> There's no fair and objective way to measure the performance of people... blah blah blah...
You've avoided the question, and assumed I'm arguing against measuring performance. Basically, wrong on both accounts.
I'm not arguing that teachers shouldn't have some form of review. However, considering the nature of a public school system, knowing what we know, how should teachers be reviewed? The goal is to have good teachers teaching. Simply saying "we need to review teachers" doesn't help matters, as the question isn't whether we need to. It's the method that is the question.
> the judgment of peers and superiors, augmented with whatever metrics are deemed useful.
No. That's merely saying "Their performance should be measured" with a few more words.
> augmented with whatever metrics are deemed useful.
So, once again, I repeat: What metrics are useful? You can't judge someone without knowing what metrics you are using.
> Just like most other jobs.
Most other jobs use specific metrics. When you come up for review, you know what you are being judged on. Take a sales position. The total revenue brought in, along with net profit from that revenue might be one metric that is useful. For programmers, meeting deadlines and customer needs consistently might be one of those metrics. For a writer, meeting deadlines for first draft, second draft, as well as overall sales of a book.
In the end, you are simply saying the same thing over and over again: Judge them by whatever useful means. You've yet to provide any opinion on what those useful means might be. If you don't know, just say you don't. But don't pretend like you are offering any insight. Essentially, you've answered the question of "What's air speed velocity of a laden swallow?" with the answer of "The air speed velocity of a laden swallow."
You're missing the point. I'm being vague because it's entirely appropriate to be vague. Performance in most jobs is highly subjective. People judge employees the way they judge code. There are vague rules of thumb for judgment -- software developers are expected to write solid code, use source control well, follow company standards, and keep their manager informed of their work -- but there is rarely any way to measure performance. Metrics are often disregarded, if they're gathered at all.
For example, did a writer miss a deadline after an unexpected event required changes (e.g., someone writing a piece about Bin Laden needing to make major changes after his death)? That's a pretty solid defense, but if the editor knows that the writer has been screwing up and consistently running behind, he might use the delay as an excuse to reprimand him anyway.
Did a programmer not finish his project because he was asked to support someone else's failing project? If his boss values him, he will run damage control and explain to upper management why the project did not get finished.
Were a salesman's numbers down because of bad PR from a safety scandal? If the salesman's peers felt that he expressed a flippant attitude towards the safety problems in front of customers, thus contributing to the company's bad safety image, his boss might fire him. That can't be measured.
Was a prosecutor's 80% conviction rate actually pretty crappy because he handled nothing but slam-dunk cases? The difficulty of his cases can't be measured. There is no difficulty grade for a legal case like there is for an ice skating routine.
Sure, there are exceptions: call center employees and data entry clerks come to mind. The point is that we don't limit performance evaluations and performance-based pay and promotion to jobs where performance can be measured. Instead, we fall back on human judgment, justified -- but not fettered -- by accepted rules of thumb about how a person should do their job. Do you really think I need to spell out what those are for teachers? There's an entire college curriculum for teacher training. I don't think they spending four years telling them to "just go do what feels right." Even today, teachers' classroom performance is evaluated before they're certified. There's a rubric. And in the end the rubric doesn't matter anyway, because it boils down to personal opinion of the evaluator. Big bureaucratic companies, including mine, have rubrics for performance evaluations where workers are given numerical ratings in various categories, but that isn't measurement. It's just a way for a manager to record his opinion about whether an employee is doing a good job or not, using numbers instead of words. No boss sits down to write a glowing review, fills out the ratings and adds them up, and says, "Oh my goodness! I thought Jason was doing a great job, but the numbers say he's only average! I must have been wrong." If the numbers don't reflect his intuitive sense of whether you're doing a good job or not, he'll fix them.
Heck, even my subsidiary's baseline bonus for last year, which is calculated according to a precise mathematical formula, got a hefty boost because the VP over our subsidiary made a convincing case to the top executives that the numbers didn't accurately reflect our contribution to the company's bottom line. A few numbers got adjusted here and there, the formula was recalculated using the new numbers, and our bonus went up. Human judgment for the win.
You've avoided the question, and assumed I'm arguing against measuring performance. Basically, wrong on both accounts.
I'm not arguing that teachers shouldn't have some form of review. However, considering the nature of a public school system, knowing what we know, how should teachers be reviewed? The goal is to have good teachers teaching. Simply saying "we need to review teachers" doesn't help matters, as the question isn't whether we need to. It's the method that is the question.