The campaign to improve teacher quality in public education depends heavily on policymakers’ ability to put in place new systems for evaluating teachers’ work—defensible evaluations that allows policymakers to credibly argue that employment decisions be made on the basis of teachers’ performance rather than their credentials, and help teachers improve their performance. In most school systems today teacher evaluations are surpassingly superficial and largely unable to serve as the foundation for tenure, promotion, or any other employment decision, much less provide the basis for effective teacher professional development.

The federal government can support the creation of new evaluation models and the scaling up effective systems already in place, both pushing states in this direction through the reauthorization of the Elementary and Secondary Education Act, and by supporting state and local innovations through grant programs like the existing Teacher Incentive Fund.

The administration’s emphasis in its Race to the Top grant competition on providing student test scores substantial status in evaluations has produced mixed results. It has caused evaluations to focus on the educational bottom line to an extent they rarely have in the past. But the price has been high.

Experts, even those who are reluctant to make the good the enemy of the perfect, have questioned the dependability of test scores as measures of teacher quality. Some significant numbers of teachers produce great scores one year and low scores the next, for example. The respected federally funded National Academies concluded in a recent report that test scores "should not be used as the sole or primary basis for making operational decisions because the extent to which the measures reflect the contribution of teachers themselves, rather than other factors, is not understood."

Perhaps more importantly, in the long run, teachers, rightly or wrongly, hate being judged primarily on test scores. Teacher buy-in is critical to the success of teacher reforms, whether reformers like it or not. So what’s the point in pressing for test scores to play a leading role in evaluations if the result is merely to alienate teachers? Better, it seems, to focus on creating effective models for evaluating classroom teaching. Research by the Gates Foundation suggests that classroom evaluations (multiple evaluations by multiple trained evaluators are the best strategy) correlate with student test scores anyway. And evaluators, unlike test scores, can give teachers feedback to help improve their craft, something that teachers value highly, surveys show.

The premise of the question seems to be that the route to teacher quality is to identify our best and worst teachers on the basis of the performance of their students on standardized tests, and to use that data to reward the best and get rid of the worst.

If test-based accountability was the key to national education success, we would expect to find policies of that sort in the countries with high quality teaching forces and top-performing students. But this is not what we find at all. Not one of the top-performing countries have policies of that sort. The top-performing countries agree that teacher quality is key to student performance but they have a very different idea about how to get it. They set pay for their teachers at levels comparable to their high status professions, they set very high standards for entry into teacher education institutions, and don't provide licenses to teach in their schools unless their teachers—even at the elementary school level—have a very firm command of the subjects they will teach and have spent a least a year really mastering their craft. And they never, never waive those high standards for teachers in the face of teacher shortages. They don't have to, because they have more highly qualified teachers than they need. The top-performing countries have anywhere between six and ten applicants for every opening in their teacher education institutions.

Sometimes it seems as though we think we can fire our way to a top-quality teaching force, that people think we can somehow produce a top-quality teaching force by getting rid of our worst teachers. But where are their replacements supposed to come from? Our most successful competitors think the way to get a top quality teaching force is to make teaching very attractive to their best high school and college graduates. That's where they are placing all their bets.

Many of these points were made in a report to Secretary Duncan last December, a report that was presented by the OECD and prepared by the National Center on Education and the Economy under the supervision of the OECD. A few days ago, the Secretary proposed that teachers be paid on scale ranging from $60,000 a year to $150,000 a year and that standards for getting into our schools of education be greatly raised. This is a proposal that is moving in the right direction. The reader can find a detailed proposal of this kind in NCEE's report, Tough Choices or Tough Times.

So I would reframe the question on the table as follows: What should the federal government do to improve teacher quality? And my answer would be: Offer some money to the states to do what the countries doing the best job of educating their students are doing about teacher quality. This does not include test-based accountability. It does include the points made above as well as others. The interested reader will find the full list of findings concerning what the top-performing countries are doing about teacher quality in Surpassing Shanghai: An Agenda for American Education Built on the World's Leading Systems, to be published in October by the Harvard Education Press.

Teacher quality is the result of a wide range of policies affecting the whole system of teacher recruitment, selection, education, compensation, training and mentoring. The single most important factor is the pool from which future teachers are selected. Only the state is in a position to build such systems. No single not-for-profit organization or school district is in a position to pull all these policies into alignment except the state. The way to do it, then, is to run a competition among states interested in taking advantage of the experience of the top-performing nations to create world-class teaching forces for their schools. Such a competition would summarize what the research says about the methods being used by the top-performing countries, including the variation among them from country to country. It would not be dogmatic about the steps to be taken, the targets to be hit or the standards to be set. The proposals would be judged on the basis of whether the state had put together a coherent, powerful proposal, based on the experience of the top-performing nations, that was well adapted to the realities and ambitions of the state and whether the reader believed that the state had the capacity to implement the plan it advanced. This would leave great latitude for states to propose plans that make sense to them.

And it would enable the United States to build its strategy for creating a high quality teaching force on empirical data on what really works at the scale of a state or nation, not ideology or guesswork.

Steve makes quite reasonable points about core components of what an evaluation policy should aim for, but they are not what has happened with Race to the Top. Most of the "winning" states have decided to make student "achievement" count for 40-50% of the teacher evaluation, with student standardized test scores being the preponderant factor in "achievement." The approach narrows education, limits reasonable autonomy, focuses on monetary rewards, reduces incentives to care about real learning, pits educators and students against each other, and reduces the craft of teaching to results that can be measured on hopelessly inadequate tests. It is clear, in the wake of NCLB, that basing accountability “in significant part” on test scores is to open the door to the narrowed curriculum, teaching to the test, cheating and, per Cambell’s Law, distorting both the measuring instruments and what is being measured. (On some of these points, see http://www.fairtest.org/flawed-ma-teacher-evaluation-proposal-report-home.)

We don’t need any more federal “nudges” like large competitive grant programs that bribe or coerce states into using student test scores to judge teachers. The nation would be far better off not having the federal government do anything on teacher evaluation, rather than perpetuate and enshrine in law, via Title I, Title II, competitive grants, or any other means, the dangerous conceptions that undergird both NCLB and RTTT.

If Congress is determined to act (though I don’t think it will reauthorize ESEA in this session), then it should follow these points:

1) Do not mandate that any state have a state-level teacher evaluation system.

2) Do not mandate that student achievement have any set weight, or that student test scores be included in any particular way.

3) Proceed very cautiously – slowly, in small steps. The goal should be to implement carefully, evaluate thoroughly, fine tune, and then build from there. In that line, it would make good sense to evaluate existing programs, both ones that appear not to work (DC’s IMPACT, which relies heavily on test scores) and ones that do (Montgomery County, MD, which does not include test scores). Between evaluation of existing programs and careful development and evaluation of some new pilots, evidence of what works and why can be obtained. (The letter Kevin links to emphasizes these points.)

4) In the evaluation of programs, do not pretend that test scores equal achievement. Not only are state test scores largely corrupted (per Campbell’s law), the tests simply don’t measure most of what students should learn.

5) Do require a preponderant role for educators in the design of the evaluation system, with meaningful roles for parents, students and community people in guiding the development and fair use of a system. Getting “buy in” for something others have designed is not appropriate or adequate.

6) Require a cost-benefit analysis. From the quite reasonable idea that high-quality evaluation can help improve teaching and learning, we have leaped to the possibly erroneous conclusion that systems mandated or coerced by the federal government are more worth the time and money than would be other uses of federal, state and local funds. For that, we lack evidence.

When we combine the lack of knowledge that evaluation systems are more valuable than other uses of resources with the high likelihood of causing more damage through again over-emphasizing testing, the reasonable conclusion is that Congress should proceed very slowly.

Cash incentive programs such as the Teacher Incentive Fund (TIF) are not enough to move states to adopt rigorous teacher quality standards and get us to a national norm of an effective teacher in every classroom. Competitive grant programs play major roles in incentivizing innovations and creating models of practice for other states to emulate. TIF, in particular, promotes systemic changes in approaches to attracting, evaluating, and developing educators. But one should not think that programs like TIF alone can take care of our country’s immense teacher quality issues. Our nation needs a broader approach that includes both formula and competitive grant programs, as well as improvements in our current efforts. There are ways to continue to move the states forward on teacher quality independent of a reauthorized Elementary and Secondary Education Act.

Law, regulation, guidance, monitoring, and bully-pulpit leadership can demand better uses of existing federal funds and requirements. For example, although some of the professional development supported through Title II is used effectively, there is little evidence that the way most districts use professional develop dollars produces a positive impact on student achievement. Better guidance and a focus on evidence-based practices could offer improvements and efficiencies.

The No Child Left Behind Act currently requires that teachers should be equitably distributed and that indicators of equitable distribution are collected. Unfortunately, these requirements are not well enforced and these indicators are not presented in a useful manner. The Center for American Progress (CAP) and the Education Trust have recommended creating A Teacher Quality Index that combines these data in ways that states can easily measure and report on progress on inequities within and among districts. This is not a case of over-regulating states. It is a case of governments acting on information already collected to fulfill current responsibilities.

On their own, states (e.g. Illinois and Tennessee) are legislating reforms in teacher policies that require new teacher evaluation systems to inform critical human resources decisions. The Department of Education could provide leadership to states in developing statewide criteria for evaluations that districts will use to inform critical human resources decisions, including tenure, compensation, professional development, teacher assignment patterns, and dismissal.

Finally, within the context of TIF, let’s improve upon and maximize what currently exists. Like the Administration, CAP has recommended the authorization of a new Teacher and Leader Innovation Fund, or TLIC, within ESEA to award grants to states and school districts to support innovative strategies that attract and support effective teachers and principals in high-need schools. States or districts could use the funds to develop more aggressive recruitment strategies, strengthen tenure processes, and institute career ladders for teachers.

To keep the momentum on teacher quality reform, Congress, the Department of Education, and the states need to find ways to do what they are already doing or already required to do more effectively. The tools exist through monitoring, leadership, and strategic funding strategies to continue the move forward.

Before we can find a solution to something, we have to believe it's out there.

House Education and the Workforce Committee Chairman John Kline, R-Minn., says “We all know there can be no one-size-fits-all federal solution for ensuring an effective teacher is in every classroom.”

But I think the oft-uttered phrase, "We all know..." isn't quite accurate in this case.

With all due respect to Congressman Kline, I think there is indeed a one-size-fits-all solution here. Dan Pink describes it thoroughly in his book “Drive: The Surprising Truth About What Motivates Us”, and presents it concisely in this wonderful talk, illustrated most entertainingly by the Royal Society of the Arts.

The gist of Pink's argument is this:

1. Pay people enough to “take money off the table”. As Pink says in his talk, “Pay people enough so that they’re not thinking about the money, they’re thinking about the work.” (This makes sense especially in situations like education where workforce stability is important and tasks are cognitively complex.)

2. Give people appropriate degrees of autonomy. (As we take away people’s autonomy, they become less motivated to improve and less accountable for their results.)

3. Help people find purpose in their work. (This should be pretty easy in education.)

4. Support people in mastering their craft. (This is easy to do, too, and we have many ways of doing it.)

Each of these things seems reasonable. And each is backed by research.

As I see it, these four ideas represent a one-size-fits-all approach to improving teacher quality. Actually, this seems to be a one-size-fits-all approach to improving the quality of the work done by anyone whose job involves performing complex cognitive tasks. If we scan our society, we’ll probably notice that the groups of people we hold in highest esteem for the quality of their work are treated more or less in accordance with this model.

We can say that we don’t think teachers deserve better pay, that we can’t afford to pay them more, or that we need better structures to pay them more equitably so that some can make a lot and so that those who make a little will aspire to improve. But research shows that this kind of incentive-based approach doesn't work very well. Getting teachers off the traditional "step and lane" model and over to something like a "rank and role" model seems like a logical step. But as Pink states, the real key is to “Pay people enough so that they’re not thinking about the money, they’re thinking about the work."

We can say that we want to exert more control over what teachers do in order to achieve greater degrees of accountability, but research tells us that decreasing teacher autonomy won’t increase teacher quality. This means handling testing and standards differently than we do now. It doesn’t mean doing away with them. It means doing more with them to support teacher autonomy—exactly the opposite of what has occurred in the last decade and where we appear to be moving with CCSSI and the testing consortia.

We can say that teaching is its own reward, and that no additional effort is necessary to help teachers find deeper meaning in what they do. But this runs counter to intuition and, again, to science. Curriculum standards, as we create and apply them today, make selecting material, delivering instruction, and assessing results—the essential acts of teaching—less purposeful. Annual tests, more-than-annual practice tests, and the teach-to-the-test culture these things inspire, make being a teacher less purposeful. If we choose to test differently, and to use test data differently, we can benefit from educational measurement without demoralizing the people we count on to improve it.

When it comes to the importance of mastery, educational and psychological research converge in an unsettling realization as we discover that teachers don’t improve much as they gain years of experience or as they pursue additional teacher education. Mastery is, literally, quality improvement. And science has shown us that it is a natural drive. An important question to answer, then, as we move forward with reform, is how we have discouraged so many people in schools—teachers, principals, and most importantly children—from expressing the natural desire all humans have to want to get better at things.

I hear all the time that there are no one-size-fits-all approaches to solving certain problems in education. But I learn about such approaches all the time. Personally, I’ve also got a holster full of silver bullets I use to solve tough problems in my own work in schools, especially problems related to literacy, assessment, classroom management, and student motivation.

Education provides us with many challenges. But a large set of thoughtful and well-researched solutions exists, even in the notoriously controversial one-size-fits-all and silver builet categories.

The real challenge is not discovering these solutions or designing them from scratch; it's choosing to use them—and choosing to see them by banishing the belief that they don’t exist.

I agree with Sen. Alexander that Congress should be cautious about legislating too specifically to states about how teacher evaluation programs should be designed. But here’s a thought: Maybe the best way to figure out how to help more of our teachers become high quality would be to ask those who already are.

In response to such requests, the National Board of Professional Teaching Standards (NBPTS) has just released: Getting It Right: A Comprehensive Guide to Developing and Sustaining Teacher Evaluation and Support Systems. It’s a concise and practical tool designed to help states and school districts develop effective teacher evaluation and support systems. Note how I worded that: We believe no teacher evaluation system can be successful or sustainable unless it is also, if not primarily, a system to support the improvement of teaching.

As we proudly note in the Getting It Right preface, “One could argue that by involving dozens of universities, research labs, scholars, teacher leaders, and premier testing organizations around the nation [for over 20 years], the National Board’s investment in developing a voluntary, advanced certification system has built the field of teacher performance assessment to which states and districts turn today.”

A truly effective teacher evaluation and support system, according to the Guide, should have four major outcomes:

· Promote high levels of student performance

· Improve teaching practice

· Create more effective school environments

· Strengthen the public’s understanding of what teachers should know and be able to do.

I would assign the members of both the House and Senate Education Committees, and anyone who is truly interested in redesigning our teacher evaluation and support systems to study Getting It Right. Understanding how to approach this complex task may help identify the best roles for each of the stakeholders, including the Federal government.

Cross-posted at TeachMoore (www.teacherleaders.typepad.com/teachmoore)

Yes, “student achievement” should be a significant part of teacher evaluation.

But is “student achievement” synonymous with “a score on a single test given in the early spring”? Might other tests be important? How about papers, projects, reports, engaged conversations, etc? Are these assumed to correlate so highly with those standardized test scores that we don’t have to worry about them? And if for some reason we believe this, should we believe it will still be the case after we ramp up the stakes and Campbell’s Law kicks into a high gear?

We all know the answers to these questions, and they don’t bode well for continuing on the present course. The problem, though, is that while it is relatively easy to administer yearly standardized tests and then calculate value-added growth scores, it’s very difficult to attach a number to these other important aspects of learning and student achievement. So in lieu of throwing up our hands and taking student achievement out of the evaluation calculus, policy makers choose to dismiss these questions and concerns and opt to use the flawed measures we actually have.

The problem as I see it is undoubtedly one of measurement; the value-added scores have substantial and well-documented issues of reliability and validity. But the even bigger problem concerns the overall effects on student achievement. In the end, the overarching (and only important) reason for adopting these evaluation mechanisms is to drive improvement in teaching and learning. On this ground, the current policy push appears to fail, and there’s a real risk that it will fail miserably.

Dr. Carol Burris and I just published a letter we sent to Secretary Duncan on this exact issue. Please read it at: http://nepc.colorado.edu/publication/letter-to-Arne-Duncan