Teacher evaluation reform, improving schools’ measurement of who is doing a good job in classrooms and who isn’t, has been nothing if not controversial in public education in recent years. But there have been many bright spots in the evaluation reform movement —which has expanded to 46 states—and they haven’t received the attention they deserve.
Here’s one example: A study by leading national researchers has found that the more precise information on teacher performance produced by many of today’s new teacher evaluation systems pays dividends in another part of public education’s human capital pipeline—teacher hiring.
Brian Jacob of the University of Michigan, Jonah Rockoff of Columbia University and several colleagues used the District of Columbia Public Schools’ (DCPS) comprehensive new teacher-measurement system to gauge the payoff of an ambitious hiring strategy introduced by the school system in recent years, a process that included written assessments, personal interviews and teaching auditions.
They found that the new Teach DC system for vetting applicants, one that replaced a model heavily reliant on teachers’ paper credentials and individual principal preferences, strongly predicted teachers’ classroom effectiveness. For example, applicants rated highly under the new hiring system subsequently earned much higher marks on the DCPS evaluation system than low-rated applicants who were nonetheless hired by DCPS. The gap, the researchers found, was twice as great as the growth in performance by traditionally hired DCPS teachers over their first three years in the classroom.
In contrast, research in recent years has consistently found that teacher licenses and completion of graduate education courses, markers that many school districts rely on heavily in hiring decisions, “have little or no power to explain variation in [teacher] performance,” Jacob and his colleagues write in their study, Teacher Applicant Hiring and Teacher Performance: Evidence from DC Public Schools, published by the Cambridge-based National Bureau of Economic Research.
The new DCPS hiring model, introduced in 2009 as a part of the school system’s comprehensive human capital reforms, provides principals a list of “recommended” candidates who successfully completed the new process. Recommended applicants are listed in an online data base, with links to their resumes, and can be filtered by subject area to help principals find candidates. Principals can also navigate through the online database to find out further information on how the applicants scored in the Teach DC process.
The new, more comprehensive vetting system adds between $370 and $1,070 to the cost of each new hire, the researchers estimate—a price tag that they suggest is “quite small relative to the anticipated long-run benefits to future students of hiring more effective teachers.”
DCPS would have had difficulty measuring the effectiveness of the new hiring process prior to the introduction of the school system’s comprehensive IMPACT teacher evaluation system in 2009. Before IMPACT, evaluations were sporadic and typically involved only a cursory classroom check-in by a principal or other administrator, the standard practice nationwide. IMPACT, in contrast, includes clear teacher standards, multiple observations by multiple observers, student achievement results, student surveys, and other factors—yielding a much clearer picture of who’s shining in the classroom and who isn’t.
There is one very big caveat in the otherwise encouraging study: the researchers found that many applicants rated highly under the new application system don’t end up getting hired, and many lower-rated applicants end up in the city’s classrooms. The problem, the study’s authors suggest, is that principals don’t rely on the new hiring data as heavily as they should.
That’s no small matter. But we wouldn’t even know the problem existed if we weren’t able to confidently compare applicants’ performance once they were hired—something that wasn’t possible before the introduction of DCPS’s new teacher evaluation system.