In 2007, only 57 percent of fourth graders in New York City and 44 percent of fourth graders in Chicago could claim even basic literacy according to the National Assessment of Educational Progress. Yet, in the same year, less than 2 percent of New York’s teachers and less than 1 percent of Chicago’s teachers were deemed “unsatisfactory” in their official evaluations. Clearly, something is missing here.
Current public-school evaluation systems do not distinguish between effective and ineffective teachers. We can dramatically improve these systems by thoughtfully incorporating information gained from student performance on standardized tests.
But union opposition to measuring teacher quality makes even the most obvious policy reforms frustratingly difficult. In fact, thanks to union pressure over the years, it is actually illegal in some states — California, New York, and Wisconsin — to evaluate teachers based on standardized test results. To his credit, President Obama has insisted that states abandon such restrictions in order to qualify for part of the $4.3 billion in discretionary funds under the administration’s Race to the Top initiative. The president and his education secretary Arne Duncan deserve praise for (so far) sticking to their guns despite outcry from these states.
A wide body of research confirms what everyone already knows: Teachers are the most important factor (within a school’s control) for developing student proficiency. This research suggests that teacher quality varies dramatically — a student assigned to a high-quality teacher learns about a grade-level’s worth more material in a year than if assigned to a low-quality teacher. This should make improving teacher quality the primary objective for any education reformer.
The most promising ideas for improving teacher quality — such as paying more effective teachers higher salaries or loosening tenure restrictions that make it impossible to fire ineffective teachers — require that we meaningfully and systematically measure a teacher’s contribution to his student’s learning. Current evaluations are not up to the task.
Most current teacher-evaluation systems rely heavily — and sometimes entirely — on classroom observations. The principal observes the teacher and makes a subjective assessment of her classroom-management techniques. The observation can last less than a single class period — according to the union contract, in Miami, Fla., the annual evaluation need not last longer than 20 minutes — and usually occurs at a time known beforehand by the teacher, allowing him or her to fix the scene. I personally remember a teacher in the public school I attended — who was mediocre but by no means incompetent — pleading with us to be on our best behavior the next day when “guests” would be visiting our classroom. I realize now that those guests were evaluating her.
These teacher observations are too infrequent to be informative. Recently hired, untenured teachers might be observed two to three times a year. But a U.S. Department of Education study of teacher evaluations in the Midwest found that in more than half the districts studied tenured teachers were only evaluated once every three years.
The subjectivity and infrequency of teacher evaluations makes them essentially useless. Further, principals are reluctant to rock the boat by publically deeming a teacher “unsatisfactory,” particularly since (1) they can’t remove a tenured teacher, no matter how unsatisfactory his performance, (2) the rarity of the distinction implies the teacher is not only “unsatisfactory” but egregiously incompetent, which is often a stronger message than the principal intends to send, and (3) the principal himself is not directly accountable for the school’s performance, giving him little incentive to enrage his teaching staff by pointing out the poor performance of a colleague.
We can greatly improve teacher evaluations by using modern standardized testing and econometric tools to create objective measures of a teacher’s classroom performance. Datasets matching student test scores to their teachers allow us to measure the influence of a teacher on student performance. Such analyses are not reliable enough to use in isolation to make any employment decisions about a teacher. But, they can certainly be used to raise a red flag when a teacher’s students are performing poorly.
Thanks to the ubiquity of standardized testing, all states produce yearly data on student proficiency. However, only 21 states have developed sophisticated data systems useful for teacher evaluations. Mr. Obama and his education Secretary Arne Duncan say they will use Race to the Top dollars to encourage more states to follow this path. Their decision to withhold funds for states that outlaw this important practice deserves an A+.
– Marcus A. Winters is a senior fellow at the Manhattan Institute. His new report with Jay P. Greene, “How Special-Ed Vouchers Keep Kids From Being Mislabeled as Disabled,” can be read online here.