January 21, 2007

Teacher merit pay

Being a 16 year grizzled veteran teacher, I've often read about the pros and cons of merit pay for teachers -- that is, bonuses and/or salary increases for superior teacher performance. On the surface, it sounds like a pretty easy concept. When you delve into the details, however, it gets more complicated. Don't get me wrong -- I am in no way opposed to a merit pay scheme that makes sense and is logically functional. But that's the rub. Many merit pay ideas (that I've seen) fail those criteria and/or fail to address other concerns.

Just to give you an example of what I mean by "making sense" and "logically functional," the state of Delaware wanted (wants?) to evaluate state teachers partially (20%) on the test scores of the students. On the surface, this sounds somewhat reasonable. However, since the state test (DSTP) measures ONLY mathematics and English, how exactly can an art teacher's evaluation be tied to students' test scores? Or a science teacher's? Many of you know that I am a Spanish teacher. How come 20% of my evaluation is based on my students' math and English test scores? Talk about "incentive" -- this turns the incentive on its head! Why waste my time teaching Spanish -- I should be tutoring my students in math and English!

Soccer Dad alerted me to an Opinion Journal piece in support of teacher merit pay. It states in part:

That column, "How One School Found a Way to Spell Success," described how teachers at the Meadowcliff School, formerly full of student underachievers, were promised bonuses linked to improvements in the standardized test performance of each student. (The column is available on OpinionJournal here.) The size of the bonus increased relative to the student's year-over-year test gains. A 4% improvement earned a $100 bonus, rising to $400 if the student gained 15% (some did). Everyone in the school was in the bonus plan, including the cafeteria ladies, who started eating with the kids rather than in their own lounge. It worked. Scores improved. Twelve teachers got bonuses from $1,800 to $8,600. The checks were handed out in a public ceremony. Oprah would love Meadowcliff.

Again, sounds great on the surface. No doubt. But I have questions. What subjects did the standardized tests measure? How exactly did art teachers -- or Spanish teachers -- affect these [standardized test] subjects? And how in the world do cafeteria workers affect them?? The above plan states teachers' pay was tied to student performance; so, cafeteria workers can get bonuses too ... for eating lunch with students? Wow. Apparently there were individual bonuses and group (school) bonuses. And there's a nit I have with this comment (from the linked article in the above blockquote): "This straight-line pay-for-performance formula awarded teachers objectively in a way that squares with popular notions of fairness and skirts fears of subjective judgment. In most merit-based lines of work, say baseball, it's called getting paid for 'putting numbers on the board.'" Doing away with as much subjectivity in teacher evaluations (for bonuses) is a good thing; however, the analogy to baseball is far from perfect. Baseball players have only to rely on themselves for their performance. They control all the "factors of production," so to speak. On the other hand, teachers [also] have to rely their students, obviously. That is a pretty significant factor of production with aspects outside of teacher control, is it not?

When you tie teacher pay (or bonuses) to student test performance, you need a basis from which to start -- a "benchmark," if you will. A student takes a test at the beginning of the [school] year and then continously throughout the year. You can thus verify improvement. But the way many schools are currently configured means that students have many different teachers over the years (and subjects), especially in the upper grades. This is probably a "minor" structural issue, however; subject/grade configurations could be modified to permit a more concrete measurement.

But what about my subject (again, Spanish)? Students come to my class with no knowledge (or virtually no knowledge) of the subject matter. How would a "benchmark" test work in this regard? (I am asking this not out of skepticism, but out of real ignorance.) Or, would the standardized final exam at the end of the year be a sufficient measure of my teaching abilities? If so, what percentage of my students would need to pass it in order for me to get my bonus? Does that percentage get raised each year? Or, does a certain "bell curve" grade distribution need to be realized for my bonus?

I Googled "merit pay teachers" the other day and the first page of articles dealing with the subject were overwhelmingly negative. The very first article is by education researcher Richard Rothstein and is titled "Merit Pay Won't Work." I must admit I am very skeptical of Rothstein, especially after reading [one of] his devoid-of-reality "solution" to inner-city education problems. Rothstein argues that "... despite the oft-repeated notion that 'merit pay' contributes to corporate success, it is hard to find private sector examples for such proposals." He goes on to point out various studies to support his point, but in my opinion they are weak. And as a point of personal anecdotal evidence to the contrary, my job before teaching was in credit card collections. Each employee who reached his/her monthly percentage goal received a bonus.

Rothstein then quotes Brooklyn-Queens Archdiocese Superintendent Guy Puglisi who says

"We have many schools where test scores are low. But if the teachers are working hard, the scores are not that much of an indication of anything. The teachers might be working doubly hard, but the scores are low because of the social conditions they face.

"I don't think merit pay is that great an idea ... what motivates teachers is the opportunity to do challenging work, and then being told how much they are appreciated."

If the teachers are working hard. There's certainly no doubt that what motivates many teachers is effecting positive change in the learning of youth. Personally, I did not go into teaching for the money! I don't know of anyone who did. But the fact remains that there are many teachers who are lemons -- and the question remains: Is it fair that these lemons get paid the same (or more, based on years of experience and/or education level) as an exceptional teacher?

One point where I believe Rothstein pretty much hits the mark is this: The evaluations needed to support a merit pay system in education are inconceivable with schools' currently weak administrative structures. Administrators' main focus these days is dealing with student discipline, and many are now hired solely for this purpose. Thus, I'd argue these folks do not have the aptitude to evaluate teacher performance, especially in key subject areas. But, again, utilizing standardized test scores on which to base (whole, or in part) teacher evaluations could certainly alleviate some of the "load" from administrative "judgment."

Myron Lieberman, like Rothstein, dismisses the notion that merit pay is much more widespread in industry (as opposed to education):

Conservatives often exaggerate the extent to which merit pay is the practice in our labor force. In dozens of industries, merit pay seldom applies to employees below the supervisory level. In fact, merit pay does not affect airline pilots and several other occupations in the public and private sectors. Superior performance often leads to promotions to positions in which merit pay is operative, but this is true in education as well.

I don't know how accurate that last sentence is; in my experience excellent teachers wouldn't touch administration with a 50-foot pole. Lieberman goes on to ask one of my main queries from above (my emphasis):

Even if teachers, teacher unions, and school management agreed that merit pay was a good idea in principle, the problems of implementing the idea would be difficult to resolve. To cite just one problem, how can we compare merit among teachers of different subjects and grade levels? With the best will in the world, different interests will lead to differences of opinion on this issue.

And that's just it. As I stated from the beginning, I have little hassle with merit pay. I just want a system that makes sense -- and wasn't put together to satisfy someone's notion of political expediency, like the proposed Delaware teacher evaluation plan. In addition, Lieberman notes how school administrators would be wary of utilizing a merit pay plan due to [its possible] subjectivity -- "This is not something administrators look forward to, especially since it's always possible to criticize the criteria or the applications of the criteria for merit pay," he writes. It's a good point, for as Greg over at Rhymes With Right notes, just as teachers know which fellow teachers are doing their jobs, they also know what adminstrators are doing their jobs, too. You could be a phenomenal teacher who's "led" by an inferior principal, have disagreements over proper education policy (the teacher clearly being right in this case) and as a result your evaluation is poor -- "payback," if you will. Does this sound familiar to any teachers out there?

The debate will continue to rage. I recommend all teachers, especially, check out the many differences of opinion regarding merit pay at the Google link above. To restate, I personally have no objection to merit pay, as long as a fair system is established to implement it. The Meadowcliff School's example may to be such a system; I'd like to see more specifics to make a concrete evaluation. Delaware's system (not merit pay, mind you, but 20% of a teacher's evaluation that can affect pay) as current constructed is not such a system.

Posted by Hube at January 21, 2007 11:40 AM | TrackBack

Comments  (We reserve the right to edit and/or delete any comments. If your comment is blocked or won't post, e-mail us and we'll post it for you.)

Hube, what do you mean, merit pay? You are a conservative? Just thank God that you retain your position and have not yet been drawn and quartered. Teachers' unions and school administrators are not normally that tolerant. You must be flying under their radar. Hush!

Posted by: Mark F. at January 21, 2007 03:37 PM

And then you get the situation I'm in down here in Texas. We've got a merit-pay program in place due to some grant money -- except it is distribited by departments, not by teacher. If our scores increase, everybody in the department gets a bonus -- even the ones whose classes don't take the test! Oh, and the test my kids will be taking? It doesn't include ANY of the material I am responsible for according to the standards set for World History. It covers only the material these kids took through eighth grade, when they took US history to the Civil War. So what do you want me to teach? My subject, or the subject they took two years ago? And does is my teaching performance really measured by a test that doesn't include my materia1?

Posted by: Rhymes With Right at January 21, 2007 08:09 PM

You've hit the nail on the head -- merit pay is a great concept, but the devil is in the details.

Maybe charter schools and private schools are about the best merit pay model you could have -- if parents are happy with all the subjects (or happy overall), they'll pay to send their children there and all the school employees benefit.

Hard to see a completely fair system that meets all the objectives people have for education.

Posted by: churchill-in-the-wilderness at January 22, 2007 09:21 AM

Private schools and businesses aren't very good at paying for performance. Politicians says so, but the data doesn't show that it is true.

What does work in businesses and private schools is those teachers who are doing a poor job are fired. The entire teaching staff isn't measured by its test scores, but they know who the poor teachers are, and they get rid of them quite easily.

However, private schools do pay some teachers more than others. Just like businesses, they pay by responsibility and by the "market". If physics and math teachers are difficult to come by, they'll be paid more. I was hired (but didn't take) a job at a private school and was to be given much more money to teach math than a similarly experienced elementary grade level teacher would have made. I was told that I was not to discuss my wages with any other staff member.

I think that is why unions often fight against this type of pay. Math teachers are harder to get, but once they are at the job, they don't necessarily do any more work than the history teacher, so they feel that they shouldn't be paid more.

Posted by: IB a Math Teacher at January 24, 2007 02:37 PM

I am not a teacher, but am subject to merit pay in my job. I think merit pay programs have their flaws no matter the industry, however, in the teaching profession would have even more flaws. It think you article was very good and hit on some very good points.

Here via the Carnival of Education

Posted by: Lisa at January 24, 2007 08:26 PM

We have an incentive plan in place at my school. We tried to come up with a plan based on student performance, but as you said, without a fair benchmark it's difficult, and even more so for non-testing subject teachers. Our finalized plan? It's based on teacher attendance. The idea is that our continued presence will motivate the kids.

Posted by: Mister Teacher at January 24, 2007 09:30 PM

The devil is in the details indeed. Let us not, by any means, denigrate administrators who labor ceaselessly (and effectively) in the fields of discipline. Without rational, swift and effective discipline, the inmates run the asylum and little learning takes place.

However, the only way that merit pay will ever be rationally applied is if we can turn around several fundamentally deleterious aspects not only of our schools but of our society. I'll suggest that we have a culture of mediocrity wherein everyone works primarily for comfort. Once an appropriate level of comfort in lifestyle and work is attained, little or no effort goes into improvement, and that very average--if you will--level of effort comes to be seen as excellence. True excellence tends to be scorned and distrusted, even attacked, and only the very worst performance attracts negative attention.

I know I'm speaking in generalities, however, if I'm incorrect, why would anyone have any objection to immediate implementation of merit pay schemes? Wouldn't the system be automatically able to fairly administer them? Wouldn't only the truly best teachers be rewarded? However, if my hypothesis is correct, it's not hard to see why we worry about such details: We know that the mediocre, even the craven, would be likely to be seen as meritorious.

Test grades are a terrible way to measure such things, even benchmark tests, because they tend to measure--particularly in these days of be-all, end-all mandatory high stakes testing--only student ability to take a very specific type of test with a very specific content. If this is the educational outcome that means most, why do we need teachers at all? Wouldn't long, low metal buildings with thousands of individual cubicles, each stocked with a computer and appropriate software provide the necessary instruction? Once a student passed the necessary tests, their education (pre-college) would be over. Surely this would be efficient and might even have the happy benefit of being gender neutral, color blind, faith and creed blind, and ultimately politically correct.

Perhaps a good way to deal with this issue is what I'll call a portfolio method. In my classes, my students each keep a portfolio in which everything we do is kept. Every handout, every assignment, every journal entry, piece of writing, test, everything they touch in my class is kept to be returned to them at the end of the year.

With such a system, a competent administrator could, with some time and effort, see what progress a given student, or even all students, have made on a wide variety of assignments throughout a year. If we accept the idea that the purpose of education is building bigger, better brains, and not merely cramming them with short term knowledge that tests well, this makes all kinds of sense. One concept, one skill builds on previous skills and enable future skills, and if all that a student has done is available, instead of only a few test scores, a far more complete picture of student progress and teacher effectiveness can be done. This should be combined with actual principal/teacher involvement in teaching. Principals should be speaking with teachers daily about how they're solving teaching problems, which lessons they've revamped to improve them, which new handouts and ideas they're developing, and how they're adapting their efforts to deal with different classes and their varying needs.

The problem with this suggestion is, of course, that it takes actual time and effort of the part of administrators. Some don't wish to expend either, some simply aren't allowed to do such things. It would also take competence, competence to recognize actual student progress, and competence to remember what, all those years ago in a classroom dimly remembered, constituted good teaching. It takes no effort, ability or competence to look at two sets of test scores and see if the second is greater than the first, thus indicating improvement. But this, sadly, is the way of things for now and probably, for the future.

Posted by: Mike at January 24, 2007 09:50 PM