Join the conversation

Good post, but we have to consider another important point. As long as there are firewalls between teachers and their students' performance, how are we going to tell what kinds of teachers do best in different environments? How do we gain the knowledge we need to make the best decisions about teacher deployment?

Good post. My experience in teaching was that I had no trouble serving either middle class or low-income students; the differential came with the behavior type of the students (roughly speaking). I excelled at drawing out withdrawn children, and was terrible at responding appropriately to in-your-face children. I may have even unwittingly caused in-your-face behavior, because some students could intuit that I wasn't good at handling disruption. Of course, over the years, the ratio between shy children and acting-out children changed, and I felt that I didn't bring what was needed to the classroom and left teaching. The only other alternative would have been to seek out a position in an all middle-class school, which I didn't want to do.

Claus, you're on the right track here. Reform efforts should focus on creating systems and work conditions that will help any school to grow their own teaching force and improve learning for all. Consider as an analogy the sports coaches who win league championships with one team and then move to the next job and struggle.

The reply from "anonymous" suggests that the answers to teacher effectiveness are behind the firewall. In fact, we need stronger, ongoing, robust and interactive evaluation practices that consistently lead to improved teaching for every teacher - and yield plenty of varied types of evidence of what's working. Standardized test scores might give some indication if a school is moving in the right direction, but most tests are ineffective tools for evaluating individual teachers, a fact that needs to be continually repeated by the research and testing experts, journalists, and educators who are aware of that problem. If we take down the firewalls in question, we run two risks. One is that teachers will modify their practices in ways that boost test scores but could have unintended consequences in diminishing the overall educational experiences of the students. The second risk is that teachers will be subject to improper evaluation practices. See my blog for examples of business experts who have warned against over-simplified quantifiable types of evaluation of individuals in complex systems.

The reason great teachers end up disproportionately in less needy schools is primarily a result of the step and level compensation system.

Markets will always find a way to clear and the teacher job market is no exception. If you don't pay people more for working in hard to fill/high needs areas, they will choose to work in relatively easier positions if the money is the same.

If we changed our compensation systems to take into account that its more difficult and time consuming to work in tougher schools you would mitigate this problem.

The step and level pay system isn't rational from a labor economics perspective. Because we pay everyone the same, the market clears on where the jobs are easier instead of where compensation is greater.

This is the same reason we can't keep people in special education for more than a few years, and the same reason you have 3000 applications for every PE and social studies job.

A smart policy decision would be to incentivize getting great teachers (and great teaching) into relatively harder situations.

Jason Glass
Eagle, CO

Look at it this way: what can we do to get our best doctors into the emergency room of a public hospital?

Tom makes a great point on several levels. An inner city teacher has to develop a mentality that is comparable to that of an ER doctor.

And nobody would disrespect doctors enough to coerce a cardiologist with high survival rates into an ER or any other medical field.

Face it, selective schools and low-poverty schools are worlds apart from inner city neighborhood schools. Yeah, after a few years many inner city teachers transfer to more tranquil schools. That's just common sense.

Jason, your remark about less challenging schools reminds me of an exchange I once had with Jay Mathews. I had been offered a part-time job (which I could not afford) by H B Woodlawn alternative program in Arlington VA. Jay told me the kids there were bright and self-motivated and did not need me as much. My response was even bright and self-motivated students need teachers who can challenge them.

The problem with the firewall image is the presumption that if you take it down it actually informs you of something meaningful about what the teacher is doing. The kinds of tests we have now it does not, because you have no indication of the prior knowledge or outside learning that influences the results of the test scores. Even so-called value-added assessment is really not robust enough to meaningfully tie back to individual teachers. William Sanders, who was responsible for what was called TVAAS in TN used to argue that one could only use teacher results for comparisons against other teachers in the same school.

If we are attempting to use tests as the sole measure of teacher quality, we are nuts.

The proper use of standardized tests would be to have them shorter, and more frequent, and turned around more quickly, to give information that is useful. The way we largely use them now, as culminating measures, uses them for punitive summative purposes and does NOTHING to inform or improve instruction.

To evaluate how teachers are doing requires far more than test scores. It should include observations, inspections, students demonstrating the ability to apply knowledge - thus science fairs, national history day and the like are far better measures of student knowledge and understanding of a domain than are tests consisting solely of selected response items. Tests which require students to EXPLAIN their choices/reasoning are superior to picking one out of four/five provided choices, which of course is very far from real world application of knowledge, which really should be the goal of education, that what we learn we can apply in the real world.

As far as the question Claus raises, teaching occurs in context and involves relationships. It is one reason I call all my parents at the start of the year, to build relationships and to accelerate the process of understanding each of my students. If one as a teacher is going to help students connect with what we want them to learn, we do have to understand what they bring to the classroom, which includes the culture in which they live, their prior knowledge.

I am considered an excellent teacher. I am not as effective for special ed kids who need a more formal structure than I prefer to us. I prefer something more fluid precisely because I am trying to stretch my kids, the majority of whom spend too much time without taking intellectual risk, pushing the envelope. I have taught multi-level classes, and was able to adjust, but it is not my metier.

Similarly, I chose not to teach in a school with a high percentage of parents who spoke Spanish and not English because I do not speak Spanish, and that would limit my ability to work with the parents collaboratively. I could have learned, but it would have taken time before i could have been as effective. Similarly, I know far more about black culture than I do the various flavors of Hispanic culture (although I continue to learn) which would have limited my ability to connect the content I teach with the lives of my students.

Just a few thoughts from someone who has just completed his 15th year of secondary teaching, 11 of which are in a very diverse high school with a wide range of academic abilities and preparation. Someone who also thinks and writes about education on a regular basis, including previously at Claus's request here.

Peace.

Great points, Claus. One thing your article doesn't mention is that frequently good teachers are driven out of low-performing schools. Such teachers bring expectations of autonomy and content expertise that influence their teaching moves not to dumb down the curriculum but to find authentic ways to make the curriculum relevant and meaningful. And they are going to find ways to do this wherever they teach. Many administrators in low performing schools do not select for this quality but rather blind compliance to narrow, easily identified approaches to teaching. There seems to be a colonized mindset from above that selects for these administrators.

Pardon my ignorance, but where are urban and suburban teachers paid the same? My salary as a rookie in New Orleans was substantially higher than what my mom made teaching in a rural district back in Ohio. Not a 50,000 difference or anything, but there was a not so insignificant difference. I always assumed it was some sort of "combat pay" bonus, but I guess I was mistaken?

It seems pretty clear to me that high-need areas ought to have a different salary scale than less-needy areas, but I certainly agree with Claus that working conditions likely play an even greater role. If my district had offered me an substantial raise (say, 7,000 dollars) to stay, but with the same working conditions I had the year before...I still would have quit.

Jason--I think Ken's and Mr. Brown's perspectives from the classroom are informative. Currently, pay isn't the biggest sorting mechanism out of hard-to-staff schools. And teachers make big decisions about where they want to teach--decisions that higher pay won't necessarily affect.

Now, if pay in hard-to-staff schools were $50K higher, that might indeed make a bigger difference. But don't think unions would be your biggest hurdle. If you started using big cuts in the salaries of suburban teachers to subsidize big increases in salaries of urban teachers, influential suburban parents would have your head. We've already seen what happens to superintendents who preside over much smaller redistribution of resources within districts. (See, for example, the drama in Hamilton County, TN and Montgomery County, MD.)

If you wanted to redistribute that wealth among districts, you'd have to change how we finance schools by moving away from the local property tax model. That would be difficult, to say the least. (Many have wanted to change that model for ages.)

Ed--I agree that great teachers can be pushed out of struggling schools, though I'd go a bit easier on the administrators in those schools. If some of them are less willing to embrace very innovative teachers, it's because it's natural for people who feel they're on the hook to become risk-averse. It seems clear to me that efforts to teach a rich and broad curriculum, done well, will lead to increases in state test scores. But I don't have to stake my job on it.

That said, there are great examples of leaders in hard-to-staff school who have taken the right risks and welcomed innovation in their schools. We tell their stories in our "Success Stories" pages.

Most of you are assuming that test scores measure student achievement when we have mounds of research going back to Coleman that tests, by their very nature, predominantly measure wealth. So if you put all the poor children in one school and the more affluent children in another school, those teaching the more affluent children will have higher test scores, and it doesn't matter a whole lot what they do.

Daniel Koretz, the Harvard expert on testing, has a current book out on this. Steven Dunbar another expert has talked about this. Bob Linn, perhaps the top expert, has pointed this out. James Popham, another top testing expert has made the point. Anybody with expertise in testing says poor kids in general will be outscored by more affluent kids. Coleman even pointed out that if you put the poor kids in schools with more affluent kids the poor kids do better, which was the rationale for busing.

Teachers faced with scientific facts act rationally: they want to be successful so they go where the success is. Administrators are the same: they would rather not deal with the discipline problems created by lead poisoning and transient students. You can improve the outcomes of many schools by bringing in good administrative resources and forging collaborative processes among the teachers, but the more affluent schools are still going to score higher on the tests because that is what tests measure.

California is a cautionary tale for those looking for an alternative to local property tax funding. With some exceptions due to grandfathering, there aren't huge disparities in funding from district to district, but the trend has been toward the lowest common denominator.

Policy makers say they care about improving urban schools, but if doing that requires raising teacher salaries to attract good teachers, I think many taxpayers will balk.