I had good, bad and average teachers all through grades 1-12...in private religious school. Note: there was no union, K-8; and there was a ridiculously weak union for SOME of the faculty 9-12.

Sorting them out is most problemmatic. Letting the principal do it is beyond silly. The principal will give the goodies to the inner circle... regardless of how good, bad or average a given teacher. That's just the way it is and perhaps you have to be familar with public school culture to realize how axiomatic that is.

Basing salary on test scores seems utterly impractical. The teachers that get the swiftest classes, wiil score the highest, naturally. I suppose a complex formula could be devised where IN THEORY teachers of kids who made the most statistical *improvement* over a given period of time could be given a bonus. The reality is that public schools are highly politically charged environments where such an outcome would be almost impossible to deliver.

>>>Is the only way to say all teachers are they saem?>>>

I'm not sure. Are we really sure this is such a problem? Is the fact that teachers are paid the same really "the problem" with what's wrong with the public schools in Amnerica. I seriously doubt it.

>>>pay them on the basis of how long they've been teaching?>>>

Theoretically, at least, one becomes more proficient at somethin over time. So a 10 year teacher should be better at it than a rookie. I agree that this model has limits. Some people burn out and/or get complacent. Younger people tend to have more energy... a huge asset when you're dealing esp with younger kids.

There are pluses to scaling pay to length of service, however. There is ENORMOUS turnover in teaching. This is a given. Scaling the pay to longevity tends to mitigate that and promote stability.