Schools reward their inhabitants in two amazingly different ways: teachers get salaries, and students get grades. Not only are these systems vastly different, but I'm not sure teachers would accept a pay scale that's built like a grading scale, nor would students accept a grading scale that's built like a teacher's pay scale.
The vast majority of schools pay teachers according to a "salary schedule," a rigid, two-dimensional matrix of dollar amounts with credits across the top and experience down the left side. The more credit hours you've earned, and the more years you've taught, then the more you'll make. I don't feel it's a fair system, but it's a system that most teachers will agree to use, putting it ahead of most any other pay system available. It's a simple system, perhaps too simple. Its simplicity allows us to easily print and read the salaries on a sheet of paper, and I think that's one reason we continue to rely on it. Now, forget merit pay for a moment. Suppose we simply wanted to add a third variable to the schedule, such as student count. The number of students you have definitely has a measurable impact in the effort it takes to be an effective teacher. (Certainly an impact comparable to credit hours, for example.) Addition of a third variable would turn the salary schedule into a three-dimensional model, not easily displayed on a piece of paper, and probably requiring the use of an algebraic formula to calculate every teacher's salary. So however incomplete and unfair, teachers opt for the simple and straightforward.
If you put ten teachers in a room and asked each to describe their grading system and practices, I guarantee you'll get ten different descriptions. My goal is to have a grading system that accurately reflects each student's ability and achievement, and I fail at that every single grading period. I always seem to find at least one student for which the numbers just don't reflect my personal feeling of what he/she has learned. As a math teacher, I think I'm especially critical of my grading methods, and thus I've never graded exactly the same two semesters in a row. Teachers are allowed so many choices: grade weighting, extra credit, curving scores (using normal curve or other methods), dropping lowest scores, partial credit, use of "pluses" and "minuses"...the list is considerable. These variables are in addition to the simple idea of being a "tough grader" vs. an "easy grader." Is a "C" average? Should every class have "A" students? How many failing students is too many? Teachers are generally allowed to include any combination of variables, and we expect our students to adapt to each of their teacher's grading systems.
As a teacher, I find it ironic that we resist making our compensation system more complex, yet we subject our students to some of the most esoteric grading rules imaginable. Maybe I'm comparing apples to oranges, but maybe not. If my ideas are valid and the writing of sufficient quality, feel free to compensate me for my efforts. Hmm...should I be simply paid by the word? Or should we develop a multi-variate rubric to assess the quality and effectiveness of the piece across a variety of audiences depending on my grammar, word choice, sentence structure, paragraph organization, and tone?