JSD: In 2007, you wrote in JSD that “educators should begin to act more like professionals.” What do you mean by that?

Elmore: My basic belief is that education is a profession without a practice. One of the reasons why we have relatively little influence over the terms and conditions of our practice is because we as practitioners haven’t developed a clear sense of what it is we do, and what its connection is to the core functions of our organizations. We don’t select and promote people based on their knowledge of how to do this work. If we meet those conditions, we’ll have a lot more influence over how the world affects us.

We’re at a point now where it’s no longer acceptable for superintendents and principals and teacher leaders to say that teaching is a mysterious thing, that it occurs idiosyncratically in every classroom and you have to take what you get. We need a much more systematic answer to that question of how we do what we do. Part of our responsibility as leaders in this sector is to create conditions so that teachers can be effective individually and collectively.

We’ve paid a fairly high price for being casual about how to organize for high-quality instruction. Americans in general have what I call attribute theories of leadership, which is you put the right person in the right place at the right time and a miracle happens. This is not a workable model if you’re interested in doing anything at scale. If you’re interested in doing things at scale, you have to focus on the knowledge and skills people have and their capacity to use them, not just their personal characteristics and attributes.

Our work with administrators suggests that most of what you learn about how to do this work you don’t learn in college, you learn in practice. It should be the obligation of the profession to create the learning environments in the practice world that make it possible for people to get systematically better at their work. The profession needs to take some responsibility for developing the knowledge and for making sure that people don’t get into positions of leadership at the classroom level or school level if they don’t have the knowledge they need to have.

THE CORE OF EDUCATIONAL PRACTICE

JSD: In your work, you’ve examined closely the relationship between changes in structures in districts and schools and changes in the teaching and learning practices in schools, or what you call the “core of educational practice.”

Elmore: The core is the teacher and the student in the presence of content. This definition grows out of a model that David Cohen and his group developed around capacity, drawing on the work of David Hawkins. The model says that there are only three ways to improve the quality of instruction and the quality of student learning: You can raise the level of the content, you can increase the knowledge and skill of the teacher, or you can change the role of the student in the instructional process. If you do any one of those things, then you have to do something about the other two pieces.

If you have that definition of the instructional core, the organization is only going to improve instruction if it supports the development of knowledge and skills on the part of teachers and students and if you can get knowledge and skills into the setting in a way that effectively connects to practice.

I believe the effect of professional development on the quality of instruction and student performance is inverse to the square of its distance from the classroom. That is, most professional development that occurs outside the school setting doesn’t have much of an impact, although it has other purposes. Most professional development that occurs inside the school but outside the classroom has the potential for affecting what goes on inside classrooms, but it doesn’t have a direct connection unless teachers are actually in each others’ classrooms.

Professional development that is likely to have the biggest impact has a reciprocal relationship between the time you spend with your colleagues in classrooms trying to solve instructional problems and then reflective time outside of classrooms to think about what you’re going to try next. The corollary to that is the most powerful professional development occurs in real time around real problems in real schools involving real people who actually have to make decisions about what to do on a day-to-day basis.

That implies a very different kind of school organization. You have to have the structure that allows people to get together. You also have to have a culture that begins to break into this isolation of teaching, to make it OK for people to be in each others’ business and to talk about their common concerns about the work, as opposed to doing only what they do in classrooms by themselves.

JSD: What kind of leadership skills does this require?

Elmore: If you’re a building administrator, you need expertise about how to organize the schedule, about how to use consultants and coaches. You have to have enough instructional expertise to know the difference between a classroom in which the math curriculum is being taught at a high level versus a classroom in which it’s being taught at a low level. You have to know how to use time and money in a strategic way, and you have to have all the interpersonal skills necessary to coax people out of their private sphere of practice into a collective practice. You need a basic understanding of the cause-and-effect relationships we’re trying to see between teaching and learning and how to diagnose problems if it’s not working.

The real constraint to this kind of practice is not preservice training, it’s the capacity of people to create school systems that support this kind of work, because that’s where people are going to learn how to do this.

School boards have to begin to understand that for systems to get better, we have to spend more money on the acquisition of knowledge and skills for the practitioner. In order to make that argument credible, school administrators have to learn how to spend the money they have more powerfully and effectively.

A COMMUNITY OF PRACTITIONERS

JSD: You have been working with the Connecticut Superintendents Network since 2001 and then more recently began working with a school leadership network in Cambridge, Mass. How do these networks function?

Elmore: This project is an outgrowth of work that the Connecticut Center for School Change had been doing. The work is focused on developing a new model of the superintendency, one that is focused primarily on the improvement of instructional practice and student performance. We looked to the medical rounds model as an analogy to the kind of practice that we wanted to try to develop.

We’re developing a community of practitioners with a common interest in improving instruction who are willing to stay in a long-term professional relationship with each other and who are willing to take on this issue of developing a practice as colleagues. The emphasis is to have the superintendents take control of the process and use it to develop their own practice, and for this network to develop their community of practice and shared body of knowledge about how to do the work.

Three years ago, we started with 12 superintendents, and then we added 12 more superintendents two years ago as a second cohort. We get together with all 24 superintendents about three times a year. Each group meets for a full day every month. Every other month we meet in a school, where we work on what’s called a problem of practice in that school. This is an instructional issue that you can see in classes and that is related to a larger strategic problem that the district and the school are trying to solve. The issue could be the achievement gap between certain groups of kids, it could be the issue of literacy in the middle grades, these kinds of issues.

The group does a highly structured observation, working in rotations through classrooms in groups of three or four. Then we do a debriefing on the spot about what we’ve observed, which usually includes the principal of the school in addition to the superintendent for the district we’re visiting. Then a month later, we gather for a full-scale debriefing. The product of this debriefing is what we call the “next level of work.” The superintendent, having listened to her colleagues talk about what they saw and what their ideas are for the next level of work, makes a commitment to a certain course of action.

The last thing we do is a process by which the superintendents create what we call a theory of action, for them to state their approach to the superintendency in a cause-and-effect model. This helps them understand the relationship between the specific things they do on a daily basis and instruction in the classroom.

The network practice has percolated down into the district. Most of the districts are now running some form of this instructionally focused observation and problem solving. I suspect that if you walk into a random school in these districts, the principal would present you with his or her theory of action.

JSD: You must see the network model as a particularly effective way for leaders to develop knowledge about the field.

Elmore: Isolation has been the big issue in this sector. There’s isolation at the classroom level, there’s isolation at the school level, there is a lot of isolation among school systems. The network model puts people in proximity to each other around common problems and gets them to develop a collective practice and a body of knowledge about how to deal with these issues.

JSD: How do schools move down this road toward a collective practice?

Elmore: This is a process that happens in stages. Schools that are not very far along in this process are badly organized to do any kind of collective work instructionally, so teachers don’t have any time to meet with each other to talk about common instructional issues.

Schools have to go through a stage of building a structure inside the school to make it possible for people to talk about these problems collectively and to work on them. What we’ve discovered through our work is that you can have these structures — you can have common planning time by grade level or by content area, you can have instructional team leaders at grade level or content level, and all those things can be working in the sense that people are meeting and talking, and you can still not be producing anything.

The next stage of this process is to get people to be very specific about the kind of problems they’re working on individually and collectively and then to commit to a course of action. They need to go to their classrooms and try it; most importantly, they have to watch each other doing it and then come back and share what they think they’ve learned.

What we’ve found, not surprisingly, is that if you don’t focus heavily on the knowledge and skills that are required to do this work, you don’t get much impact. When we asked teachers who seem to have good teamwork skills where they developed those skills, to a person, none of them said that they developed them in schools. In order of importance, what the teachers told us was that they developed their team skills in church work, in political campaigns, and in union activities. So what that means is that the organization in which they work isn’t giving them the knowledge and experience that are required to be effective in a team. You have to put that knowledge and skill in the organization and help people learn how to do it.

TRANSPARENT CLASSROOMS

JSD: What do you mean by the “norm of transparency”?

Elmore: The norm of transparency is essential for really powerful improvement at the school level. It is virtually impossible for teachers to learn how to improve their practice if they can’t watch each other teach. We have plenty of evidence of organizations that have done everything to put teachers together to talk about practice, but they haven’t gone to the last stage of making it possible for teachers to watch each other practice and to work with each other in classrooms.

The knowledge that teachers bring about their own practice into these groups is not very accurate, unfortunately. You have to have people circulating through each other’s classrooms. You have to be able to ask your colleagues to watch you do something and then tell you what it looks like.

JSD: Have you found that teachers working in that sort of environment develop a more accurate sense of what’s actually happening in the classroom?

Elmore: It’s a pretty straightforward causal connection. You’ve developed much more specific language for talking about your practice and understandings of specific cause-andeffect relationships about what you’re trying to teach and what the evidence is that students are actually learning.

We’re much better at understanding these cause-and-effect relationships in the lower grades than we are in the middle and upper grades. That’s one reason why elementary school teachers are doing much better by international standards than middle grades teachers or high school teachers. Our performance in the middle grades on basic literacy and numeracy is terrible. If you’ve spent any time in middle grades and upper grades classrooms, you know exactly why, and it’s that the practice is, on average, terrible. Teachers are working in an environment in which the kind of specific knowledge that they need to respond to the particular issues students are bringing into the classroom just isn’t there, and the professional development isn’t there.

In Boston, we’ve got about 40% of the population who are in the lowest quartile of functional literacy skills required just to read and use the textbooks that they’re taking home every day. They don’t have the academic vocabulary needed to do the work. In the 7th and 8th grade, we make these distinctions among those kids who are capable of mastering higher-level academic content and those who aren’t. So we’re virtually guaranteeing that a large proportion of the population of these kids are not going to able to read and understand these textbooks that we’re asking them to use in the course of what we’re teaching.

JSD: How can the field turn this around?

Elmore: We need some equivalent to Reading Recovery for middlegrades kids — that is, a highly focused clinical intervention that increases kids’ academic vocabulary and fluency and comprehension at a markedly faster rate than what we’re currently doing in the classroom. Then we need much more focus on the part of subject matter teachers on what they know about the functional vocabulary and the processes that are necessary for kids to master the content. We have to get out of this business of “some kids can learn and some can’t” and into the business of saying, “What is it about our teaching that isn’t connecting with kids?”

We’re very soon going to be in crisis mode starting at about the 7th or 8th grade.

With NCLB, there are more schools coming into various stages of school improvement than anybody knows how to handle. Soon it’s going to be obvious that we’ve got a bigger problem out there than anybody can fix with the current state of knowledge, and just banging schools upside the head isn’t going to do anybody any good.

A DIFFERENT MODEL OF ACCOUNTABILITY

JSD: What impact are current accountability systems having on schools?

Elmore: The effect of accountability policy has been to further devalue professional knowledge of the field. The consequence of relying so heavily on testing and sanctions has been to reinforce the idea that educators already know all that we need to know to solve these problems, and the reasons why schools aren’t performing is that educators are just contrary and incompetent people.

What we’ve lost, what is clearly visible with some sense of urgency in other countries that are struggling with accountability, is that we’re engaged in an enormous human investment enterprise here. The gains we get in student learning and performance are going to come as a consequence of increased knowledge and skills on the part of adults. And that the only way it’s going to happen.

We, as educators, do not command the kind of respect and social authority that are necessary to make this case. The solution is partly a matter of developing a clinical practice and beginning to act like professionals and to get our own house in order. It’s partly a political issue — educators need to get better organized around these issues and to begin to push harder on them.

Our so-called allies in federal policy are not paying attention to the central problems of this sector, and they’re not paying attention because their failure to pay attention doesn’t carry any political cost.

The accountability system in NCLB only looks and sounds inevitable if you’ve never been to another country. There are models of accountability in Canada, Australia, and Europe in which schools are given feedback on the performance of their students, and they are given support and challenges to improve. They’re given information from various parts of their constituencies about how their clients evaluate the schools. They are put into a reasonable school improvement structure that is calibrated around what people actually know how to do.

You come to the United States, and you discover that we’re incredibly overinvested in testing and sanctions and incredibly underinvested in capacity building. You don’t have to convince the people in schools that they’re failing.

They know that, and they know it in all the ways that NCLB wants them to know it — they know which populations of students are failing, they know in which subjects they’re failing, and they know how many years in a row they’re failing. What people in schools don’t know is the actual process by which you take a school from point A to point B, what are the resources that are required to do that, and how do you organize and mobilize those resources.

That requires developing a practice around school improvement, and that’s what they’re doing in places like Canada and Australia, and it shows in their international performance.

JSD: What’s your best hope for a reauthorized NCLB?

Elmore: I would be delighted if the legislation would put a lot more emphasis on supporting research and development around school improvement processes and empirical evidence about how schools actually improve, as opposed to how we fantasize about how they improve. I would be delighted if they put money into exemplary programs for building up the profession around the practice of school improvement and some encouragement to engage in alternative certification and professional development models based on some of the work that exemplary districts are doing.

My problem with Adequate Yearly Progress is not that we shouldn’t have some kind of performance-based incentive structure. My problem with AYP is that we don’t know how to do what we’re requiring schools to do.

It’s not the policy makers who are going to make this period of educational reform successful — it’s the people on the ground who are going to do it. They don’t have all of their best ideas stashed away in some desk drawer somewhere — they’re doing what they know how to do. If they’re not doing the right thing, we need to figure out how to put them in a situation where they can learn how to do it differently.