Corrected: A previous version of this article misspelled Hannah Putman’s last name.
In many professions, including law and nursing, the percentage of candidates who pass a licensing exam on their first try is routinely reported. But that data for the teaching profession hasn’t been made public—until now.
New data show that many aspiring teachers do not pass their state’s licensing exam on the first attempt. And nearly a quarter of those candidates who fail do not try again, quashing their plans to teach. That’s even higher for test takers of color—30 percent don’t retake the test after failing the first time.
The first-time pass rates on elementary teacher licensure tests and the disaggregated data on race and ethnicity have been under wraps for more than two decades and , a Washington-based think tank that advocates for more rigorous teacher preparation. NCTQ spent two years collecting the data from state education departments, sometimes via public records requests, and ultimately obtained data from 38 states and the District of Columbia.
States generally require aspiring elementary teachers to take a licensing exam that covers the foundations of the content they’ll teach to their students: English/language arts, mathematics, science, and social studies. Teacher candidates have to pay to take—and retake—the exam, which can cost up to about $200.
“First-time pass rates may be a reasonable indicator of the quality of preparation candidates are getting,” said Hannah Putman, the managing director of research at NCTQ and a project lead on the report. And “every time a teacher has to retake [the exam], it costs them time, money, and anxiety. A lot of people who fail the test don’t retake the test.”
Other professions that require entry exams, like nursing, law, and accounting, report first-time pass rates to monitor the quality of programs’ preparation. NCTQ and other experts argue that those in teacher education should readily report it, too. (The American Association of Colleges for Teacher Education, which represents teacher-prep programs, declined to comment before the NCTQ report’s release.)
“I’m excited about this work not because I think it tells us anything super specific about how to improve teacher education, but it points us in some interesting directions and shows us just how much variation there is among different institutions in [regards to] pass rates,” said Dan Goldhaber, the director of the Center for Analysis of Longitudinal Data in Education Research at the American Institutes for Research, who studies teacher licensing exams and previewed NCTQ’s report.
There are significant differences among teacher-prep programs within even the same state: NCTQ found an average 56 percentage point gap between the institutions with the highest first-time pass rate and those with the lowest pass rate. And six states—Connecticut, Florida, Louisiana, New Jersey, South Carolina, and Virginia—have at least one teacher-prep program where not a single test taker passed on their first attempt.
“When you start to look program by program at the difference in the likelihood [of passing the test], it’s opening the door to thinking about what it is that the candidate is getting from the program in terms of encouragement and additional help,” Goldhaber said.
Congress tried to ‘get a handle’ on program quality
In 1998, the U.S. Congress passed Title II of the Higher Education Act in an attempt to hold teacher education accountable. Teacher-prep programs were asked to report their passing rates on licensing exams, among other data. States were also required to rank their programs and report additional information about teacher quality in the state.
“We were looking for a way to get a handle on the quality of teacher preparation,” said Charles Barone, who was, at the time, the legislative director for Rep. George Miller, a Democrat, who spearheaded much of this effort.
But many institutions were reporting 100 percent pass rates on teacher licensure tests. That’s because they required candidates to pass the exam in order to complete their preparation program—and they were only reporting the data for program-completers.
“The data was effectively useless,” said Kate Walsh, the president of NCTQ.
In 2008, the law was rewritten to include, among other measures, a requirement for programs to report the percentage of students who passed any single assessment. Unlike the summary pass rate data, those data are not limited to only program-completers. That more-granular data are more useful, Putman said, but still have limitations: “We don’t know how many people are passing it [on the first try], we don’t have data on race or ethnicity, we don’t know how many times people are taking [the exam].”
Also, if a licensure test has multiple subtests, the pass rates are reported at the subtest level but not the composite level. This masks the true percentage of people who fail the test overall, Putman said.
Reporting a fuller picture of pass-rate data is standard in all other professions with licensing exams and should be considered a “consumer protection mechanism” for teacher candidates, said Barone, who is now the vice president of K-12 policy for the advocacy group Education Reform Now. Candidates should know if the program they choose can successfully prepare graduates to be able to pass a test that’s a requirement to go into the classroom, he said.
“It’s one piece of the puzzle [for program quality], but it’s an important piece,” he said.
The Higher Education Act, which is supposed to be renewed every five years, hasn’t been reauthorized since 2008. Congress was considering an update last year before the pandemic happened. If lawmakers pick it up again next year, Barone said he hopes they will require states to collect and report first-time pass rates.
“That seems to me a no-brainer, and NCTQ has proven that it can be done,” he said.
Twelve states opted not to share their data with NCTQ or only supplied incomplete data. They are: California, Georgia, Indiana, Maine, Mississippi, Montana, North Carolina, Oklahoma, South Dakota, Utah, Washington, and Wisconsin. Walsh said one of the schools chiefs for those states declined to share the in-depth data, saying, “We’re going to look bad.” (She declined to name the state.)
Experts debate the value of the licensing exam
NCTQ says that 20 states and D.C. have what the group considers to be a strong testing system—meaning the licensure test is required of all candidates, it’s structured to separately score each content area, and the minimum passing score is aligned with what experts and practitioners recommend. In states with those strong testing systems, the average first-attempt pass rate is 45 percent. But 79 percent of test-takers eventually go on to pass the test.
In the 30 states with what NCTQ considers to be a weaker testing system, the average first-attempt pass rate is 76 percent, and 89 percent of test-takers eventually pass the test. However, NCTQ warns that having lower guardrails for entry into the profession means that incoming teachers may have gaps in their core knowledge.
“We shouldn’t be expecting teachers to be learning the lessons the night before they teach them,” Putman said.
Yet teacher licensing exams have been criticized as serving as a barrier into the profession, especially for aspiring teachers of color, who fail the test at higher rates than white candidates. As states and school districts struggle to attract teachers, some whether licensing tests are necessary.
Just this month, California policymakers said teacher candidates , which test reading, writing, and math skills. Candidates can now bypass that requirement by taking relevant college-level coursework.
“These tests are meant to accurately measure readiness to begin teacher preparation, not to be a barrier that keeps potentially great teachers from learning to teach,” said Mary Vixie Sandy, the executive director of the California Commission on Teacher Credentialing, in a statement. “As alternatives to high-stakes testing, these measures will right-size the role of testing and allow a broader and more diverse array of people to make a career out of teaching.”
For the report, NCTQ reviewed the available research on licensure tests, and found that 11 studies found a statistically significant positive relationship between a teacher’s test scores and their effectiveness in the classroom. Two studies yielded mixed results, one found no statistically significant relationship between the test and classroom effectiveness, and one found a negative relationship between California’s basic skills test and students’ reading achievement.
It’s a tradeoff, Goldhaber said: “Licensure tests are likely to be an impediment to diversifying the teacher workforce. That is bad. On the other hand, licensure tests are one of the only preservice measures of teacher effectiveness—that is good.”
Still, Emery Petchauer, an associate professor of teacher education at Michigan State University who studies licensure exams, said the research on licensing tests relies on a limited indicator of student learning—standardized test scores—and doesn’t take into account the full picture of teacher effectiveness. For example, he said, licensure exams can’t measure how teachers advance equity in their classrooms or how they build relationships with students.
“Multiple data sources give us the best understanding of something,” said Petchauer, who was not involved in NCTQ’s report. “I get worried when a single high-stakes standardized test can trump other indicators of what a teacher knows and is able to do.”
Even so, since licensure exams are a requirement to enter the classroom, programs should better support their candidates so they can pass the exam, he said.
The NCTQ analysis found that programs with fewer students who receive Pell grants, a measure of students’ financial needs, tend to have higher first-time pass rates. But NCTQ found 161 institutions with relatively high percentages of Pell grant recipients where the first-time pass rate is higher than the state average.
Said Walsh: “We do think that states ought to be asking some hard questions of institutions that have really low first-time pass rates. … We shouldn’t be afraid of this data. This data can help programs get better.”
