Good teacher pay strategies are never written in a vacuum: they're part of a well-thought out system of incentives and professional supports designed to attract and keep the best teachers. That's why a study on performance pay released by Vanderbilt University's National Center on Performance Incentives has us scratching our heads.

The study, which sought to measure the influence of merit pay on student achievement, paid a healthy sum of money to--up to $15,000--to middle school math teachers in Nashville if they raised test scores. While the plan had some positive impact on 5th grade scores, it generally fell flat. Most telling, it left the teachers involved in the study unconvinced that performance pay rewarded their efforts in the classroom.

First off, it's no surprise that the findings showed no correlation between performance pay and increasing student achievement, meaning that the very premise of the study might be called into question. Performance pay is a reward system designed to send strong signals that the profession honors and rewards results but, perhaps even more critically, it should increase the profession's appeal to individuals who might not otherwise consider teaching, convinced that the profession disdains excellence. It's a silly notion to think that teachers leave their "A" game at home, absent the promise of a little extra pay.

Further if performance pay can improve instruction among existing teachers, the Nashville trial wasn't designed to measure that improvement. Compare Nashville's initiative to the IMPACT system introduced in DC Public Schools. In the Nashville experiment, teachers and principals were unaware who won bonuses and were not provided with resources to voluntarily improve their own instruction. Under IMPACT, performance pay is used not to compel teachers to work harder, but to fairly reward the most successful teachers. Further, DCPS offers teachers not only the opportunity to learn from each other, but a myriad of voluntary professional development resources to improve their instruction.

IMPACT was ratified by three-fourths of participating teachers this summer and continues to be generally popular with DC teachers. The Nashville system, on the other hand, was largely rejected by its participants, who said it was unlikely performance pay would change how they teach. By testing performance pay on its own, and by making sweeping conclusions about their own findings, researchers encouraged the same old chatter from education wonks across the spectrum--declarations that the study said everything or said nothing about the impact of performance pay. We're holding out.