Will you please clarify your response? It seems like you're conflating 2 different things. The two things that I see here are: 1) How do we determine who is

Will you please clarify your response? It seems like you're conflating 2 different things. The two things that I see here are: 1) How do we determine who is a good teacher versus a bad teacher? 2) Should we fire "bad" teachers, or should we just fire teachers who underperform to some standard? #1 is a very difficult question. Malcolm Gladwell has a great essay on how teachers are like NFL quarterbacks: we can't really tell ahead of time who will succeed. (http://www.gladwell.com/2008/2008_12_15_a_teacher.html http://www.gladwell.com/2008/2008_12_15_a_teacher.html) #2, are you against firing teachers who are "bad", assuming we had some decent way to separate the bad from the good (not just based on test scores, or even the improvement in scores)? Because the teachers union (at least here in California), is adamantly against even firing the bad teachers. Granted, it's a moot point until we can come up with a good teacher evaluation heuristic (and I personally do not believe it will put much weight on test scores...) Also, WRT to evaluating teachers through students' test scores: have you ever seen The Wire? One of the main characters (Bunny Colvin) moves from being a cop to an educator. At the end of the year, he is forced to "teach to the test" (e.g., only cover test material) rather than helping his students learn (which he was doing before). I believe this to be an accurate criticism of why we shouldn't use student test scores to eval teachers: it screws up the incentives, and makes the goal for students to beat a test, rather than to get educated.