Texas’ reading licensure test looks like both a quality screen and a hiring bottleneck

A new October 2026 working paper on Texas’ Science of Teaching Reading exam argues that the state’s standalone reading licensure test is doing two things at once: keeping some teacher candidates out of classrooms and, at the same time, identifying candidates whose students later post stronger reading results. For state leaders and educator-preparation programs, that shifts the practical question from whether to scrap the exam to how to reduce preventable failure without weakening a standard that appears to predict classroom effectiveness. (edworkingpapers.com)

Texas has required the Science of Teaching Reading, or STR, exam since January 1, 2021, for five certification fields tied to early literacy and middle-grades reading instruction. The test is not a small add-on: the official exam page describes a separate assessment with 90 selected-response items, one constructed-response item, a passing score of 240, a five-hour appointment window, and a $136 fee each time a candidate sits for it. In other words, the STR functions as a genuine gate in the certification process, not just another module folded into a broader elementary test. (tea.texas.gov)

That gate is the focus of Andrew Avitabile’s new paper, posted through EdWorkingPapers and last updated October 5, 2026. EdWorkingPapers is a prepublication research series designed for rapid dissemination, which means readers should treat the study as important new evidence, but still as a working paper rather than the final word. Even so, the study matters because it links exam attempts, certification records, employment, and student achievement data in one state system, allowing a much fuller look at what happens after candidates pass or miss the cutoff. (edworkingpapers.com)

What the study found

Avitabile uses two main strategies. One compares candidates right around the passing threshold, essentially asking what changes when a candidate lands just below rather than just above the cutoff. The other compares certification fields that were affected by the new exam with fields in the same preparation programs that were not, before and after the 2021 rollout. That combination does not answer every question, but it gives the paper more leverage than a simple statewide trend line or pass-rate snapshot would. (edworkingpapers.com)

On the supply side, the results are hard to ignore. Narrowly failing the STR lowered the probability of teaching the following year by 6.6 percentage points, and the effect was concentrated among alternatively certified candidates. Looking more broadly at program output, the paper finds that certifications in the fields touched by the STR jumped ahead of implementation, then fell after the requirement took effect and had still not recovered five years later. That is the core pipeline warning in the paper: a higher bar can change who reaches classrooms, but it can also reduce how many do. (edworkingpapers.com)

On the quality side, the paper finds something many licensure tests never clearly show. STR scores predicted student reading achievement more strongly than candidates’ scores on other certification exams, and students taught by teachers who passed the STR on the first attempt scored 0.024 standard deviations higher in reading than students of teachers who had failed. That does not mean the test captures every trait that matters for reading instruction, or that every candidate who needs a retake will be a weak teacher. It does mean the exam appears to contain real information about later teaching effectiveness, which makes lowering the cut score a riskier policy move than critics of testing sometimes assume. (edworkingpapers.com)

Why this matters beyond one Texas test

The paper notes that 20 states now require prospective teachers to pass a standalone science-of-reading exam. Texas therefore looks less like an outlier than an early case study in a broader policy shift: states want more evidence that new elementary teachers understand how reading develops and how to teach it. What Texas adds to that conversation is evidence that a reading-specific test can act as both a screen and a barrier at the same time. (edworkingpapers.com)

That dual role matters because Texas is already navigating a fragile teacher pipeline. In an April 2026 workforce report, the Texas Education Agency said alternative certification has been significantly replaced by uncertified hires and that uncertified teachers still make up the largest share of new hires who are new to teaching. The same report says the gap between people admitted to educator-preparation programs and those who actually obtain certification has widened. Put alongside Avitabile’s finding that the STR’s workforce effect is strongest for alternatively certified candidates, the implication is uncomfortable but useful: if the state keeps a predictive exam without improving preparation, some districts may not get a stronger certified pipeline so much as a thinner one, while continuing to rely on less stable staffing routes. That last point is analysis, but it follows directly from the paper and TEA’s own workforce data. (tea.texas.gov)

The content of the exam helps explain why it may predict later results. Texas’ official framework says the STR assesses reading pedagogy, foundational skills, comprehension, and candidates’ ability to analyze student reading needs and respond instructionally. It also explicitly grounds the assessment in scientifically based reading research and the state’s English language arts standards. This is not merely a phonics compliance check; it is meant to test whether a novice teacher can connect reading science to instructional decisions. If the exam is measuring that reasonably well, its predictive value is less surprising. (teadev.tea.texas.gov)

The more useful policy lever is preparation

The paper’s most actionable finding may be the least flashy one: structured science-of-reading coursework in preparation programs raised first-attempt pass rates by about 4 percentage points. That points toward a more productive policy response than either celebrating the test or abolishing it. Texas and other states can keep the standard while changing when and how candidates encounter it: align coursework earlier, treat the STR as a capstone to taught content rather than a surprise hurdle, and make practice with the constructed-response portion routine before candidates ever pay for a test appointment. Texas already points candidates to preparation materials, including sample questions and a $10 interactive practice exam, and TEA has also published support resources for educator-preparation programs as state standards shift toward stronger cognitive-science expectations. The evidence in the paper suggests those kinds of upstream supports matter more than simply arguing over the passing bar. (edworkingpapers.com)

There are still real limits here. The study is Texas-specific, and the cleanest causal estimate on supply comes from candidates near the passing threshold, not from every teacher candidate who ever fails the exam. The paper is also a working paper, not yet a peer-reviewed journal article. Policymakers in other states should therefore resist easy copy-and-paste conclusions, especially where cut scores, prep-program quality, or district hiring pressures look different. But they should also resist the opposite mistake of treating all licensure tests as equally empty signals. This paper suggests the STR is not empty. (edworkingpapers.com)

The next thing worth watching is whether states use this kind of evidence to redesign preparation rather than weaken expectations. In Texas, that conversation may get sharper once the state’s new Teacher Vacancy Collection begins yielding statewide position-level data; TEA says the first submission, covering vacancies from the 2026-27 school year, will be due September 23, 2027. When those data arrive, the most important question may no longer be whether science-of-reading tests reduce supply. It may be whether states learned to build the preparation systems that let strong candidates clear the bar on the first try. (tea.texas.gov)