Willingham takes a look at the issue of dismissing low-performing teachers and urges teachers’ unions to take a firmer hand. He notes that natural tendency to focus on the negative outliers at the expense of the positive stories hurts the unions in the court of public opinion.
True enough, but there are a two problems with his analysis. First, right now the teachers’ unions are in a purgatory of their own creation. They don’t want to use data to evaluate teachers and they don’t want to use managerial discretion. I guess that leaves the Magic 8-Ball? In practice, a combination of data and managerial discretion is how most professional fields operate and the most promising avenue for education. But the unions structural inability (they are not designed to lead on issues like this but to protect) to embrace really meaningful reform here leaves them unable to truly propose breakthrough ideas. The most far-reaching idea they have, peer review, is a great start and helps address the problem of observably bad teachers but does not get the field anywhere near where it needs to be in terms of performance-oriented management and growth of human capital. The numbers in the places where it has been tried speak to that. This gets at the larger tension between industrial-style unionization and how to organize a profession. Forward looking teachers’ union leaders are trying to sort that out.
Second, he states that,
“The issue of firing teachers has been poorly framed. It’s usually described as an issue of getting poor teachers out of the classroom in order to improve overall quality of instruction. That’s important, of course. But how big a difference is this really going to make to American education as a whole? If you had a perfect diagnostic to evaluate teachers, how many would you dismiss tomorrow? One percent? As many as three percent? If you had a perfect diagnostic and dismissed incompetent teachers the students in their classrooms (and their colleagues down the hall) would be glad. But the impact on the overall national quality of instruction would be minimal. Such dismissals could, however, make a dramatic difference in the public’s perception of the profession.”
I’m not sure what Dan’s basing this on? That figure would certainly vary by locale. But teachers themselves and school administrators in lower-performing school districts say it’s higher, in many cases substantially so. One has to be wary of the downstream effects of any teacher policies in terms of impact on recruitment and so forth, but if schools were to address some significant percentage of low-performing teachers, say 5, 10, or even 15 percent, the impact on student learning could be quite powerful – especially if schools and the work of teachers was organized differently and more professionally. The field has always taken a quantity approach to teachers rather than a quality approach. The evidence on class size reduction, teacher effects, and other alleged predictors of performance suggest that this approach may be profoundly misguided.
I’m not arguing here for some mass termination policy, only that we shouldn’t casually dismiss the potential impact of such an approach. Getting highly ineffective teachers out from in front of kids is no small thing in terms of outcomes.
In related news, Charlotte, N.C. says performance will matter in any layoffs.
I agree with your first point: the unions (and teachers) are in a situation of their own making. . .that’s really my point as well. Naturally they don’t want to use data or managerial discretion to evaluate teachers, because all of their emphasis is on keeping teachers in their jobs not on evaluating them. And so the response has always been “that method has lots of flaws.” My point is that stonewalling and calling methods flawed is costing teachers more than making use of the methods, even if they are flawed. Even more, teachers need to own this problem.
On the second point: I thought you were arguing for a mass termination policy until your last sentence. If you could identify the worst 15% of teachers their replacements might well be better (although I think both of us agree that that action will not be taken). But I do worry about the downstream effects. Job security is a reward that teachers see as compensation for low pay, lack of prestige, and difficult working conditions. So if the plan is to take that away, I’d say we’d better brace ourselves for the downstream hiring effects.
re: quantity vs. quality. I agree, but first we have to come to better agreement on what constitutes teacher quality, how to predict it, and how to measure it validly before entry to the field. I’m don’t mean to be obnoxious–I really mean that entry to the field is understudied, so far as I know.
Do we ever notice how much education reform in urban areas gets so much attention while reform in suburban or rural schools gets second billing? Hard-nosed reformers in inner city school systems are like rock stars in the education policy realm. Their job is incredibly difficult and necessary, I do not begrudge them of that. Nevertheless, there is no way much of the reforms going on in urban schools would work in many suburban areas, let alone rural schools. I think what tends to happen is the latest, hottest, most innovative ideas comes out of urban reform discourse and people want to simply map these reforms onto other school systems. I think this reaffirms the one size fits all approach rather than challenging it.
This all comes from a former teacher in an urban school who is now working at the teacher education level in a rural environment. Things like school choice, vouchers, and the wanton firing of teachers just would not work out here. As for choice, where do we expect these kids to go? No one like KIPP or TEP or any other flash in the pan ideas are going to leave the cities to open up charters in the Midwest. As for firing teachers, who will replace them? Teach for America graduates do not come out here because the rural Midwest oftentimes does not have that socio-political cachet that looks so great on a medical or law school application.
I do not think the analytical eye is really on the ball here when it comes to many of the reforms suggested here.
Dan,
When the AFT American Educator published your work, don’t you bother to read the rest of the journal? How can you miss the AFT’s efforts to push the Toledo Plan? Results have been consistent over two decades with around 8% of new teachers being removed every year. In my experience, the problem has been the inability of administrators to share control that prevents agreement.
As for past practices by unions, I also see that differently. Again in my experience, unions seek to solve the worst but the most solvable problems. If an incompetent teacher – or administrator – is also protected by federal law, for instance, they just transfer them to a spot where less harm will be done. The problem comes when the “dance of the lemons” gets out of control as it does in the toughest schools. But if you think its hard to staff high poverty schools today, just think of how much worse it would be if incomepent value-added models were implemented in a primitive way.
As Randi Weingarten told the NYT, it often takes only five minutes to identify a failing teacher, and the same often applies to identifying failing schools. Perhaps its confirmation bias, but I continually hear people from all types of urban districts that 1/3rd of principals are incompetent. I don’t think the problem is nearly that large. I just think principals seem incompetent when they have no choice but to play their role in an incompetent system and keep their true feelings to themselves. Regardless, I wouldn’t trust principals judgements if they weren’t SUPPORTED by evidence. Under the Toledo Plan pushed by the AFT, mentor teachers tend to be much tougher on weak teachers than principals. But again, we seek supporting evidence.
I don’t think its hard to sort out a better system. We need data-informed, but not data-driven accountability. We need something much much more efficient than the legal system, but we should borrow its basic model. We negotiate process for terminating (or sanctioning or rewarding) teachers yet we also need a type of bill of rights. For instance, “data mining” would be banned. The “first cut” in identifying ineffective teachers should be based on observations of behavior, with evidence complementing or supplementing or disproving the judgments.
Which gets me back to the way unions and schools got painted into this corner. Often, the union wants to see a bad teacher fired. We have legal obligations to defend members, even the guilty, but we don’t want to win. And when due process is violated we have no choice but to hold our noses and object to the misbehavior by the administrator in question. Think about it for a second and you’ll probably conclude that both of you would do the same. We can’t accept an awful precedent, for instance, even though we are tempted.
This reminds me of the last time I was on jury duty. We all thought the accused was guilty. I was the last holdout. The second-to-last holdout kept cross-examining me based on the judge’s instructions until long after I knew I would have to vote to acquit a person who I knew was guilty. Afterwards when we were allowed to ask questions, both attorneys and the judge were surprised by the verdict. They knew something we didn’t know, however. The prosecuter had just picked up the case and quickly read it as he went from one trial to another. Schools are like that, in that we are all so overwhelmed that we can’t even win the easiest cases.
John, your post vividly describes the problems posed by the situation. No argument that the situation is fraught with difficulties. My point was that teachers (through unions) can exercise choice about the type of mistakes that are made. I argue that they are currently making an ineffective choice.
The analogy you draw between teacher evaluation and jury duty misses a key distinction: as a juror you have no say in the writing of the laws.
Dan,
Good point. But I was actually refering to the dynamic of two cultures separated by a common language. (And all this fight is over the distinction between data-informed vs. data-driven evaluations and the mixed messages that are being sent by “reformers.”) Our job as jurors allowed us unlimited time to parse the judge’s words. The legal professionals lived in the world described by Tom Wolfe where the “grub” is pushed non-stop through the overworked system. I wouldn’t repeal the Bill of Rights, but I would want a system that’s not so overwhelmed. And since they were all getting paid $100+ per hour, they were relatively more competent than our equally overwhelmed schools.
Teacher identifiers, without strong protections, are a loaded gun. Getting back to your point about drafting laws, we are willing to negotiate but when the true believers won’t consider “trigger locks” it pushes us into an impossible situation. When it comes to writing laws, union leadership has to bring along the members. No profession would allow a nationwide data warehouse, to borrow Diane Ravitch’s characteriztion, without protections. We don’t even want such a system for fighting foreign terrorists without protections.
Fortunately, we still have due process laws that would enventually kill data-DRIVEN accountability, but it would do so at the costs of millions in lawyer fees. But there’s another cost that you are especially qualified to expand on. If the DFER plan became law, think of the increase in strokes, heart attacks, and other stress related diseases among teachers. My understanding is that the worst stress in terms of health is when the person has little or no control over their environment.
I’m still confused by the “ineffective choice” that unions are making. I’ve seen national and local offices investing a great deal of money and political capital in advocating the Toledo Plan. I’m not privy to the inner circle that’s lobbying Duncan, but in my experience union leaders are sincere in pushing for a more rational system. Do you have a different experience?
And one last point on why the legal system was overwhelmed. Data-driven tactics in the war on drugs swamped the system filling prisons with non-violent criminals, removing huge numbers of parents from poor communities, and increased racial conflict and gang ethos, and all of which further damaged our schools. As the originator of data driven law enforcement said, data-driven accountability is like chemo-therapy. Keep it up too long and it becomes poison. I was writing a book on the subject in the late 80s, and virtually everyone in the legal system knew that Congress had unknowlingly set us on a path where institutional pressures would fill the prisons with Black and Brown people, but the outside reformers didn’t know what they didn’t know about the Street. The same applies today, I’d argue, about education “reformers” who know more about math than the realities inside schools.
You write, “they don’t want to use data to evaluate teachers,” a statement that is categorically false.
They most certainly DO want to use data. They just don’t agree with you about what data to use. They think that classroom observations provide the best data. They think that scores on standardized tests do not provide a good measure of teacher quality.
Obviously, each form of data has its problems. I would suggest, however, that the former is more flexible, and the latter is almost entirely out of the hands of the decision-maker in this case. That is, the principal or other supervisor does not control the test, so who is to say that its goal align with his/hers? On the other hand, classroom oberservation provides multiple opportunities to gather data about teacher performance, and to address improving it far before the standardized test.
This should be a basic lesson of schooling, understanding the nature of evidence (data), and the idea that there are different forms of evidence that might be more or less useful, more or less acceptable to different people.
Ceolaf,
You are right. I didn’t tackle that issue because the comment was already too long.
But you know what’s weird? Why do we need a reminder of the definition of data? What you wrote should just be assumed.
I definitely appreciate your clarification of my oversight.
They just don’t agree with you about what data to use. They think that classroom observations provide the best data.
So the question is, why do they think that? Do they have evidence to refute Medley & Coker’s (1987) famous warning that classroom observations by principals “are only slightly more accurate than they would be if they were based on pure chance,” or Kenneth Peterson’s claim that “seventy years of research on principal ratings of teachers shows that they do not work well”?
John said “in my experience union leaders are sincere in pushing for a more rational system.” That has been my personal experience as well, although I readily admit that my experience is limited and it’s not my impression that our opinion is shared by the public. And yeah, I see a big difference between the AFT and NEA on this issue.
I really didn’t mean for my post to be taken as a criticism of union action on this matter. The editor at Britannica blog initially added this bit after the colon to my title “How teachers can get more respect, part 1: the role and responsibility of the unions.” (that’s still the file name at Britannica.) I removed the after-the-colon part because I don’t see it as the responsibility of the unions. I think it’s up to teachers.
The conversation here (and elsewhere) is focusing on the difficulty of making accurate decisions, whether test-driven (which I have said elsewhere is a terrible idea) or made by administrators (which I see as a somewhat less terrible idea.) All are valid points, although if you set as your goal firing really egregious teachers, I would guess that validity improves. But I return to my original point, which was addressed to teachers: something is in the air. The President is talking about getting rid of poor teachers. It appears likely that something is going to be done, so you may as well try to take control of the situation so it’s something you are doing, rather than something that is done to you.
It’s amazing how many false choices are being offered here. Why not use data to decide which teacher needs professional development in a particular area? That’s what you want to do with students. Evaluating teachers with data could be a good thing if the assessment was fair and had some flexibility. I haven’t seen ANY good evaluative tools yet. One should also notice that the only data systems so far have been created for Math and English teachers, and the methodology so far is pathetic. How will you design merit pay for Phys Ed teachers, or Visual Arts Teachers? Ultimately, the unions are there to protect the basic rights of all teachers against authoritarians who make decisions based what is easiest for the authoritarians, not on what is right. If decisions account for basic employee rights, it would be politically undesirable for the unions to fight them. It is possible to fire teachers who aren’t worthy of tenure, and avoid the complications of lengthy legal battles once an incompetent administrator has given tenure to a poor teacher.
Great discussion – but as Loren notes the debate swirls mostly around false dichotomies and either-or thinking. Unions are vilified when they raise legitimate questions regarding the use of current standardized tests to judge and pay teachers. The reformers are attacked for trying to break-up the single-salary schedule that stifles innovation and creativity. Incompetent teachers are unfortunately protected by their unions -leaving students bereft of the education they richly deserve. But incompetent teachers usually remain in teaching because the administrators who are evaluating them are more incompetent than they are. As John noted when teacher unions seek to take responsibility for their own (through peer review) they are often rebuffed by administrative elites who want to remain in control.
So are we trapped in a not-so-civil policy war, with no possibility of reconciliation? As Clay Rise recently noted, “….education policy isn’t the problem, its education politics that needs to change.” Perhaps, we need to look hard at ourselves and question, “Do we really want a teaching profession in this country led by our best teachers – not administrators or policy wonks from inside-the-Beltway.
An alternative view and vision — e.g., Teacher Leaders Network — could plow “and/both” thinking and a 3rd way to doing the business of meaningful reform that honors the profession that makes all other possible. Are you ready? Can you stop the blame game?
I found the article entitled “If it bleeds it leads….And, Quantity v. Quality. I chose this article because it reflects some of the views that I hold from my teacher’s union experiences.
Willingham states that the union is created to protect it’s members not evaluate their performance. My experience has been the same. I worked with many teacher’s that had tenure and felt very comfortable in their position. Classroom motivation and accountability for their students was minimal. It was extremely frustrating to work with teachers who were beyond burned out on a daily basis. Some mentioned that the money and benefits were to good to lose. I find that appalling! There are many new teachers seeking their first chance at a classroom full of kids but no openings were available. My district also had an issue with a teacher not testing the special needs children, which is illegal, and received minor chastising.
The union has many benefits but just as many drawbacks. It is great to be protected but tenure can cause some to feel too comfortable in their position.