October 13, 2006

Fixing Public Schools

Does teacher certification affect student performance?
Print this story

In 2003, New York City fired several thousand uncertified teachers. The city was doing its best to comply with a state law, passed a few years earlier, that said only certified teachers could work in public schools. Professor Jonah Rockoff decided to take a look at the performance of these fired teachers. As it turned out, they weren’t any less effective, on average, than the certified teachers who remained on staff.

For the past two years, Rockoff has studied the relationship between certification and teacher effectiveness. With his research partners, Thomas Kane of Harvard University and Douglas Staiger of Dartmouth College, he compared how similar students taught by traditionally certified, uncertified and alternatively certified teachers fared on standardized exams from 1999 to 2005. Because of its size and diversity, New York provided the perfect laboratory. “The good thing about New York City is we have so many data points,” Rockoff says. “We can tell whether a result is statistically significant or not, even if it’s very small.”

Almost every state allows its districts to hire alternatively certified teachers, who account for about one-third of all new teachers hired in the United States each year. Though the rules differ by state, alternatively certified teachers typically must have a bachelor’s degree, pass state exams, complete special training and, once they begin teaching, enroll in a teaching master’s degree program.

In New York, most alternatively certified teachers come from the Teaching Fellows program, which recruits professionals without any prior teaching experience, puts them through a teaching boot camp and sends them off to the classroom. The city also hires alternatively certified teachers through Teach for America, a nonprofit group that places teachers in school districts across the country, and international recruitment. Under the state’s emergency provisions, New York is allowed to hire alternatively certified teachers to cope with its perpetual teacher shortage. The city hired more than 50,000 new teachers during the years covered by the study.

To measure the effectiveness of the city’s teachers, Rockoff had to control for factors that might make one group of students perform better than another, such as the students’ prior test scores. The study focused on grades four through eight, since all students in those grades must complete standardized city exams.

What they found challenged the conventional wisdom about teacher certification requirements. There were no major differences in performance among students taught by traditionally certified, alternatively certified or uncertified teachers. However, Rockoff found that there were wide disparities in effectiveness within each of the teacher groups.

“We’re able to measure pretty accurately at the teacher level how students are performing,” he says. “Having a highly effective teacher or having a mediocre teacher makes a large difference in student achievement.” The difference between having a highly effective and a highly ineffective teacher is about one-quarter of a standard deviation, or about half of the achievement gap between students who are poor and those who are not.

The next step for Rockoff and his colleagues is predicting whether a teacher will be effective or not before the hiring decision is made. So far, researchers have had little success answering this question. For example, Rockoff examined teachers’ college grade point averages and the selectivity of the undergraduate institution they had attended. He found little evidence that either is linked to classroom performance, though these are important factors for being accepted into programs like Teach for America.

Rockoff is now working on a follow-up project with incoming New York teachers that will focus on many nontraditional measures, such as personality types and cognitive ability, that may be linked to effectiveness. He will also test whether it is possible to predict effectiveness by studying videotapes of teacher interviews or by observing teachers give a short lesson to a real class. Currently, candidates for both the Teaching Fellows and Teach for America programs must prepare mock a lesson, but they don’t actually deliver the lesson to children.

For now, Rockoff suggests that school administrators and policymakers reassess their thinking on teacher qualifications. “Rather than worry about whether a particular teacher has certification or which program they come from, just be worried about whether they’re highly effective or not,” he says. “That’s what really going to make a big difference in student achievement for a district or a school.”

Jonah Rockoff is assistant professor of finance and economics at Columbia Business School.

Read the Research

Thomas Kane, Jonah Rockoff, Douglas Staiger

"What Does Certification Tell Us About Teacher Effectiveness? Evidence from New York City"

Download PDF 
View abstract/citation