A recent PPIC survey on education shows 69% of Californians believe student achievement should be closely tied with teachers’ salaries. This shows a greater concern about teacher quality, likely brought on by recent attacks on teachers unions. This avenue of discourse often blames teachers more than praises the tough job they face in real classrooms.

I am a teacher in a low-income urban high school, and while I do agree that student achievement should be partially tied to teachers’ salaries, it needs to be one of many measures we use to reward good classroom instruction. Unfortunately, I’m not sure anyone out there is doing this right.

The first important point we need to keep in mind is there still isn’t a proven way to effectively evaluate teachers. The Bill and Melinda Gates Foundation and other organizations claim to be perfecting it, and there are already programs out there that compensate teachers with bonuses when they raise scores. I wouldn’t mind making some extra money when my students’ scores rise, who wouldn’t? But we have to remember our assessments of both students and teachers are far from perfect.

I give my students many tests. The unfortunate thing about our state and district-wide assessments is they are all multiple choice. In addition we often find errors in the tests, or the data from the tests is inconclusive. Some of the tests are culturally biased as well. These tests are so flawed it is sometimes impossible to judge anything from the data gathered. I just did an assessment of data over the whole year for one of my classes, and it didn’t show me much. Overall my class showed growth, but the assessments were so up and down it makes me think that I could have given them any set of tests on any subject and it would have showed the same thing—that is—over the course of the year they got a little bit better at taking a certain kind of test.

Let’s take a subject specific perspective. Part of the problem for English teachers like myself is how they assess Language Arts skills in a multiple-choice test. The actual writing part is either very small or non-existent. So even though my students took seven persuasive essays through three drafts and turned them in as a portfolio, they weren’t asked to write anything when it came to state-wide test time. As an English teacher, I would love to show you how much writing my students have done, but all the tests they take for English don’t require any writing. To me that is unacceptable, and it is unacceptable for anyone to tell me I am or I am not doing my job if you haven’t seen my students’ writing.

Then there are a variety of nuances on campus that make assessing a teacher’s influence very, very difficult. The sheer variety of classes we teach seem to make a huge difference in test score trends. For example, in my English department we have three levels of English Language Learner classes, three levels of sheltered classes, regular English, plus accelerated, honors and AP. Not to mention special programs like Puente and AVID. I’m not aware of any distinctions being made between these types of classes when it comes to data assessment.

Then there is the difference in rigor on the grade level exams themselves. Looking at CST data of these classes, many things become apparent. 9^th graders seem to make the most gains out of any grade level. Whether the 9^th grade STAR test is easier, or developmentally students make a huge leap between junior high and high school, it seems like 9^th grade teachers would be compensated the most simply because they teach 9^th grade.

In addition to the problems we face in the classroom, studies have shown that putting the carrot of better pay in front of teachers doesn’t necessarily make them teach any better. In fact, I have written about this topic for New American Media before. In a nutshell, good teachers can’t do much better than they’re already doing. They’re already arriving early, staying late, grading papers all weekend, running after school tutoring programs. If you told a teacher like this they’d make more if they did more, they’d rightly ask just how do you expect them to do more. Even so, I am in favor of giving high achieving teachers more money, because even if they aren’t going to start working harder all of a sudden, they can still be compensated for the great job they’re already doing.

This is why if we tie teacher pay to student achievement, it should be as a bonus, but definitely not a determination of what salary we get. If you end up with a class of 30 boys, you should still be guaranteed a base salary (we should also take into consideration you have 30 boys in a classroom, and no evaluation system out there is doing that as far as I know). We should be tying student achievement to bonus pay, not salaries. This is an important distinction, and there are already some programs out there trying this. The TAP (Teacher Advancement Program) has begun a trial in the Lucia Mar District here in California, and while they do tie teacher pay to student performance, it is only in the form of a bonus. Not penalizing a teacher’s already meager salary is the only way I could advocate for such programs as TAP.

Here’s the good news. Despite my reservations, there are some interesting points to be made in favor of tying student achievement to teacher pay. First, good teachers would be compensated with more money, and I am definitely in favor of that. And as far as I can see, the data, while flawed, over time will still distinguish teachers whose students consistently outpace the school and district averages.

However, test scores should not be the only measure we analyze to properly compensate great teaching. We need to do a variety of things. We need authentic evaluations by administrators and other teachers. We need to chart the growth of students over four years at least—the two years before they came to you, the year you had them, and the year after. We need to improve the assessments we are giving our students, and include more authentic criteria. For example, to see how well they are doing in English class, they should have to write something—but this requires extra time and extra money, neither of which education has in this climate. We also need to be able to distinguish the difference in expected growth in each type of classroom. A 9^th grade accelerated class is going to show more growth than a sheltered 10^th grade class. Teachers already would rather teach the accelerated classes, and if it is more lucrative when it comes time to get in the money, we’ll have even less teachers willing to take on the difficult Sheltered students.

This is a very sticky point in this debate that argues a more competitive environment will turn teachers against each other. I agree with that, and the only way around it is to weigh each student and each type of class within each grade level correctly. If you teach a sheltered class, your test scores should only be weighed against other sheltered classes of the same grade level, and even then, we need to take truancy into account, gender distribution, resources available, and a lot of other things I guarantee no one is looking at right now. If we don’t get this specific, teachers will bicker. I have one 9^th grade class of geniuses, and another who have a hard time spelling their names right. What if I got two of the first and another teacher got two of the second? The other teacher would not be happy, and would not get the same bonus I would get simply because I got the well-behaved classes with more girls.

I have no problem being evaluated and held accountable for the job I’m doing. My door is always open, and I welcome discussions that will hold our profession accountable for teaching the next generation. With the negative perceptions we face, opening the realities of what we do every day for all to see is a wonderful idea. If we want to tie student achievement to teacher pay, lets just make sure it isn’t the only assessment we use to determine the job teachers are doing. My ninth graders just put on the entire play of Romeo and Juliet, memorizing lines of Shakespeare and delivering stunning soliloquies, and I have yet to come across an assessment that tests that—except of course for my own.

So even though it is most likely not being done right, should we still offer bonuses to teachers whose students are most prepared to ace multiple-choice tests? If it means more money for teachers, I have to say yes, and hope we get better at assessing teachers and students in the meantime.

A version of this article appears in New America Media, the nation’s largest ethnic news organization.