Moneyball and Teacher Pay
Tuesday, February 12th, 2008 By Matthew Carr
Ever since the book Moneyball, the story of how one Major League Baseball executive used statistics in the front office to build winning teams on the field, was published in 2003 there has been no shortage of writers trying to find other professions for which similar innovations might work. In a recent New York Daily News piece, Kevin Carey makes the connection to merit pay for teachers.
Teachers who may by all appearances seem similar - same age, experience, level of education and demeanor - can be hugely different in how well they help their students learn.
And so, (New York City Schools Chancellor Joel ) Klein plans to start using something called “value-added” data to figure out differences among teachers. The data compare annual test score gains in a teacher’s classroom to statistically predicted gains, given students’ backgrounds, previous academic history and a range of other factors.
At first, school officials are going to measure the performance of some 2,500 teachers - a small fraction of the 80,000-plus in the system. There are no plans to attach huge rewards or penalties to the results, at least not yet.
But the promise of the program is clear. Down the line, teachers who make great progress with previously low-scoring students could get stellar ratings. Those who coast with previously high-performing students - or fail to bring struggling kids up to grade level - won’t.
As I’ve noted in recent posts, getting accurate data on teachers’ contributions to student learning is not an easy task. Serious attention will have to be paid to both the evaluation systems and nature of the data to be collected and used to build performance measures. But it is not an impossible task, and allowing the perfect to be the enemy of the good is no excuse.
To those concerned about fairness, let me point out: Using value-added statistics to inform teacher evaluations is far fairer than judging teachers by their students’ raw test scores. It is far fairer than judging teachers by their principals’ like or dislike of their work. And, yes, it is far fairer than making no distinctions at all. If some teachers are consistently more effective than others, it’s only fair to recognize those differences.
Also, some of my colleagues in the Department of Education Reform at the University of Arkansas have recently completed their year two evaluation of the Little Rock merit pay program. You can find the results here.
