, circulated Thursday through Brown University’s Annenberg Institute for School Reform, uses a combination of survey responses and statewide administrative records from Tennessee to create a framework for identifying how and where teacher shortages emerge. 

Those data principally come from school years leading up to 2019–20, anchoring the results in the pre-COVID era. But they will inevitably resound in debates over the pandemic’s effects on the education workforce, which have come to revolve around the central paradox of teacher shortages: Even as countless school and district officials say they’re struggling to fill positions, national labor statistics show only slight movement in teacher turnover rates the last few years. 

Matthew Kraft, a Brown economist and one of the paper’s authors, said that ambiguity around shortages arises from the decentralized nature of K-12 employment, which can bely the realities experienced by many teachers and administrators.

“Teacher shortages are real, period,” Kraft said. “Teacher shortages, however, are not universal. We’re trying to help people understand that it’s actually accurate for people to disagree about this because they’re answering from different perspectives.”

To conduct a comprehensive examination of K-12 employment, Kraft and his collaborators gathered response data from the Tennessee Educator Survey, an annual poll administered to thousands of teachers and school administrators by the state department of education. Tennessee offers a fairly representative setting, including several major urban districts along with substantial suburban and rural populations. 

Specifically, the study leveraged a survey item asking administrators if their school had a vacant teaching position at the beginning of the 2019–20 school year. Respondents at roughly 1,100 of Tennessee’s 1,740 schools answered that question, which provides a reasonable proxy for teacher shortages; while virtually all schools occasionally have to deal with unfilled jobs, classrooms that begin the academic year with insufficient or uncertain staffing may struggle to accelerate learning and establish relationships between students and teachers.

The responses, along with a decade’s worth of state information on student and teacher demographics, paint a picture of shortages that are both widely distributed throughout the state and narrowly experienced among schools themselves. To begin with, just 609 of the sample’s more than 40,000 teaching positions were vacant in the fall of 2019, with secondary schools accounting for 73 percent of all vacancies. But while three-quarters of all schools reported no vacancies, and just 6 percent of schools said that more than two positions were vacant, schools in virtually all of Tennessee’s commuting zones  — — reported at least one teaching vacancy.

Secondary schools in a handful of areas, including Memphis and Nashville, stand out as sites of acute shortage, but vacancies were seen throughout Tennessee rather than concentrating within specific regions or counties. About 80 percent of the variation in vacancy levels, in fact, was accounted for by particular schools within districts, rather than across school district borders. 

What’s more, the existence of vacant positions was significantly associated with certain features of schools and school communities. In keeping with earlier research showing that K-12 teachers — to a greater degree than other professionals with a similar level of educational attainment — near where they grew up, the authors discovered that schools employing a higher proportion of early-career educators who themselves attended a nearby high school (another question included on the statewide survey) reported fewer vacancies.

Teacher compensation also played an important role. A 0.5 percentage point increase in teachers’ scheduled salary bumps was correlated with a 36 percent drop in vacancy rates, while increases in a combined measure of self-reported working conditions (e.g., better school culture, more administrative support, strong relationships among faculty members) also drove down teaching shortfalls at the beginning of the year.

But perhaps the strongest indicators of unfilled teaching jobs were school-level turnover rates. This result may seem intuitive — more vacancies are naturally seen in workplaces where more people regularly leave — but Kraft and his colleagues highlight the finding as a validation of earlier work on ; for state and district leaders hoping to identify and support schools that struggle to staff classrooms, it is an important observation that past turnover tends to predict future turnover, especially when real-time data on teacher quit rates are typically hard to come by. In all, turnover rates were found to be 39 percent higher in schools with fall teacher vacancies than those without.

“For 20 years, we’ve been seeing that revolving doors predict shortages,” observed coauthor Danielle Edwards, a postdoctoral research associate at the Annenberg Institute. “There’s something important about that measure, [which] doesn’t include the past year’s turnover. It’s the idea that a particular school might always be losing teachers, and we need to identify those schools because it’s also going to make it hard for them to get teachers in the future.”

Certain academic specialities are also harder to find than others. While about 20 percent of school districts said they didn’t receive enough applications for open social studies positions, nearly two-thirds reported insufficient interest in math, science, foreign language, and special education roles. That unevenness suggests that standard salary schedules may be “ill-equipped to address the wide variability in…subject-specific needs,” Kraft said.

He concluded by arguing that the frequency with which teaching jobs sit open across different communities, and the commonalities between schools that see higher vacancy levels, shows that teacher shortages both before and after COVID are the downstream effect of policies that can be altered — whether by differentiating teacher pay to attract applicants with especially valuable expertise, or improving working conditions so that all school employees feel more valued.

“Shortages don’t just appear out of nowhere by some miraculous force; they are a function of decisions we make, and they can be influenced by macroeconomic patterns…like a global pandemic. But underlying them is an infrastructure that we’ve designed and have agency to change.”

Teacher turnover rates are actually about the same as they were before the pandemic, according to through the Annenberg Institute at Brown University. Flush with pandemic relief money and faced with the generational challenge of fostering learning recovery, school districts are hiring for more positions and leaving vacancies open for longer.

A wide-ranging analysis of employment trends from national and state-level sources, the brief does confirm that the K-12 workforce shrank significantly after the onset of COVID-19 and its disruptions to schooling. After roughly a half-decade of steady growth, total public school jobs decreased by roughly 9% through May 2020. The initial drop represented more than twice the number of positions erased during the financial crisis of 2008. 

But the data also suggested that those positions were disproportionately cut from non-teaching ranks. Occupational records from both national and state sources showed measured declines among nurses, administrative support staff, paraprofessionals and other predominantly non-instructional employees. Across all the states included in the study, there was actually generally less teacher turnover during the summer of 2020 — likely the residue of that year’s severe economic slowdown, which discouraged many from leaving their jobs. (During the summer of 2021, seven of those states saw an average turnover increase of 1.2 percentage points, effectively bouncing back to pre-pandemic levels.)

Confusion about the state of the education field has emerged due to a lack of consistently reported data on millions of school employees, the authors argue. In fact, the report was only made possible by combining several overlapping federal data sets — each with its own liabilities — with additional findings from 16 states that publicly reported annual statistics on turnover through the first year of the pandemic.

Matthew Kraft, a Brown economist and the paper’s lead author, said he was “very concerned” about the increased burnout teachers reported experiencing over the last few years. While a true mass exodus of educators hasn’t yet occurred, Kraft said that profession-wide exhaustion could someday trigger one. But he added that short-term instability in the education workforce has “obfuscated” longer-term issues of working conditions and public funding that demand more thorough examination.

“There’s no doubt that this story [of educator dissatisfaction and turnover] is catching our national attention, and it’s generating headlines,” Kraft said. “The problem is that most of those stories are asking a question for which there is a nuanced response, and nuance isn’t communicated effectively in our sound-bite world.”

Kraft and his co-author, Joshua Bleiberg, culled figures from four surveys conducted by the federal Bureau of Labor Statistics, each collecting regular reports from tens or hundreds of thousands of employers in both the private and public sectors. That information allowed them to not only generate month-by-month estimates of the total number of elementary and secondary education jobs, but also form a clearer view of the large swings in hirings, resignations and layoffs between March 2020 and May 2022.

The pair supplemented that picture with files from 16 state education agencies — though these additions were complicated by the states’ differing definitions of turnover. For the purposes of their study, Kraft and Bleiberg described it as the percentage of teachers in one school year who did not return to the same school or district in the next year.

One possible explanation for the vacancies that did linger was a period of weak job growth after schools were closed in spring 2020. According to one federal survey, K-12 and higher education institutions collectively hired 32,000 fewer educators per month over the first six months of the pandemic. That belt-tightening was likely caused by worries that the austerity measures of the last global economic downturn would be repeated, Kraft remarked.

“We had lived through the lessons of the Great Recession, which substantially cut education funding over multiple years and led to hundreds of thousands of teachers being laid off,” he said. “So schools were cautious, and I think rightly so, about filling positions even from natural turnover.”

After the slashed budgets of the 2010s, few if any observers predicted the federal government would allocate nearly $200 billion in pandemic relief to American schools. If that understandable misapprehension guided decisions during the early phases of the crisis, a general absence of accurate, real-time data has further clouded the picture ever since. 

The deficiencies of public data sources are several, Kraft and Bleiberg note. Some surveys don’t clearly differentiate among K-12 employees, such that job additions or attrition among non-instructional staff can be conflated with those affecting teachers. Others make it hard to differentiate between public K-12 schools and private institutions (or even colleges and universities). And as with virtually all data regularly collected by the government, figures are subject to serious revisions even months after their initial publication. 

Chad Aldeman is the policy director of Georgetown University’s Edunomics Lab, a research group that studies education finance. In an email to The 74, Aldeman called national teacher employment data “at best a patchwork quilt of federal, state and local databases, much of it several years old.” That disorganization makes it difficult to answer even basic questions, such as how many job openings exist throughout the nation’s K-12 schools and which specific positions principals and superintendents are hiring.

In normal circumstances, that kind of opacity paves the way to misguided policy choices. But at a time of unprecedented tumult in the labor market, it might come at the cost of critical, one-time resources that could otherwise be spent helping students climb back from years of lost learning. Aldeman said he was aware of cases in which districts were poaching from their neighbors, or even cannibalizing their own workforce, to fill specialist roles.

“I don’t think state and federal policymakers are taking these data gaps seriously,” he wrote. “Instead, states seem to be spending their own money blindly, and I don’t see many thoughtful plans to track the spending alongside student outcomes to make sure the increased staffing levels actually translate into better services for students.”

Kraft said that public confusion over the nature of teacher shortages is a serious concern, pointing to showing higher vacancy rates at high-poverty or predominantly minority schools. The difficulties those schools face in hiring, and the increased stress suffered by their staff, are persistent problems that call out redress through higher pay and better working conditions, he argued; misbegotten narratives based on incomplete information could only make them harder to solve.

“We are failing these communities by failing to understand the nature of the problem, Kraft said. “And by failing to understand the nature of the problem, we may well diagnose it incorrectly and prescribe remedies that fail to address the underlying, structural inequities.”