« New Jersey learns Kansas City’s lessons the hard way

Stop a Silly Argument: Learn from Success (Joe Nathan) »

Terry Moe on merit pay

Researcher Terry Moe holds forth on merit pay.  The column is behind the WSJ’s subscription firewall, but here’s an important excerpt:

But public schools, unlike most government organizations, do have a quantifiable bottom line — namely, student achievement, which can be measured quite reliably through standardized tests. Under NCLB and state accountability systems, students are already being tested regularly. It is a natural next step to connect the dots to teachers: measuring the learning that takes place in each classroom, and creating an incentive system that rewards teachers on that basis.

There are many ways to design such a system, and districts have a great deal of flexibility — about the portion of pay devoted to merit, about adjustments for student background, about how teachers of nonacademic subjects (like music and art) are to be evaluated, about the use of school-wide measures of performance (to encourage teamwork), and so on. It is a framework that is readily adapted to local circumstances and concerns. The idea is not to impose a "one best system," but simply to take reasonable steps, based on objective data, to see that good teachers are rewarded and mediocre ones aren’t.

I think points like these are going to be raised during NCLB reauthorization.  But as he also says, the devil is in the details.  Maybe the union will actually engage these points rather than try to dismiss them–because, as Moe says later in the article, merit pay is getting traction. 

Sphere: Related Content

1 comment »
  • Malcolm Kirkpatrick

    November 1, 2006 | 1:50 AM

    I guess I’ll have to buy a copy and read the article. here’s my uninformed reaction.

    I don’t believe that merit pay can work in the typical government school system, as these currently operate. There may be reliable standardized tests of Math, English, and (some) Science, but:
    1) would a gain of, say, 0.2 standard deviation in Math be worth the same as a gain of 0.2 standard deviation in English?
    2) what measure of teacher effectiveness would apply to Auto Shop or Art?

    It’s strange to see this analysis coming from Terry Moe, since his Politics Markets, and America’s Schools (co-authored with John Chubb) excluded standardized tests of Social Studies in their composite measure of school effectiveness since standardized tests of Social Studies did not correlate with anything (which is pretty funny if you think about it).

    A competitive market in pre-college education would develop many different pay structures. Some of these would work better than others. In the current system, insiders would turn any merit pay system into a reward scheme for cooperative supporters of the status quo, by larding classes of supporters with good students and salting the classes of whistleblowers and non-union members with disruptive students.

Leave a reply