A multitude of research has shown that quality teaching is necessary for students’ achievement and positive labor market outcomes. Rigorous evaluations have been hailed as a way to improve the teacher workforce by recognizing and rewarding excellence, providing detailed and ongoing feedback to improve practice, and identifying low-performers who should be let go. While plenty of time has been devoted to how best to provide teachers with feedback, less time has been spent examining how evaluation systems contribute to the removal of underperforming teachers and the resulting changes in the teacher workforce.
This study examines The Excellence in Teaching Project (EITP), a teacher evaluation system piloted in Chicago Public Schools (CPS) in 2008. The program focused solely on classroom observations and used Charlotte Danielson’s Framework for Teaching (FFT) as the basis for evaluation (unlike many current systems, which rely on multiple measures including student test scores). Roughly nine percent of all CPS elementary teachers participated in the first year of the pilot, which was considered a “low-stakes intervention” since scores on the FFT rubric were not officially included on teachers’ summative evaluation ratings.
Prior to the use of the FFT, teachers in Chicago were evaluated against a rudimentary checklist of classroom practices. This overly-generous model led to nearly all CPS teachers (approximately 93 percent) receiving one of the top-two ratings in a four-tiered rating system. EITP, on the other hand, utilized the detailed, research-based set of components of the FFT and required teachers to be evaluated multiple times a year. Principals were trained extensively on how to effectively use the framework, and were required to have conferences with teachers before and after observations. Because FFT provided teachers and principals with far more detailed information about instructional performance than the previous system, the framework produced more variation in teacher ratings.
The pilot started with forty-four randomly selected elementary schools in 2008-09; the following year forty-nine schools were added. CPS worked with the University of Chicago Consortium on School Research to craft an experimental design for implementation, and the University of Chicago randomized schools to take part in the first and second cohorts. Both treatment and control schools were statistically indistinguishable in regards to prior test scores (reading and math) and student composition.
Despite the fact that the experimental design was only maintained for one year, researchers were able to determine how the pilot impacted teacher turnover. While there was no average effect on teacher exits, the researchers did find that teachers who had low prior evaluation ratings were more likely to leave the district due to the evaluation pilot. In fact, by the end of the first year of implementation, 23.4 percent of low-rated teachers in schools using the EITP pilot left the district, compared to 13 percent of low-rated teachers in control schools. Non-tenured teachers were also “significantly more likely” to leave. Overall, the first year of the pilot saw an 80 percent increase in the exit rate of the lowest performing teachers and a 46 percent increase in the turnover of non-tenured teachers. The loss of teachers who were both low-performing and non-tenured suggests that “contract protections enjoyed by tenured teachers provided meaningful job security for those who were low-performing,” as there was no difference in the exit of low-rated tenured teachers. Also worth noting is that teachers who remained in EITP schools were higher-performing than those who exited, as were the teachers who replaced exiting educators.
These findings suggest two important conclusions. First, teacher evaluation reforms like the EITP pilot can indeed impact the quality of the teacher workforce by inducing the exit of low-performers. In turn, by replacing low-performing teachers with higher-performing ones, achievement should in theory rise (though the researchers did not specifically test this hypothesis). Second, given that low-rated non-tenured teachers were significantly more likely to leave than low-rated tenured teachers, the researchers were able to surmise that “tenure reform may be necessary to induce low-performing tenured teachers to leave the profession.”
SOURCE: Lauren Sartain and Matthew P. Steinberg, “Teachers' Labor Market Responses to Performance Evaluation Reform: Experimental Evidence from Chicago Public Schools,” The Journal of Human Resources, (August 2016).
 The researchers note that although “the leave rate of low-rated treatment school teachers is imprecisely estimated because very few teachers received low ratings, it is remarkably stable and large in magnitude.”
 In CPS, teachers who are in their first, second, and third year of teaching are non-tenured.