Taking Teacher Quality Seriously

A collaborative approach to teacher evaluation

By Stan Karp

Illustrator: Randall Enos

Randall Enos

So what’s the alternative? If narrow, test-based evaluation of teachers is unfair, unreliable, and has negative effects on kids, classrooms, and curricula, what’s a better approach?

By demonizing teachers and unions, and sharply polarizing the education debate, the corporate reform movement has actually undermined serious efforts to improve teacher quality and evaluation. Corporate reform has detached these issues from the conditions that make improving teacher quality possible. Class sizes are growing and professional development budgets are shrinking. Federal and state plans are pouring hundreds of millions of dollars into data systems and tests designed to replace collaborative professional culture and experienced instructional leadership with a kind of “psychometric astrology.” These data-driven formulas lack both statistical credibility and a basic understanding of the human motivations and relationships that make good schooling possible.

But better alternatives do exist. Beneath the polarizing, political rhetoric of current “reform” efforts, there is a considerable common ground among educators, parents, and administrators on the need for:

  • Better support and evaluation before new teachers get tenure (or leave the profession, as nearly 50 percent do within five years).
  • Reasonable, timely procedures for resolving tenure hearings when they are initiated.
  • A credible intervention process to remediate and, if necessary, remove ineffective teachers, tenured or nontenured.

Good models for each of these ideas exist, many with strong teachers’ union support. One promising example is the Montgomery County, Maryland, Professional Growth System (PGS), which has taken a collaborative approach to improving teacher quality for more than a decade. Several defining features make the Montgomery County model very different from the test-based “value-added” or “student growth” approaches. PGS:

  • Was negotiated through collective bargaining rather than imposed by state or federal mandate.
  • Is based on a clear, common vision of high-quality professional teaching practice.
  • Includes test scores as one of many indicators of student progress and teacher performance, without rigidly weighted formulas.
  • Includes a strong peer assistance and review (PAR) component for all novice and underperforming teachers, including those with tenure.
  • Takes a broad, qualitative approach to promoting individual and systemwide teacher quality and continuous professional growth.

Developing and sustaining good teachers, rather than “getting rid of bad ones” has always been the main goal of the Montgomery County system. But real consequences for persistently poor performance are part of the process. New York Times education reporter Michael Winerip wrote that the program “has worked beautifully for 11 years,” providing teachers with “extra support if they are performing poorly and getting rid of those who do not improve.” In 11 years, the PAR component of PGS has led to approximately 500 teachers being removed from the classroom in a system of about 150,000 students with approximately 11,000 teachers and 200 schools. Over the same period, nearly 5,000 teachers have successfully completed the PAR process.

But PAR is only part of a professional growth system designed to improve teacher capacity throughout the system, not just identify and remove ineffective teachers. It’s a qualitative approach growing out of a shared vision of high-quality professional practice. Ironically, one of the system’s greatest strengths is the labor-management collaboration that is often expressly forbidden by recent “evaluation reforms,” which mandate top-down formulas and ban collective bargaining over the details.

The PGS begins with “six clear standards for teacher performance, based on the National Board for Professional Teaching Standards” and includes “performance criteria for how the standards are to be met and descriptive examples of observable teaching behaviors.” The six standards are:

  • Teachers are committed to students and their learning.
  • Teachers know the subjects they teach and how to teach those subjects to students.
  • Teachers are responsible for establishing and managing student learning in a positive learning environment.
  • Teachers continually assess student progress, analyze the results, and adapt instruction to improve student achievement.
  • Teachers are committed to continuous improvement and professional development.
  • Teachers exhibit a high degree of professionalism.

An extensive system of supports and professional development activities, including detailed protocols for assessing progress toward these goals, is outlined in various handbooks, evaluation rubrics, and contractual agreements. The system also provides resources necessary to turn these ambitions into real commitments.

How Does PAR Work?

For example, the PAR system relies on 24 consulting teachers (CTs) who are recruited from master teachers with at least five years of experience in Montgomery County Public Schools. The CTs make a commitment to work for three years as CTs and then return to a school for at least two years in a teaching or other nonadministrative position. CTs receive special training to work intensively with an average of 16 to 18 new teachers and/or experienced teachers referred to PAR by their principals. The supports provided by CTs include:

  • Informal and formal observations.
  • Written and verbal standards-based feedback.
  • Equitable classroom practice.
  • Coaching sessions.
  • Lesson planning.
  • Model lessons.
  • Co-teaching modeling.
  • Peer observations.
  • Classroom management.
  • Time management.
  • Alignment of school support.

CTs document their work but do not do formal evaluations. There are no numerical ratings or tiered categories. Teachers are evaluated as either meeting or not meeting the standards. CT reports go to the PAR panel, composed of eight teachers appointed by the union—the Montgomery County Education Association (MCEA)—and eight principals appointed by the administrators’ association. The panel reviews the documentation and makes a recommendation for nonrenewal/dismissal, an additional year of PAR, or release to the regular PGS evaluation process that covers all staff.

If either the teacher or the principal disagrees with the panel’s recommendation, they can initiate an appeals process that allows all parties to present additional information and speak to the panel, which ultimately reaffirms or alters its original decision. A tenured teacher dismissed through PAR does retain tenure rights and can appeal a dismissal decision. But, in practice, the PAR process generally documents fully the basis for such decisions and formal challenges to PAR decisions are rare.

Building Trust in the Process

After more than 10 years and several contract cycles, nearly half the Montgomery County teaching staff have personally gone through the PAR process. This has done a lot to “establish its credibility,” says Chris Lloyd, a high school communications teacher who is now the MCEA vice president and co-chair of the labor/management PAR panel. “Teachers overwhelmingly view the process as one that treats teachers fairly and is not a ‘gotcha’ process. It’s designed to help teachers grow, not to fire them.”

For example, PAR has built-in safeguards against one of the biggest complaints teachers typically have about evaluation procedures: arbitrary judgment by a single principal or supervisor. If a principal determines that a tenured teacher is not meeting PGS standards, a CT observes the teacher and provides a report. If the principal and CT disagree about the teacher’s performance, a PAR hearing is held that includes the participation of the teacher in question. During the past year, Lloyd says, there were approximately 50 hearings involving cases where principals and CTs disagreed about whether a tenured teacher was failing to meet professional standards. About half those teachers were referred to a second year of PAR support, approximately 10 were dismissed, and the others were released to the regular PGS process.

Although the system is spelled out in detail, what really makes it possible is the level of trust and cooperation that grew out of years of developing a collaborative approach to issues of teacher quality. The commitment to collaboration between the MCEA and the district is summarized in unusual contract language:

We define collaboration as a process in which partners work together in a meaningful way and within a time frame that provides a real opportunity to shape results. The purpose of the process is to work together respectfully to resolve problems, address common issues, and identify opportunities for improvement. To be successful, the collaborative process must be taken seriously and be valued by both parties. The process must be given the time, personal involvement and commitment, hard work, and dedication that are required to be successful. The partners will identify and define issues of common concern, propose and evaluate solutions, and agree on recommendations.

“It wouldn’t work without the level of trust we have here,” MCEA President Doug Prouty told the New York Times. “It took three to five years to build the trust to get PAR in place,” added former Superintendent Jerry D. Weast. “Teachers had to see we weren’t playing ‘gotcha.’”

Beyond PAR, the larger PGS system is based on the premise that “good teaching is nurtured in a school and in a school system culture that values constant feedback, analysis, and refinement of the quality of teaching.” Formal performance evaluations are part of “a multiyear process of professional growth, continual reflection on goals and progress meeting those goals, and collegial interaction.” The aim is to support “a collaborative learning culture among teachers in each school, integrating individual growth plans into school plans, and utilizing student achievement and other data about student results.”

There are separately articulated PGS standards and evaluation protocols for administrators, nonclassroom professionals, and support staff. Ideally, this contributes to a schoolwide sense of accountability and collective purpose that helps sustain healthy school communities, and there is significant evidence that it works.

Over the past decade, student achievement as measured by Maryland’s state assessments has increased across-the-board in every student subgroup—by race, ethnicity, and income level. Achievement gaps have narrowed at all grade levels in both math and reading. In grades 3 and 5 math, and grade 7 reading, the gap narrowed by 16 points; in grades 3 and 5 reading, it narrowed by more than 20 points. Beyond the test scores, 84 percent of Montgomery County’s students go on to college and 63 percent earn degrees.

The collaborative approach has extended beyond teacher evaluation issues. For example, Broad Acres Elementary School serves a population almost completely composed of students of color and free/reduced lunch students. In 2000, it was on the verge of state takeover and a reconstitution that would have included wholesale replacement of school staff and leadership. But a collaborative approach initiated by MCEA and embraced by the school leadership led to a sustained process of renewal and reform that has dramatically improved student performance and school culture. According to the school’s principal, “The reason Broad Acres succeeded was teacher leadership and everyone holding themselves accountable for every student.”

PAR or Race to the Top?

These successes and the national debate about teacher quality have brought new attention to the Montgomery County PGS/PAR model. Obama’s Secretary of Education Arne Duncan told Weast, “You’re going where the country needs to go.” Yet the PGS approach is exactly the opposite of where Duncan and federal policies have led the country.

Under the Obama administration’s Race to the Top (RTTT) competition, states have been pressured to tie teacher evaluation to student test scores. Maryland won a $250 million RTTT grant by promising to base teacher ratings on state test results. Implementing the grant in Montgomery County would have meant dismantling a successful system developed by collective bargaining that improves results for teachers and students. After failing to get a waiver from the U.S. Department of Education to continue using the PGS system, the school board withdrew from the state’s RTTT plans and forfeited its $12 million share of the grant funds.

If federal policy were serious about improving teacher quality it would be investing precisely in programs like peer assistance and review. One Harvard study estimates the cost at $4,000-$7,000 per participant. Instead, the federal government has poured hundreds of millions into the development of test-based data systems and pressed states to use them to rate both teachers and the college certification programs they came from. It’s wasted money chasing bad policy.

Montgomery County is not the only district that has implemented collaborative peer approaches based on collective bargaining. Long-standing peer review programs in Cincinnati and Toledo, Ohio; Rochester, New York; and elsewhere have shown various degrees of success.

A recent in-depth study of two California districts using PAR programs reached some striking conclusions about the current push for new and better teacher evaluation models. The study compared the types and quality of support provided by CTs in two districts using PAR, one near San Diego, the other near Sacramento. It also compared the work of the CTs with the more traditional performance reviews done by principals. Finally, it observed and analyzed the work of the joint labor-management PAR panels that reviewed the evaluations and recommendations of both the principals and the CTs. According to the study’s authors:

What we found belies conventional wisdom… Integrating support and evaluation can be a more effective approach to improving instructional practice than isolating one from the other. The programs… clearly show that PAR is a rigorous alternative to traditional forms of teacher evaluation and development.

In an era when policy makers are calling for better teacher evaluation, our research shows that peer review is far superior to principals’ evaluations in terms of rigor and comprehensiveness. Equally important, peer review offers a possible solution to the lack of capacity of the current system to both provide adequate teacher support and conduct thorough performance evaluations.

The study confirmed another benefit that Montgomery County union leaders and administrators had previously demonstrated. Collaboration about core issues like teacher quality and evaluation has ancillary benefits. The PAR panels “turned out to be problem-solving arenas where district officials and union leaders collaboratively addressed operational and policy problems that might otherwise have ended up as grievances or gone unresolved… We were struck by the collaborative labor-management interactions that form the foundation of PAR. Though both [districts] have in the past experienced rocky union-district relations, PAR has served as a springboard for building strong connections. More than simple collaborative efforts, through PAR, management and unions are doing the hard work of confronting tough, high-stakes issues and reaching accord on how to proceed when decisions carry real and human consequences.”

The Montgomery teacher evaluation and professional growth system has taken on increasing importance amidst reform debates that regularly target teachers’ unions as defenders of poor performance and obstacles to improving teacher quality. The MCEA’s foresight in making teacher support and professional growth a major collaborative initiative with administration helped create a system that has served as an alternative to the current mania for narrow, test-based teacher evaluation.

It’s been a long haul, not a quick fix. And, as Chris Lloyd says, it requires a convergence of progressive union leadership, collaborative administration, and well-functioning school boards, a combination not easily found in the current environment.

Just as with student assessment, evaluation can be a tool for improving teaching and learning or an instrument of bad policy and external control. The key in both cases is to make sure that people, not tests, are the point of departure and that real collaboration among all parties shapes the process.