Merit pay for teachers has gotten a lot of play recently (for examples, see here). Without a doubt, the principle that some teachers ought to get paid more than others has gained political currency around the country. More and more politicians - generally a risk-averse group - are coming out four-square behind merit pay, even if it means taking on the unions.

While I'm certainly glad that merit pay is gaining ground as the "right" thing to do, right doesn't always make might. The groundswell of public support could quickly seep back into the cracks depending upon how we proceed from here. Merit pay could be doomed to failure unless governors support the careful experimentation that's needed to solve some of this reform idea's great dilemmas.

Structuring merit pay well is hard to do. The systems need both to be fair and to be hefty enough actually to impact a talented teacher's decision to enter or leave the profession, i.e. to alter behavior. As we've seen in California - where the unions have formed a formidable bloc in opposition to Governor Schwarzenegger's attempt to reform teacher tenure and institute merit pay - our collective naivet?? combined with unrestrained enthusiasm plays right into the hands of groups opposed to change.

The thorny problem of how best to determine a teacher's effectiveness - the only fair basis for deciding who gets merit pay - has by no means been worked out to the degree required for wide scale adoption. Challenges include:

  • Value-added measures of student learning, while certainly promising and the most reliable option on the table, cannot be used to measure the effectiveness of teachers who work with very young children, in high schools, and in non-tested subjects like art, music and history.
  • Evaluations by principals or peers, when done with care and consistency, do correlate highly with student learning gains. However, this method must overcome a long history of ill-designed instruments and weak training of evaluators, not to mention widespread teacher suspicions that principals will play favorites.
  • Letting the teacher decide for him or herself what goals to achieve, as Denver has done, brings its own challenges - both in administering a program predicated on unique goals for each teacher, and in the possibility that a teacher's goals may not align with those of the school, school system, state or taxpayers.

Though these problems appear daunting, policy makers must tackle them. Some experimentation is surely in order as well as some pilot programs coupled with rigorous evaluations. Meanwhile, we do know some things that offer useful parameters for moving forward:

  1. Merit pay needs to be based on multiple factors. It should always include some measure of student achievement, but it also needs to include evaluations by school principals and senior faculty. It's neither workable nor even fair to base a teacher's income on a one-shot test. Critics have a valid point here.
  2. Merit pay bonuses must be large enough to persuade teachers to do something they might not otherwise choose to do. The $1,500 bonuses currently offered by a number of states and districts are likely inadequate. My hunch is that 10-20 percent of base salary is more like it.
  3. Merit pay programs should acknowledge individual successes, not just school-wide achievements. In other words, all teachers in the same school should not receive the same bonus - though perhaps all should receive some bonus. For good reasons why, see this recent study by Eric Hanushek, Steve Rivken, John Kain, and Daniel O'Brien(
  4. Schools must help weak teachers achieve. Professional development funds ought to be directed at supporting teachers as they gain the skills they need to qualify for bonuses.
  5. States and districts require a long-term strategy for sustaining any merit pay program. Too often, teachers are promised bonuses that prove short lived, ending as soon as the first budget crunch. One idea worth pursuing is to persuade teachers in experimenting schools to give up their automatic step increases, contributing these funds to the bonus package. Tweaking the existing uniform salary schedule is a promising way for merit pay packages to survive the test of time.
  6. The resources needed to do merit pay right for all schools statewide do not exist. For now, it would be better to allocate resources to high-need districts than to spread limited resources too thin.

The teaching profession has no choice but to remedy an outmoded pay structure that is woefully insensitive to current labor force realities. Daunting though the challenges appear, there's no question that it can be done. Will it be perfectly fair? No system is - but the system we're saddled with now is remarkably unfair to teachers and, even more importantly, runs counter to what works best for kids.


Kate Walsh is president of the National Council on Teacher Quality ( This editorial is adapted from a recent edition of the Teacher Quality Bulletin.

Item Type: