Is grade retention a gateway or gatekeeper for English learners?

Achieving reading proficiency by the end of third grade is a pivotal benchmark in students’ academic journey, marking a transition from “learning to read to reading to learn.” This shift is necessary for students to prepare for more challenging material in later elementary grades and is positively associated with high school graduation.

For most multilingual students classified as English learners (ML-ELs), reaching this benchmark on time is challenging. According to the most recent NAEP results, just 10% of fourth-grade ML-ELs scored Proficient or Advanced in reading, compared to 37% of their non-English learner peers. This gap is not surprising, but it reflects opportunity gaps these students face: ML-ELs often attend underresourced schools, are taught by less-experienced teachers, and have limited access to rigorous instruction. Further, ML-ELs are assessed using standardized tests designed for native English speakers; these tests fail to consider ML-ELs’ unique circumstances, and ML-ELs often do not receive effective accommodations for them.   

To improve early literacy, many states have adopted third-grade retention policies if students fail to meet literacy benchmarks on their third-grade assessments. In theory, an extra year for struggling readers would help them to build foundational skills before moving on to more demanding curriculum. However, it could also harm students emotionally and socially. Research shows that third-grade retention, when paired with supporting services, has a positive effect on student achievement in the years or grades immediately after retention, but the advantage typically fades out over time.

Most existing studies on third-grade reading retention policies focus on the average effects on all students; however, the consequences of these policies on ML-ELs are rarely explored and are not well understood. This post examines how these policies may impact ML-ELs, a particularly vulnerable student population, and summarizes findings from my new study that evaluates both short- and long-term outcomes of a third-grade retention policy in Texas.  

Should we expect grade retention to help or harm English Learners?

The decision to retain any struggling student in third grade comes with a set of possible tradeoffs, and these tradeoffs only get more complex for ML-ELs. On one hand, an additional year may act as a gateway to future learning opportunities for ML-ELs, providing  them space to further develop English fluency, catch up in reading, and attain language reclassification sooner. Even more, retention paired with target support, such as small-group tutoring sessions or specialized help from a qualified bilingual educator may lead to greater academic gains. Consequently, this may set them up for accessing rigorous core content in the future with ease.

On the other hand, retention separates students from their peers. ML-ELs already face integration challenges, and this choice could harm them socially and emotionally. This could lead to further disengagement from school and, because retained students reach the legal age out of compulsory education sooner than promoted students, they are at higher risk of dropping out sooner. This possibility is particularly concerning for ML-ELs, who already graduate from high school at lower rates. Additionally, grade retention may reinforce teachers’ low expectations of ML-ELs, which could translate into tracking them toward a less rigorous curriculum and diminished college and career opportunities after high school.

Texas’ third-grade retention policy: 2003 to 2009

To better understand how these choices impact ML-ELs, I looked to Texas, which was an early adopter of a test-based reading retention policy and has a large population of ML-ELs. I focused on students who took their third-grade tests between 2003 to 2009. I then used the state’s longitudinal data to track students over time and across different state agency records, providing an opportunity to evaluate many outcomes ranging from short-term test score gains to long-term outcomes into young adulthood like earnings.

Texas’ test-score-based policy enables a regression discontinuity (RD) approach to estimate the causal effects of grade retention.¹ The RD design compares students who scored just below to those who scored just above the threshold for grade promotion. For these students, passing or failing is as good as random assignment. Yet, these estimates cannot be generalized to a broader population of all ML-ELs; they are only valid for those students whose scores in the reading test are around the threshold.²

Finally, Texas’ retention policy required schools to provide accelerated instruction to students who were retained after failing the state reading test. Schools had flexibility on how to implement their accelerated instruction program, but by law, it was required to be in small classes (10 students per teacher) and taught by a certified teacher at least once a week. My results, accordingly, capture the combined effect of retention and additional reading support.

Grade retention plus additional reading support improve short-term outcomes, but the effects fade over time

I estimated effects on a host of both short-term and long-term student outcomes.³ The third-grade retention policy—which included the instructional support for retained students—clearly helped improve all the short-term outcomes. For example, in comparison to similar students who were promoted, marginally retained students scored 0.94 SD higher on their fourth-grade reading test (when they took it one year later than their original classmates). This gain is equivalent to moving from the 14th percentile to the 36th percentile of the test score distribution. By fifth grade, the effect is still important but decreased to 0.76 SD. I also estimated a positive effect on students’ probability of exiting English learner status. Marginally retained students were 11 percentage points more likely to be reclassified sometime in grades 4 to 11 than marginally promoted students.

Despite the large gains in the short term, the positive effects do not extend to and beyond high school. All the long-term outcomes I estimated showed no statistically significant differences between marginally retained or promoted students (see summary table of results below). I find no effect of supplemental services and grade retention on students’ probability of ever graduating from high school; formally dropping out of high school; enrolling in a Texas institute of higher education within three years of expected graduation; or their earnings in their early 20s.

School resources matter for retained students

Additional instructional support can play a critical role in boosting struggling students’ learning opportunities, and this appears to be true for supporting ML-ELs, too. Though I lack information on the specific supports schools offered to retained students, I do have information on school expenditures on accelerated instruction supports (e.g., small classes for students who failed the reading test) and on bilingual education.

To explore the role of these school resources, I conducted a subgroup analysis in which schools are divided into two groups: schools with high per-pupil spending in accelerated education or bilingual education and schools with low spending on those items. The results showed that the positive effects of retention on reclassification and high school completion were higher in schools that spent more money on these supports. Though these results are not causal, they suggest that schools’ instructional supports play an important role in continuing to influence the gains for ML-ELs beyond the retention decision alone.

Working to improve support for ML-ELs

The evidence from this study indicates that third-grade retention, when paired with supplemental reading instruction, can improve ML-EL students’ math and reading achievement while also helping them exit the EL status sooner. However, retention decisions do not appear to make a significant difference on longer-term outcomes.  

In light of the finding that additional learning time benefits ML-ELs’ academic outcomes, states and districts could consider strategies that extend learning time without the social or emotional costs of being held back. For example, summer school could help prevent summer learning loss, which is greater among ML-ELs. Similarly, an extended school day with targeted instruction for struggling ML-ELs may be a promising strategy.

Additionally, there may also be value in expanding interventions for ML-ELs. In Texas, ML-ELs who barely passed the reading test did not receive supplemental services but likely could have benefited from them, too. Increased investment in bilingual education and remedial programs can provide sustained support for all ML-EL students and help them close opportunity gaps more effectively.

Finally, I echo other experts and scholars suggesting that before retaining students, schools and teachers need to understand the risks, not just potential short-term benefits. This could be particularly true for ML-ELs, who have experienced far greater setbacks due to the COVID-19 pandemic.