The Effect of Information on Employee Evaluation in the United States

Jonah Rockoff
Fieldwork by:
New York City, United States of America
223 principals
2007 - 2008
Target group:
  • Teachers
  • Managers
Outcome of interest:
  • Productivity
Intervention type:
  • Information

Principals are responsible for evaluating teachers to improve student outcomes. However, information gathering can be both subjective and costly. Researchers evaluated the effect of distributing “value-added” teacher evaluations to principals in New York City. Their findings suggest that principals found the reports useful and used them to inform their overall perceptions of teacher effectiveness.

Policy issue

Teachers are constantly evaluated by principals, but these evaluations may be based on imperfect information. Information-gathering costs principals time and money, and the process itself may lack rigor and objectivity.

In their role as employers and evaluators, principals observe and critique teachers, allocate training and development resources, and ultimately decide if and how they will employ teachers at their schools. Economic research provides substantial support for distributing standardized teacher performance reports to principals to help them make these managerial decisions.

Studies suggest that productivity varies greatly across teachers, and that principals play an important organizational role in ensuring student achievement. In spite of principals’ substantial influence on the education system, little research to date has examined principals’ perceptions of teachers, as well as the avenues by which to inform those perceptions.

Context of the evaluation

Researchers focused their study on New York City principals. A baseline teacher evaluation revealed that principals generally approve of their staff – three-quarters of teachers received an above-median rating. Principals reported formally observing teachers an average 2.2 times the previous school year, with around 6.4 total observations per teacher.

Descriptive data also suggest that principals believed they could objectively evaluate teachers; more than three-quarters of principals strongly agreed with the statement: “I know who the more and less effective teachers are in my school.”

At the same time, less than half agreed with the statement: “Anyone can learn to be an effective teacher.” This suggests that principals had strong prior perceptions about teachers, and that principals did not believe training could help all teachers.

Details of the intervention

In 2007, the New York City Department of Education enacted a suite of policies to give principals more decision-making power to improve student performance. The Department also instituted a new accountability system that, depending on school performance, could result in principal bonuses of up to $25,000 or dismissal.

To assist principals with their newfound autonomy and responsibility, the Department also offered them the opportunity to receive teacher “value-added” (i.e. impacts on student test scores) estimates. Each teacher was compared to “peer” teachers citywide, with similar levels of teaching experience working in classrooms with similar student composition.

Researchers collaborated with the Department to design a randomized evaluation around the 223 principals who completed a baseline survey in August 2007 which solicited principals’ evaluations of their teachers. Of the 223, researchers grouped principals based on school configuration (elementary, middle, and kindergarten through eighth grade) and randomly selected 112 to receive value-added reports, conditional on the principal’s attendance at a training session organized by the researchers. The remaining principals received no information and served as the control group.

In May 2008, after the intervention, researchers again surveyed principals on their evaluations of teachers to determine to what extent and how the reports had been used.

Results and policy lessons

Principals found reports useful and changed their evaluations accordingly.

Program take-up: Principals provided largely correct answers to survey questions testing their understanding of the reports, and 85 percent rated reports as either “useful” or “very useful.”

Perceptions: Principals’ prior teacher perceptions were more in line with valued-added information when they had supervised the teacher for a longer period of time. Additionally, principals were more likely to change their evaluations in response to value-added estimates if their prior perceptions proved imprecise. This suggests that value-added estimates provide principals with useful, but not overly influential information.

Turnover and student achievement: Researchers also observed some instances of more exiting by teachers with lower value-added estimates, as well as marginal improvements in student test scores within the treatment group. However, these results are not statistically significant.

These results suggest that standardized teacher performance data are useful, and that government provision of evaluations can help overcome otherwise costly information gathering.