Thursday, July 17, 2008

The Theory behind Merit Pay

Liam Julian posted a couple quick blurbs on merit pay today, including one partially in response to a comment I left on the first. Merit pay is, as I'll explain, fine and dandy on paper -- but I still worry about its feasibility in real life.

By "merit pay," I mean plans to pay better teachers more money -- which has taken all sorts of forms in different schools and districts.

The basic theories behind merit pay, as far as I can tell, are that:
-Teachers will work harder if they know that better teaching will result in more money
-It's more fair to pay teachers based on how good they are than on seniority or education
-Successful teachers are more likely to stay in the profession if their success is rewarded
-Brighter and more driven people are more likely to enter the profession if they know that their success will be rewarded
-Less successful teachers are more likely to leave voluntarily if their pay isn't advanced

All of these are eminently reasonable assumptions. On paper, they make a lot of sense. It wouldn't surprise me at all if any or all of these were borne out by experiments with merit pay.

That said, I also refuse to assume that all of these will be borne out in real life. Economists like to imagine that everybody is a "rational actor" that acts in their own best interest. That theory is, generally speaking, usually true. But there are exceptions -- large exceptions. Not to mention practical hurdles. And these mean that there are a lot of unanswered questions about merit pay, including:

-How much harder are teachers able/willing to work for more money?
-Would teachers motivated by money behave differently from teachers motivated intrinsically?
-How many more people would consider teaching if merit pay were common?
-How large of a role do salaries play when teachers leave the field?
-What types of people would enter teaching if pay were different?
-How well can we measure how "good" a teacher is?
-Will teachers buy-in to any measure of their success?
-Can a rewards system that is both fair and easy to understand be created?
-What type of behavior should be rewarded?
-What types of challenges or additional responsibilities should be rewarded?
-Do teachers know how to alter their behavior so that they will be more successful? (i.e. if teachers work harder, will they necessarily be better?)

I see multiple meritorious answers to all of these questions. If you'll bear with me, I have two quick anecdotal pieces of evidence that point in opposite directions. When I was teaching, I lived, ate, slept, breathed teaching. I gave it my all. If somebody had offered me $1 million if my kids improved, it wouldn't have changed my behavior any b/c I was already doing everything I could. A friend of mine, meanwhile, taught in a similar situation. He came in and out-performed all of the other math teachers in his middle school (as measured by student test scores) and then left, in part b/c he wouldn't be paid any more the next year as a result of his success -- he found it quite depressing that he would continue to earn only his 3% (or whatever) salary step bumps and make less than the older teachers in the department no matter what he did.

In short, I see arguments on both sides of the coin. I, for one, refuse to assume that merit pay either will or will not work. In order to answer that question (prepare for the most hated four words in education research) more research is needed.


RDT said...

Something I've never seen discussed is what models of merit pay are effective in similar contexts to public school teaching.

What models to private schools use? Do they use test scores? Do they use subjective evaluations? How does it affect morale?

Most colleges have something a like peer-evaluation system. What are the lessons from those?

One of my beefs with merit pay advocates is they seem to ignore a whole world of actual practice and potential data, and instead pull out a the simplest Econ 101 "money motivates people" model.

Anecdotally (and I realize there is a difference between anecdote and data) most people I've talked to who work in areas where their boss's evaluation determines their raise tend to feel the whole process is a joke -- either everyone gets the same anyway, or the bosses' judgments seem arbitrary. And managers I've talked to tend to resist pressure to do significant differentiation.

Many academics seem pretty satisfied with the peer-evaluation processes, but it's pretty labor intensive -- and usually focuses on research rather than teaching, in part because that is seen as easier to evaluate.

Anonymous said...

Interesting topic. I've always been suspicious of merit pay myself because I've seen principals evaluate teachers in rather arbitrary ways. As you noted, how would "success" be measured?

One thought: My friend works in a school that implemented a merit pay system recently. She said that, as a result, teachers are less likely to share their lesson plans, etc. with other teachers.

Under the school's system, only a certain, limited number of teachers can receive "merit pay" in a given school year. Therefore, if Teacher X gives Teacher Y some great lesson plans or classroom management tips, Teacher Y's merit pay application gains, while Teacher X now has a reduced chance of garnering one of the limited merit pay slots. Interesting twist...

Roger Sweeny said...

The question isn't whether we should have a system that pays teachers differently because some are better than others. The question is whether we should switch from the present one to something better.

Just about every public school system in America today already has a merit pay plan. Most systems have one pay scale for teachers with a Bachelors, a higher scale for those with a Masters in Teaching, a yet higher scale for those with a Masters plus 30 hours of graduate Education credit, and so on.

It is assumed that the more education courses you have taken, the better teacher you are. Most of my fellow teachers laugh at this idea. But most of them take extra courses anyway. Education courses are highly subsidized and the "payback period" is relatively short.

Most teachers are fairly risk averse and don't want to switch from a system they know to one that they don't. The people who run and/or teach in the ed. schools obviously don't want to lose their present privileged position. Most of the people who run the state ed. departments and staff the legislative ed. committees come from the ed. schools, and aren't likely to rock the boat.

My students are shocked when they learn that their school system makes no attempt to find out how well their teachers are actually teaching. It just seems crazy to them that the powers that be look at nothing but credits and seniority when determining pay. However, few people in the system have an incentive to change things.

Dustin Cornelius said...

Like the author said, merit pay does sound good on paper, particularly that good teachers will tend to stay if rewarded and mediocre teachers will leave. However, the current system already accomplishes this to some degree. The pay is typically high enough to encourage the people who really want to be there to stay, but low enough that those who aren't committed will tend to quit.

Being a teacher myself, I would like to see more compensation for teachers, but I don't think merit pay is the way to accomplish this. Linking a teacher's pay to someone else's idea of good teaching could unfairly alienate certain teachers by giving parents and administrators an unfounded bias in which teachers are better.

Visit Thoughts on Our Education to join more discussion on education.

Unknown said...

attorney dc makes the important point; if we reward a teacher because his students did well one year, do we take the bonus away if next year's students don't perform as well? And do we fault the teacher? Student? Parents? Society.

Merit pay is a problem in gestation. At birth, it will be suggested that it be drowned!

Anonymous said...

TFT: I remembered another piece of the merit pay program at my friend's school: Each teacher may only receive merit pay a certain number of times over his or her career at the school (I believe the number was 3 times, but it may have been higher - 4 or 5).

In any event, merit pay as a way for rewarding excellent teachers is a bit odd when (1) Only a few teachers per school can receive the money each year - and must compete against each other to receive it, and (2) Any teacher in the school - no matter how talented - can only receive the "merit pay" for a couple years out of their teaching career. Very odd, in my opinion.

P.S. This instance is in a Virginia public school - not sure how merit pay is playing out in other parts of the country.