Efforts to water down the state’s teacher evaluation system may lead to worse outcomes for Michigan’s neediest students. Spurred by misguided claims that job performance measures have a negative impact on teacher employment, advocates are pushing for reduced accountability measures that would put Michigan kids even further behind.

Michigan’s teacher evaluation system could provide a rigorous and standardized process for improving teacher quality and student outcomes. There is perhaps no time in the state’s history when a proven system for holding teachers accountable for their performance has been needed more.

Only 5% of eighth grade students in the Detroit public school system — the state’s largest district — demonstrated proficiency on the 2022 NAEP reading test. This was not significantly different than their performance in 2019 and 2009. They have scored significantly lower than students in nationwide urban districts each time they have been tested since 2009. An evidence-based evaluation system is critical if Michigan is to reverse this long-term performance trend.

Yet critics of the state’s teacher evaluation system seek to water it down, which will make it easier for ineffective teachers to remain in the classroom.

In a recent Detroit Free Press op-ed, Josh Cowen, professor of education policy at Michigan State University, questions the effectiveness of the evaluation system and blames it for teacher shortages. He criticizes the 2011 tenure reform legislation that requires schools to use student test data in evaluating teacher performance.

The 2011 tenure reform also requires tenure decisions to include evaluation ratings. Before then, teachers could receive promotions and earn tenure even if they weren’t producing results in the classroom. The 2011 reform helps ensure that effective teachers get retained — and prevents the promotion of consistently ineffective ones.

Michigan’s teacher evaluation system, if implemented as intended, should gradually improve teacher and student outcomes. It is modeled after similar evaluation systems, in other states, that have produced promising results.

The D.C. Public School’s IMPACT evaluation system has improved both teacher performance and student achievement since it took effect in 2012. It has done so well, in fact, that it inspired researchers at Stanford University, Brown University and the University of Virginia to study it in detail to find out why and how it works.

The IMPACT evaluation system includes student test data, among other measures. The D.C. public school system uses data from the PARCC, a standardized test administered to students in grades three to 11. Every teacher in the district gets an annual evaluation, which is made up of PARCC student data (25%), other student assessment data (15%), classroom instruction (40%), student survey data (10%) and community engagement (10%). Tennessee uses a similar system that has also proven highly effective.

Michigan’s evaluation system is designed to include multiple measures of teacher performance, too. The evaluation score is based, in part, on student standardized test data (20%) — but only for teachers of math and English in grades four to eight — and other student assessment data (20%). The remaining 60% of the score is based primarily on the teacher’s classroom instruction. A small percentage of the score also considers nonacademic factors such as classroom management, professional development and contributions to the school community.

Each of these evaluation systems assesses a teacher’s impact on student growth. This requires schools to compare students’ standardized test scores in a particular subject on several occasions. Statistical analyses that control for outside factors (socioeconomic status, for example) allow for a fair measurement of the teacher’s effect on student growth, or value-added contribution.

For underperforming teachers, standardized testing helps administrators identify areas for professional growth. It also removes bias that may enter through evaluations that focus more heavily on subjective measures, such as observing the teacher in a classroom.

Yet this last point is misconstrued by Cowen, who claims in his op-ed that “[t]est-based teacher evaluation in Michigan all but stacked the deck against teachers of color... Black teachers were less likely to be given strong marks than white teachers within the same school.” But the use of test data in evaluations is not the cause of these observed racial discrepancies — and the article Cowen cites to support his contention confirms this.

Teacher evaluation systems that fail to incorporate standardized test data risk placing too much emphasis on classroom observations and other subjective performance measurements. That subjectivity is where bias might creep in. The best evaluation systems incorporate a variety of measures, including student assessment data, to ensure a more consistent and fair process for rating a teacher’s performance.

Cowen also claims, “Although few teachers in Michigan rate poorly enough for mandatory removal, even one low rating can push a teacher out of the profession, instead of giving the teacher time to grow into the job”. But state law requires underperforming teachers to be given targeted support and ample time for improvement. It is only after receiving the lowest rating (“ineffective”) on three consecutive year-end evaluations that a teacher may be dismissed.

Cowen’s assertion that “M-STEP, like other standardized tests, is a guidepost, not a detailed roadmap” is on target. This is precisely the intent of incorporating student growth measures from standardized tests into the evaluation system. The student test data, in combination with other key performance data, help provide a comprehensive assessment of a teacher’s overall effectiveness.

Tying tenure decisions to teacher performance helps ensure Michigan’s schools are staffed by the best teachers available. Research consistently highlights the strong link between teacher quality and student performance. A teacher evaluation designed to improve student outcomes should include student test data, along with other measures of teacher effectiveness. Michigan students deserve no less.

Permission to reprint this blog post in whole or in part is hereby granted, provided that the author (or authors) and the Mackinac Center for Public Policy are properly cited.