jobs profit from reward schemes, especially in creative fields like computer science. Why should teaching be any different? People take those jobs mainly for idealistical reasons, it’s a well known fact, and it’s good, because a certain fervour about your field is essential in passing it on to future generations.

Being a good teacher is really hard, it’s not a given that you get better results if you throw more time into preparation etc. So even if economic incentives really did inspire teachers to try harder, a teacher would probably improve a lot less from it than your average professional.

On the other hand, economic rewards should probably be used to encourage teachers to apply for inner-city jobs, and also in competing for high-demand knowledge (science, mathemathics).


harry 12.16.04 at 2:34 pm

Russell, no I haven’t, but now I will — you’ve piqued my interest. Thanks. Don’t worry about mistaking me for henry — everyone else does.

bigmacattack — I didn’t understand a word of your response. Sorry. I must be dull at the moment. Still, I shall talk to some experts about this and figure out what the problems actually are (as opposed to what they seem to me, a non-expert, to be).

If the variance in salaries predicted is 1-2% then it is almost certainly not worth bothering with merit pay. Remember there are going to be new administrative costs, and the whole point is to make teaching more attractive to talented people so they are more likely to enter and less likely to leave the profession.

And I’m very surprised by the comments people make about the reward structures in the private sector. Even in the non-unionised private sector seniority is a major predictor of wage rates, and not of productivity. Also, as several people have said, it would be bad economics to erode the sense of public service that leads soem people into teaching at a lower wage rate than they could command in other jobs. Excactly what measures erode and what measures support that sense of public service is soemthing we might want to talk about


BigMacAttack 12.16.04 at 3:26 pm


Ok good point. But maybe if pay was increased and based on merit that wouldn’t be the case with the bottom 10%.


We consistently rank in the middle or lower.

We spend the second most even when adjusting for PPP



Peter 12.16.04 at 4:59 pm

Let’s add another variable to this list for incentive pay … teachers with very large classes should get extra money. First, it’s a lot more work for them and second, it’ll keep impoverished school districts from cutting teachers when they should be cutting things that have less impact on the student’s education. (They really shouldn’t have to cut anything in an ideal world, but …)


JennyD 12.16.04 at 6:51 pm

Harry: You don’t make sense. Why would pay a terrible English teacher more than a superb guidance counselor?

Here’s a better idea: Why not get rid of tenure in K-12 education, and have teachers work under a five-year contract in their first job, and in seven year contracts after that? It would give school districts the opportunity to fire bad teachers without hassle at the end of contract. It gives good teachers the opportunity to bargain for a better contract if the district wants to keep them.

And districts do want to keep good teachers. There’s a shortage of teachers, good or bad, and you need to keep your good teachers or you’re in trouble.

I’m not worried too much about nepotism because the turnover among superintendents and school board members is pretty rapid, so it’s unlikely that someone would be held onto because they have a friend on the board. Likewise, because so many contracts come up regularly, the board is less likely to shove aside a good teacher to make room for a pal because there will be another job open at the same time.

But Harry, teaching is not magic. It’s not like being a movie star, and some people are just born with it. In fact, some people are better skilled, better trained, and have spent more time improving their practice than others who are teachers. Let’s find a way reward them.

FYI, see the series running in the Herald-Tribune in Florida. It’s about teachers who fail the basic skills test to be teachers, and notes that kids in poor schools are more likely to have teachers who failed the test, and are more likely to learn less in school.



harry 12.16.04 at 7:17 pm


why do you think I am opposed to it. I am precisely trying to provoke thinking about how to do it.
Getting rid of tenure is precisely the kind of thing that would raise costs. Government run organisations are able to pay people less than private companies because they can provide tenure, a non-pecuniary benefit which makes the job more attractive to some talented people. If you abolish tenure you’ll pay for it by having to raise the financial reards in orde rto attract the same people. I’m not opposed to doing that in pricniple, and agree that tenure is very problematic, but if you are going to propose it you are going to have to say how you’ll pay for it. The proposal has the added problem that you’d need administrators who would make better choices than the current tenure/attrition system does. I’m not convinced we have them, and getting them will take time.

Thanks for the Florida article btw. Since no-one is probably reading any more I can safely say that Americans trying to get a grip on what is going on post-NCLB would do well to read some history of UK education; I think you’ll be able to predict the next ten years of debates here if you look at the last 20 years of debates there.


JennyD 12.16.04 at 9:57 pm

Chris Correa http://www.chriscorrea.com linked to this post today, so you may be getting more comments. He’s an education academic, as am I. There are a few of us floating around, blogging on education topics of substance. Some of the blogs are quite good and thoughtful. Come check them out (and update your blogroll:-)).

I must have misread your comment about the english teacher and counselor. No harm done. But this was an interesting conversation, one I would like to have on my own pages someday.


MW 12.19.04 at 1:54 am

I’m suprprised that nobody’s compared college-level instruction with K-12. Instructors at colleges and universities are, for the most part, not unionized. They do not have standard pay rates–some disciplines pay more than others and the ‘stars’ may be paid considerably more. Colleges and universities also compete for students and are almost universally considered to be the best in the world. Our K-12 schools are merely average among industrialized countries even though our level of spending is very high.

One obvious approach would be to try to make K-12 schools work more like our colleges and universities. How would instructors be evaluated? That would be up to the individual institution — no doubt different institutions would choose different methods and priorities.

Completely apart from that, I am quite surprised that nobody seems to have mentioned evaluating teacher quality according to how well they served their customers — the students taking their classes and their parents. Universities use course-evaluations routinely, but K-12 schools seem to use them not at all.


Teacher Wanting More 12.23.04 at 1:53 am

I am currently a teacher. So, to all those who posted comments and have never taught…please just try it for one month and you’ll see and understand that teachers do not get paid enough for the job they do, and do well.

All this conversation about merit pay is very interesting. Fine, policy makers come up with a way to pay “the good teachers” more than “the bad teachers”. Yet, at some point there needs to be some serious thought applied to the FACT that teachers, as a whole, are terribly underpaid. Our profession is one in which we have to wear many hats.

Contrary to what some may perceive, teachers are probably some of the best managers/ administrators in a school setting. You really have to be a teacher to understand the complexity of this career. There are numerous roles and responsibilities under the umbrella of educator. Especially in elementary education some of those roles are planners, nurses, counselors, managers, peace makers, motivational speakers, expert implementors of curriculum, paper handling experts, and in some ways parental figures.

Merit pay for some and not others? This whole debate is fine as long as you increase the overall pay for all first. Give teachers credit for the great jobs they are already doing; and most teachers really do a great job (no matter what the media tries to picture otherwise).

Comments on this entry are closed.