An abbreviated version of this article appeared in educational Horizons, Volume 88, Number 1, Fall, 2009 under the title, "Teacher Merit Pay: Is It A Good Idea?"
by Gary K. Clabaugh
President Obama's education agenda, which turns out to be George W. Bush's agenda squared, has a particular feature that could have an unfortunate impact — merit pay for teachers.
Whenever I think about teacher merit pay I'm reminded of a situation that occurred when I taught seventh grade. Our school's scarce audio-visual equipment was "stored" in the classroom of the principal's favorite teacher. The practical consequence was that this teacher, we'll call him George, had first claim on it — a privilege he routinely abused.
How did George become the principal's favorite? It wasn't that he was the most skillful teacher. He actually bored the kids half to death. His talent was boot licking. The man stroked the principal's ego like Paganini bowed a violin. And since he taught nothing of consequence nor dared anything different, he never made waves. The principal loved him for that too. This is how George got the AV equipment, as well as choice assignments; and this is what would have won him merit pay if such a thing had then existed.
Yes, teacher merit pay could easily turn into bonuses for brown-nosers. And even if standardized test scores become the only criteria, favoritism could still play a role in who gets the money. That's because the principal's favorites often end up with the easiest classes, and particularly difficult kids are quickly reassigned to some less favored soul.
Actually one doesn't even have to be the principal's favorite to gain such advantages. Sometimes being a secretary's favorite is enough. I know a school secretary who annually let her favorite teacher pick the kids she wanted in her class. The secretary was this teacher's friend and neighbor. The other same grade teachers always got, as one of them dejectedly put it, "the dregs."
Will favoritism result in unfair competition for merit pay? It's a good bet. As a matter of fact, it is a very good bet.
The idea behind merit pay is that teacher productivity will increase because teachers will try harder. And since the most skilled teachers will make more money they will stick with the job, while the least capable teachers will make less money, feel rejected, and quit.
For this to happen, however, decision makers must have accurate information about which teachers are particularly skillful in teaching subject matter, and which might make some other positive difference in the lives of children. Why the later? Simply put, academics are not all that matters. In fact, for some children poor academic performance is the least of their problems.
Many important things that good teachers do are hard to measure; and research tells us that merit pay increases job performance only when that performance can be easily measured.[i] For most jobs, including some a lot less complicated than teaching, accurate measurement is just not possible. That's why only one in thirty occupations feature straightforward performance contracts.[ii]
Significantly, teaching has never been one of the thirty. That's because the full scope of a teacher's actual job performance is notoriously opaque. How would a school administrator know, for example, which teachers actually improve the quality of children's lives? Yet what could be more important?
Most of what happens in schools happens behind closed classroom doors. That's why administrators can't really tell which teachers are smiling and friendly. They can't tell which teachers routinely extend a helping hand or offer comfort. They can't tell which teachers consistently protect the weak from bullies. They can't even tell which teachers are simply kind. Yet such things surely are at least as important as standardized test scores and probably of greater consequence. How, then, will merit pay be fairly awarded?
Suppose, for example, a youngster comes to school with a poor self-concept; but due to the patience, skill and caring of her teacher, she leaves school with a new sense of self worth. Surely such a teacher-induced outcome is meritorious even if the child's test scores remain unchanged. But will such merit be rewarded? It seems most unlikely that it will, because it can't be well enough measured.
Besides which, even if such crucial teacher attributes could be reliably measured, they still will go unrewarded so long as standardized test scores are used to determine when a school meets muster. Spotlighting any school's test scores makes it irrational for a school administrator to pay a teacher extra for anything other than improving those scores. And this is doubly true if administrators are vying for test score based merit pay themselves.
Merit pay proponents tell us not to worry; they're working on incorporating the more subtle aspect of teacher performance. But how will they ever accurately measure the many subtle but crucial interpersonal aspects of a teacher's job? That is precisely why standardized test scores will likely become the sole criteria for merit pay, penalizing teachers who focus on the whole child.
Remember too there will always be ways to game any merit pay system. No sooner was NCLB in place, for instance, than we began to read of teachers and principals changing standardized test answers or cheating in some other way. Merit pay will only make such cheating worse.
Also bear in mind that the greater the pressure, the more likely teachers will teach to the test. Out the window goes creativity and intrinsic motivation. That already is a serious problem even without the added temptation of merit pay. Add the lure of dollars and teaching to the test will become irresistible.
What is more, all this emphasis on standardized tests, in effect, puts the test makers in charge of the nation's schools. Is that what we want to do?
The best teacher I ever had was Dr. Frederick Fuhr — he taught me seventh grade world history. I had him fifty-five years ago, yet I still feel indebted. What stood out for me was how Dr. Fuhr dealt with the fact that he was a paraplegic. Both of his legs were useless because of polio. They were encased in hip to ankle braces. He struggled down the hall on crutches swinging his useless legs pendulum-like beneath his powerful torso. Despite this handicap Dr. Fuhr was a compelling teacher —if, that is, you were willing to learn some history. I still remember what he taught me about the ancient Greeks and Romans.
But what I really remember is his example. Dr. Fuhr taught me about courage and how to deal with adversity. Sadly, however, I can conceive of no merit pay system that would reward him for teaching me that. Here was a man who could have stayed home, collected disability checks and wasted his life feeling sorry for himself. Instead he was the best teacher I ever had. He also was the only teacher in the school to earn a doctorate.
I don't know if Dr. Fuhr's classes would have scored well on a high stakes test. A fair number of the kids in that room were too immature, unimaginative, angry, scared, stupid or preoccupied to appreciate what he taught. But should that cost a man like this money?
One particularly undesirable aspect of merit pay is that it will inevitably increase competition and decrease cooperation among teachers. I know of a novice teacher, for example, who was hired for a first grade position one month into the school year. The other first grade teachers were instructed to pick five kids each for transfer to the novice teacher's class. Some picked only the kids they found most problematic. Others were nice and sent a random mix of kids.Now imagine that our novice teacher worked in a school with merit pay. Wouldn't all of her fellow teachers, eyes fixed on those extra dollars, see to it that she only got their problem kids? In this way merit pay could be positively poisonous.
The random variables that one normally encounters in teaching also will eliminate fair chances of winning merit pay. Consider a teacher who has a socially and emotionally disturbed child show up on his or her class rolls, for instance? Now suppose that the administration, trying to save money, fails to support the teacher's legitimate request that this youngster be transferred to special education. Consequently, the youngster disrupts the class for the school year. Should the teacher be financially penalized for the educational consequences of a fiscally driven administrative decision?
And what about the other forms of maladministration that teachers encounter? Failing to honor a teacher's legitimate request for assistance with a particularly troubled child. Confusing important things with cosmetics. Making foolish decisions when wisdom is demanded. All such administrative malfeasance is beyond any teacher's authority. But they can have a devastating impact on a teacher's effectiveness. How are merit pay advocates going to adjust for that?
There also is the child's home life and neighborhood to consider. Research repeatedly reveals the adverse impact of divorce and separation on a child's success in school, for instance. Poverty is another factor that limits academic success. So are child abuse and juvenile gang membership.
Then there is tardiness and truancy. Lots of inner city schools have absentee rates of 25 to 30%, plus large numbers of kids who show up an hour or more late. It shouldn't cost any teacher money when they fail to teach a child who often isn't there. After all, teachers don't set the policies that discourage or inadvertently encourage such truancy and tardiness.
In short, there are many in and out-of-school factors that are well beyond a teacher's control yet have a negative influence on school achievement. How will merit pay plans take any, much less all, of these things into account? They won't, and that's the trouble. In what sense, then, is teacher merit pay based on student achievement either fair or wise? It is likely to be neither. So, instead of encouraging teachers to achieve better educational results, merit pay will likely discourage and demoralize them. Isn't there more than enough of that in teaching already?
Don't get the idea that this merit pay thing is going to blow over. Education Week reports that the Obama administration is going George W. Bush one better in this regard. The U.S. Department of Education's proposed guidelines for awarding that $4 billion in Race to the Top money includes this non-negotiable requirement. States must not have any laws in place that bar the use of student achievement data or forbid student test scores to be used in decisions about teacher compensation and evaluation.[iii] At least two key states, California and New York, now have such laws.
Underscoring the Obama administration's determination to hold teacher's accountable for student achievement, U.S. Secretary of Education Arne Duncan recently said, "Being able to link teacher and student data is "absolutely fundamental —it's a building block. When you're reluctant or scared to make that link, you do a grave disservice to the teaching profession and to our nation's children."[iv]
Maybe he is right. But Secretary Duncan is presupposing that he can accurately collect all the needed data and reliably establish links. That's a tough, tough job. Besides, he seems to have totally forgotten about all those crucial things we don't even pretend to gather data on. Aren't learning things like compassion, consideration and caring, for example, at least as important as math and reading?
Secretary of Defense Robert McNamara relied heavily on data collection during the Vietnam War. Casualty figures were used, for example, to measure individual officer's performances as well as military progress. But it turned out that the data collected did not reflect on-the-ground reality. Commanders vying for promotion repeatedly turned dead Vietnamese civilians into dead Viet Cong, for instance, creating an entirely false picture of progress.
MacNamara's obsession with bookkeeping helped us lose the Vietnam War. Is the Obama administration risking a similarly unwanted outcome in public education by over relying on high stakes test data? We shall see.
[i] Ibid p. 92.
[ii] Tim Harford, The Logic of Life: the rational economics of an irrational world, Random House, New York, 2009, p. 91.
[iii] Michele McNeil, 'Race to Top' Guidelines Stress Use of Test Data, Education Week, July 23, 2009, p. 1.