In the discussions of reauthorizing NCLB almost no one is talking about whether the school is the right unit of accountability. And that’s because we continue to assume that schools are the problem and they can be fixed. But what if school improvement was found to be a myth?

—-

It’s really hard to convince people that something as evidence-based as “school improvement” is today, largely a myth. After all, from the earliest studies of effective schools to research about Chicago school reform, school improvement as an idea and a policy is based on decades of research and practice that tell us that what happens inside schools matters for student achievement, and that people working inside schools can change what happens in a way that improves student learning.

But there’s a problem with that evidence. A problem I discovered when I wrote my dissertation, “The Myth of School Improvement.”

The problem is that there are two hidden variables in the research about “effective schools” and “improving schools” that were somehow missed in the rush to create policy to hold “schools” accountable.

Variable 1: Context stability

The first variable is something that I call context stability. Context stability is a combination of low teacher turnover, stable leadership, and a demographically consistent student population. Context stability is also about having continuity in curriculum and materials, programs and program staff from year to year –or something that researchers studying Chicago schools called, “coherence.” If you dig deep into the research on effective and improving schools you find out that all of them had continuity in staff, leadership and student demographics during the period studied. Staff and leadership stability was a condition of effectiveness.

Anyone who works in schools today can tell you that context stability is very uncommon, especially in schools deemed “in need of improvement.” Teacher turnover is an ongoing problem, particularly in schools serving large percentages of students living in poverty where the average teacher stays less than five years. And ironically, the federal School Improvement grants have convinced many district administrators that it’s a good idea to move school principals around. And in many locales, particularly urban ones with open enrollment policies and large immigrant populations, student demographics can change dramatically from year to year.

And the real rub is that context stability itself doesn’t last forever. Most research about effective or improving schools is done in a 1-5 year period. Give me an effective school or improving school and wait three years. The effective principal or effective program will have gone, and it’ll be back to square one.

Variable 2: Autonomy

The other important variable we failed to consider is autonomy. During the previous eras of school reform people working in schools had much more control over their curricula, their technology and their programs than they did today. The research on Chicago’s improving schools was conducted during the 1990s during an unprecedented experiment in local school control. Nowadays districts and states often dictate what materials teachers can use, what programs they can implement, and even what page to be on in a pacing guide. Some researchers say that schools should be responsible for “crafting coherence” but in my experience, that’s more pie in the sky idealism than reality, particularly when district-school administrator power dynamics are involved.

If you really think about it, schools are just buildings that have a constant and complicated flow of policies, programs and people moving in and out. School administrators and teachers have very little control of that flow of information, people and practices—they can only manage those things within the confines of district, state and federal policies.

It’s not surprising we missed this enormous problem when developing the logic of school improvement. I blame my field: We take for granted that a school is a relevant “thing” because we once believed the key to educational improvement could be found in organizational science and management. Those of us at Stanford–with our offices right down the hall from organization scientists– followed right in the footsteps of the organizational development field. And because of this, the field anthropomorphized the school building as an actor. We started writing about characteristics and qualities of “schools” so that leadership and professional learning communities became properties of the school organization, and not the dynamic interplay of the adults in the building.

Think about it: the idea that we hold a pile of bricks responsible for improvement is really quite nonsensical.

Now many policymakers might say, “yes, I agree, it’s about teaching and learning—it’s about the teacher, that’s why we’re holding teachers accountable.

But if schools have very, very little control over the people, policies and programs at their doorstep, consider that teachers have absolutely none. Is it really fair to blame Ms. Johnson for the failings of a poorly designed curriculum? For mistakes in the common core standards? Is it right to hold Mr. Hopkins accountable for low test scores because the ADHD kid went off his meds during a critical week of lessons just at the very same time that the school counselor went on maternity leave?

I think most people would agree that it’s not fair. And while the idea of value added modeling may be “evidence-based” that evidence was analyzed much in the same way as the analysis that gave rise to school improvement policy: correlational analyses of variables measured at one moment in time. We know value added scores lack stability– and if you think about what happens in classrooms over time–with the flow of people, policies and programs–it’s pretty obvious to me why. I don’t fault the economists for not getting this. They haven’t spent enough time in schools to understand–they’ve never had the chance to teach a group of diverse children in an under-resourced school.

My heart aches for kids who receive a substandard education. I believe that much of what happens inside schools needs to change. But the only way we will make great strides in this area is to give up on the myths that have led us to that point– to be really precise and logical about what it is that we are trying to improve and how to measure it.

In 2015, educational improvement is a problem that belongs to the entire field and institution. It’s a problem that we all own, as tax-payers, citizens and workers, as parents, teachers and youth. The policy remedies for holding a whole field accountable are complex, but I believe in democracy. We can do this.