Why I’m Still in Favor of Merit Pay for Teachers
I’m not convinced by some economists’ assertions that offering merit pay and bonuses doesn’t make employees more productive. One of the economists professing that opinion is Dan Ariely, who describes his research in Wired. Here are some of the tasks he asked his subjects to perform:
We asked them, for example, to assemble puzzles and to play memory games while throwing tennis balls at a target.
When the subjects were offered big rewards, they did poorly on the puzzles.
I have no doubt that Ariely and his collaborators ran their experiments under rigorous laboratory conditions, but it’s a stretch to conclude from them that merit pay is bad. What goes on in the laboratory is far removed from day-to-day classroom activities. Good teaching depends more on verbal and interpersonal skills than on hand-eye coordination. And a controlled experiment with tennis balls is of necessity finished within minutes, whereas teaching takes place over the course of many months. The long work of establishing a rapport with students and building knowledge isn’t comparable to putting a little puzzle together.
To learn the true effects of bonuses and incentives, it’s better to look at studies that examine how real teachers in schools respond to merit pay.
“To learn the true effects of bonuses and incentives, it’s better to look at studies that examine how real teachers in schools respond to merit pay.”
I disagree. It’s much more difficult to isolate variables in field studies than under experimental conditions. It’s very difficult (if not impossible) to justify a causal relationship via a field study.
I sympathize with your concerns about the relevant differences between teaching and the experimental conditions. Do you have any hypotheses about why merit pay might not effect attention, memory, concentration or creativity, but would effect the development of interpersonal or verbal skills?
Comment by Andrew Hanson — February 5, 2010 @ 11:22 a.m.
I explained in this post why I think the negative effects on short-term memory and concentration may not be important:
http://www.showmedaily.org/2009/09/incentives-for-teachers.html
I expect that merit pay could motivate some teachers to explain concepts better or form better relationships with their classes. I don’t dispute the finding that prospects of reward can distract people and make them lose their concentration for a few minutes, but I don’t think their performance would be hampered all year.
Comment by Sarah Brodsky — February 5, 2010 @ 11:35 a.m.
It seems like you’re reaching a little bit to explain away the findings of the research. I think the research is probably sound. Many similar experiments have suggested that individuals respond more to immediate, short-term rewards and tend to think less about the long-term consequences of their actions. Getting better at explaining concepts or developing relationships with students will require a great deal of hard work, which requires attention, memory, concentration, and creativity in both the short and long term.
I think there is a better case for merit pay than the motivation story you want to tell: it effectively raises teacher pay, making the field of teaching more lucrative to enter for talented individuals who would make great teachers (e.g., because they have interpersonal/verbal communicative skills). I don’t think the problem is that teachers aren’t motivated enough to work hard; it’s that teaching isn’t a relatively lucrative field to enter, (which SuperFreakonomics gave a novel explanation for). It would be more effective to recruit more talented individuals (e.g., by offering merit pay), than to try to get teachers to work harder.
You could also justify merit pay on desert. The teachers who achieve the highest growth for their students are likely the most talented and therefore deserve the highest pay. That system would undoubtedly be more just than awarding pay based on seniority or education (the current system). You could also do this simply by raising teacher salaries, but I see nothing wrong with offering “merit”-based pay, even though I disagree with your analysis.
Comment by Andrew Hanson — February 5, 2010 @ 1:02 p.m.
The lab experiment in this case seems a poor fit for what they are studying in the world. An experimental volunteer is probably perfectly willing to do the memory and coordination tasks because -thats the experiment they signed up for-. But when money is on the line they become nervous and botch it. They are perfectly willing to do the novel experience of the experiment for its own sake, for one day. But if it were their job to play memory and tennis ball games for the next couple decades, I would think that merit pay would show definite strong correlation with better results for the arbitrary goals.
Comment by vroman — February 10, 2010 @ 4:06 p.m.