Screen or Barrier? Science of Reading Certification Exams and Teacher Supply in Texas

• In Event: The Teacher Pipeline Under Pressure: New Evidence on Interest, Certification, Hiring, and Retention

Abstract

Whether to require prospective teachers to pass high-stakes certification exams is a contentious question in K-12 education policy. Some states have eliminated certification testing in response to teacher shortages and equity concerns, while others have expanded requirements, particularly in early literacy, where 20 states now mandate a standalone Science of Reading (SOR) exam. This divergence has produced a policy debate that often presents a false choice between maintaining standards and ensuring an adequate supply of diverse teachers. This paper examines the effects of Texas's SOR certification exam, mandated in 2021 for new elementary and middle school teaching candidates, and asks whether structured preparation can mitigate the barriers these exams create.

Using administrative data on certification exams, teacher employment, and student achievement, I combine multiple empirical approaches to evaluate the impact of the requirement on teacher supply. Regression discontinuity estimates show that failing the SOR exam delays Standard certification by 8.5 percentage points within six months and reduces the probability of teaching the following year by 6.5 percentage points. These effects are not evenly distributed. Certification delays are concentrated among non-white candidates, who experience delays nearly four times larger than white candidates. Employment consequences fall most heavily on alternatively certified candidates already teaching in classrooms, for whom failing reduces the probability of teaching the following October by over 10 percentage points. At the market level, the share of newly issued certifications in affected fields declined by 5 to 19 percent after the mandate, with no recovery over four years. Survival analysis confirms this reflects a genuine slowdown in certification rather than substitution toward other fields.

At the same time, value-added models show that SOR exam scores predict student reading achievement more strongly than scores on the general pedagogy exam, suggesting the test captures instructionally relevant knowledge. However, the pass-fail threshold does not separate more effective teachers from less effective ones, meaning the supply costs fall on candidates who would perform comparably in the classroom.

Rather than accepting this tradeoff, I provide early evidence of a third way. Event study analyses show that Educator Preparation Programs offering state-sponsored SOR Academy coursework saw improved candidate pass rates and exam scores, with no corresponding changes on an unrelated certification exam. These findings suggest that states can pair new certification requirements with targeted preparation support rather than choosing between quality and access.

Author

• Andrew Avitabile, University of Virginia
  Presenting Author