California’s Academic Criterion for English Learner Reclassification: Implementation and Impacts

Introduction and Policy Background

EL classification is a legally required status for multilingual public-school students in K–12 schools who meet federal requirements for language support services. By law, the EL category is designed to be temporary. As students develop proficiency in English and no longer need or benefit from supplemental language supports, students are exited from the category. The process of exiting EL status is called reclassification and, in California, has been governed by a combination of federal, state, and local rules.

Federal law requires that states use standardized statewide procedures for entry and exit into the EL category. At a minimum, those procedures must include an annual assessment of each EL-classified student’s ELP using an assessment that measures English speaking, listening, reading, and writing skills.Office for Civil Rights. (2015). Every Student Succeeds Act. U.S. Department of Education. https://www.ed.gov/laws-and-policy/laws-preschool-grade-12-education/every-student-succeeds-act-essa Federal law leaves the selection of the ELP assessment, the setting of one or more thresholds on that assessment, and any additional rules related to reclassification up to individual states. Most U.S. states (33 states) use only their ELP exam to determine reclassification eligibility. Across these states, some require a single overall ELP score as the sole reclassification criterion (24 states), while others require that students meet thresholds on specific proficiency domains (reading, writing, speaking, and listening) within that assessment (9 states).Morales, C., & Lepper, A. (2024). The changing landscape of states’ English learner reclassification policies. Center for Education Efficacy, Excellence, and Equity (E4), Northwestern University. https://e4.northwestern.edu/2024/04/25/the-changing-landscape-of-states-english-learner-reclassification-policies/ Until recently, seven states, including California, Florida, and Texas, did not limit reclassification criteria to the state ELP assessment. Instead, they required additional criteria such as academic criteria (academic test scores or proficiency levels and/or course grades or grade point averages) and/or teacher, parent, or administrator approval. California recently changed its laws around this.

California had been alone in requiring four separate reclassification criteria areas, areas that have been in place since 1976. Specifically, California state statute had required (1) that a student receives a minimum summative proficiency level of 4 (“proficient”) on the state’s ELP assessment, the English Language Proficiency Assessments for California (ELPAC); (2) teacher evaluation, including, but not limited to, an evaluation of a student’s curricular mastery, an evaluation that is locally determined but that can involve a tool developed by the state called the Observation Protocol for Teachers of English Learners; (3) parental opinion and consultation as per a locally determined process; and (4) a locally determined measure of a student’s basic skills in ELA relative to English-proficient students. These criteria were additive, not substitutive, meaning that a student was required to meet all four criteria to be eligible for reclassification. Also noteworthy, three of the four criteria (criteria 2 through 4) were locally determined by individual school districts.Hill, L., Lee, A., & Hayes, J. (2021). Surveying the landscape of California’s English learner reclassification policy. Public Policy Institute of California. https://www.ppic.org/publication/surveying-the-landscape-of-californias-english-learner-reclassification-policy/

Our research was conducted prior to the passage of AB 2555 and focused specifically on the 4th criterion, the basic skills criterion. The California Department of Education (CDE) described this criterion as:

Comparison of the performance of the pupil in basic skills against an empirically established range of performance in basic skills based upon the performance of English proficient pupils of the same age, which demonstrates whether the pupil is sufficiently proficient in English to participate effectively in a curriculum designed for pupils of the same age whose native language is English.California Department of Education. (2025). Reclassification. https://www.cde.ca.gov/sp/ml/reclassification.asp

To understand how the basic skills criterion operated in California and how EL students fared on outcomes based on this criterion, we examined (1) how districts designed and operationalized the criterion, (2) the extent to which the criterion posed a reclassification barrier to students and for whom, and (3) how being held in EL status due to the academic criterion was related to students’ academic and behavioral outcomes.

Study Design: Data and Methods

Via web searches and personal communications, we collected reclassification criteria by grade for a stratified random sample of California school districts, which were selected based on their setting type and the relative size of their EL population. We combined these criteria data with student-level data from CDE for the years 2021–22 through 2023–24. We limited our sample to districts that used the Smarter Balanced Assessment Consortium (SBAC) ELA test for their basic skills reclassification criterion (and as described below, districts may have used more than just the SBAC to determine basic skills eligibility).We initially sampled 120 school districts at random but retained only districts that used the SBAC to satisfy the state’s academic criterion for reclassification. This restriction was necessary because our analysis required student-level test scores, and CDE data does not include individual student scores for alternative assessments that are used to measure basic skills, such as the i-Ready assessment. Some districts used multiple basic skills assessments. In these cases, we retained the district if the SBAC ELA test was one of the assessments used. Our final sample consisted of 76 districts serving just over 112,000 student-year observations, representing about 10% of the state. The demographic characteristics of our sample closely matched those of the full state.

Among the 76 districts, we evaluated what basic skills tests they used, how they used those tests, and where they set their SBAC basic skills threshold(s). We then examined the extent to which the basic skills criterion was creating barriers to reclassification for students by assessing the likelihood of students passing the criterion, how that varied by student and district characteristics, and how passage rates for the academic criterion compared to those for the ELP reclassification criterion (the only other test-based reclassification criterion). Finally, to shed light on the effect of being held in EL status due to the academic criterion, we conducted a fuzzy regression discontinuity analysis. In effect, this analysis compared the next-year academic and behavioral outcomes of students who met the state ELP reclassification threshold (ELPAC level 4) and were either just above or just below their district’s academic criterion.

Findings

How Did Districts Set Their Basic Skills Criterion?

Districts in our sample varied widely in terms of whether they paired SBAC with any other basic skills ELA assessments (see Table 1), falling into the following groups:

• “SBAC or” districts: The largest group of districts by far (80.2%) operationalized their basic skills criteria such that, in any given year, a student could meet one of a menu of two or more ELA assessment thresholds. More than 85% of sampled students were in this category.
   
• “SBAC only” districts: 14.2% of districts (serving 9.7% of students) required that students met a district SBAC threshold and did not include any other assessments in their basic skills criterion. In these districts, the annual SBAC ELA test functioned as a one-time, high-stakes test, determining students’ reclassification eligibility.
   
• “SBAC and” districts: 5.8% of districts (serving 5.2% of the student sample) required that a given student meet the SBAC district threshold and a minimum of one additional ELA test. This was the most challenging category, with multiple high-stakes ELA tests annually.

Among the “SBAC or” and “SBAC and” districts, there was an array of alternative assessments. Common assessments included the i-Ready, Star Reading, and MAP Growth assessments.

Districts also varied in difficulty in terms of the SBAC scale score or proficiency level they required for basic skills criterion eligibility, falling into one of three, somewhat equivalently sized categories (see Table 1). Comprising the largest group of our sample, 45.1% of districts set their proficiency level at “standard met” (level 3). These were, on average, smaller districts, since they served only 31.8% of the student sample. Another 32.1% of districts, serving 40.3% of students, set their threshold at “standard nearly met” (level 2). Finally, 22.8% of districts, serving 27.9% of students, set their threshold somewhere between levels 2 and 3. No districts in our sample set their threshold above level 3 or below level 2.

State statute at the time specified that the basic skills reclassification criterion should be set relative to the performance of same-age native English-speaking peers. “Peers” can be interpreted in multiple ways. For example, an EL-classified student’s score could be compared to the state mean of same-age non-ELs, the district or school mean of same-age non-ELs, or otherwise comparable students such as non-EL peers who share similar demographic characteristics to the student. We evaluated several approaches to estimating the ELA performance of plausible comparison groups and concluded that an appropriate threshold to reflect the standard in the state code might be an SBAC proficiency level of 2.2 to 2.3.At the state level, the average SBAC ELA proficiency level of non-EL students is 2.53, indicating that the 32.1% of districts that required a level 2 may have been considerably under that average, while the 45.1% of districts that set a level 3 were considerably above the mean. We also considered the average score of non-EL Latino/a students in the state (2.27), since 84.4% of ELs in California are Latino/a. Finally, more than three quarters (78.5%) of EL students live in low-income households that are eligible for free or reduced-price meals. Thus, another appropriate comparison group might be non-EL students who are eligible for free or reduced-price meals. Statewide, these students’ average SBAC score is 2.17. Using this benchmark, we concluded that about a third of districts were set close to, but below, that threshold (level 2); a third had thresholds set considerably above (level 3); and a third were relatively close to that threshold range (between levels 2 and 3).

In effect, the variation between districts on both the setting of thresholds and the operationalization of basic skills criteria meant that where a student attended school could have a large influence on their likelihood of passing the state academic reclassification criterion. Some students were in districts where they had to meet multiple high-stakes assessment thresholds over the course of the year and/or reach an SBAC level well above the state mean, while others were in districts where they had multiple opportunities to meet the criterion and/or the threshold was set below the state mean. Overall, however, we calculated that 89% of students were in a district that was either an “SBAC or” district or an SBAC level 2 district (or both), suggesting that most districts set their academic criterion with an eye toward enabling student passage.

To What Extent Did District Academic Criteria Pose a Barrier to Reclassification?

Our study also aimed to understand whether the basic skills criterion posed a barrier to reclassification by holding students in EL status after they reached the state English proficiency threshold (ELPAC level 4). We found that, for most students, it did not. Across our sample, more students met their district’s basic skills reclassification criterion (as measured by SBAC scores) than met ELPAC level 4. This indicates that the basic skills criterion was more attainable than the English proficiency criterion for most students.Because CDE data does not include ELA scores aside from the SBAC ELA, we were able to proxy basic skills criterion passage through the SBAC alone. We likely undercounted criterion passage, since most students in our sample (85%) were in districts that allowed one or more non-SBAC test scores to substitute for the SBAC requirement. Specifically, across the 2 years examined (2021–22 and 2022–23), 26.6% of EL-classified students cleared their SBAC threshold compared to 21.1% of students who cleared the ELPAC criterion.

While the basic skills requirement did not pose a barrier to students in general, there were important settings and subgroups in which the academic criterion frequently held students back:

• Middle and high school students: Passing the basic skills threshold was more difficult as students advanced in grade level. Students in elementary grades were more likely to pass the basic skills than the ELP criterion, but students in grades 7–8 were more likely to be held back by the SBAC academic criterion than by the ELPAC English proficiency criterion (see Figure 1).

• Students in rural settings: For students attending school in rural districts, the basic skills criterion was more difficult than the ELP criterion (13.4% passing the SBAC compared to 18.1% passing the ELPAC; see Figure 2).
   
• Districts with higher SBAC thresholds: The basic skills criterion was difficult to clear for students attending school in districts with higher SBAC thresholds. For example, in districts that set their SBAC threshold to proficiency level 3, only 17.4% of students cleared the threshold compared to the 23.3% who cleared ELPAC level 4 (see Figure 2).

As a result, there were large numbers of students who demonstrated English proficiency on the ELPAC but were not reclassified. We calculated that among the 23,693 EL students in our sample who scored ELPAC level 4 in a given grade, 4,684 had not exited EL status by the subsequent grade/year (19.7%).We also estimated the proportion of students across the full state who scored an ELPAC level 4 but were not reclassified in the subsequent grade. That proportion, 18.0%, is similar to what we found (19.7%) in our sample.

We found limited alignment between ELP and basic skills criteria passage. Across the sample, more than 20% of students met each of the two main reclassification criteria, but only 12.7% of students met both in the same year. This means that nearly half of the students who met SBAC did not meet ELPAC and vice versa.The ELPAC proficiency threshold was originally established at the performance level where the distribution of EL students’ ELA achievement most closely resembled the distribution among non-EL students statewide. This alignment was recently re-evaluated and confirmed using more recent data (Faulkner-Bond, 2026). Our analyses did not evaluate that threshold. Instead, we examined only students who already scored “proficient” on the ELPAC (level 4) and compared those who did and did not meet their district’s SBAC-based reclassification criterion.

Reclassification mapped onto eligibility fairly well—but not perfectly. Across our sample, 91.5% of students who met district basic skills and ELP thresholds were reclassified the subsequent year. Adherence to the ELPAC threshold was high; only 10.5% of students who met the SBAC but not the ELPAC threshold were reclassified. Among districts that allowed alternative basic skills assessments (“SBAC or” districts), it was not surprising that 63.4% of students were reclassified if they met the ELPAC threshold but not the SBAC threshold. We expect that many of these students met alternative basic skills assessment thresholds that we were unable to observe in state data. These patterns were quite stable across subgroups and contexts.

Did Remaining in English Learner Status Based on the Academic Criterion Impact Student Outcomes?

Employing a fuzzy frontier regression discontinuity design,Our main models employed triangular kernels, local linear fits, and robust and clustered standard errors for all inferences. Models contained a set of student level covariates and year and district fixed effects. we found that students who were held in EL status did no better or worse with regard to their next-year academic achievement, on average, than their peers who were reclassified (see Figure 3). While statistically nonsignificant, two trends stand out. First, there was a moderate, nonsignificant, positive impact of being held in EL status on students’ next-year SBAC ELA scores. Among English-proficient EL students near the SBAC threshold, those who scored just below the SBAC cutoff (and therefore were held in EL status) saw more growth in their SBAC ELA scores in the next year than students who scored just above the SBAC academic criterion and were reclassified. For the average student, this would reflect an increase from the 50th to the 55th percentile in the distribution of SBAC ELA scores. Subgroup analysis suggests that this pattern was mostly driven by students in elementary grades.

The other trend relates to absenteeism. We created an indicator for missing 10 or more days in the school year. Students who miss 10 or more days are at elevated risk of chronic absence and may be appropriate targets for early intervention. The magnitude and direction of the results, while not statistically significant, were consistent with the possibility that holding otherwise-eligible students in EL status based on their SBAC scores results in a large increase (18.6 percentage points) in the likelihood of missing 10 or more days. On average, 38% of students in our sample missed at least 10 days of school, so an 18.6 percentage point increase reflects a 50% increase to nearly 57% of students who missed 10 or more days of school.

The lack of statistically significant effects of being held in EL status might suggest that the academic criterion was working as intended, facilitating a relatively smooth transition from EL to reclassified status. Yet, examining our results more closely, we found something different—namely that the effects of being held in EL status depended on district settings. Specifically, we found that reclassification effects varied not only across districts with different operationalizations, but also among districts with the same operationalization of the basic skills criterion. For example, settings with a more rigorous basic skills criterion showed positive effects of holding students in EL status, while settings with a less rigorous criterion did not (a somewhat counterintuitive finding discussed in our full report). Even among districts with the same operationalization, some districts showed positive effects of holding students in EL status, while others showed negative effects, and the magnitude of effects varied. For example, among districts with SBAC level 2 basic skills thresholds, the estimated effect of being held in EL status on students’ subsequent-year ELA achievement ranged from –0.3 to +0.4 standard deviation units (see Figure 4).

These findings suggest that there was a complex relationship between districts’ reclassification criteria and other district-specific features that may have impacted multilingual students’ educational outcomes. While we lack the data to make comparisons of the EL services that were offered in different districts, we did find that districts with more strenuous academic criteria were different from those with more flexible or lenient criteria in terms of district size, student populations, overall student outcomes, and locale. Those with more strenuous criteria were smaller, served lower proportions of Latino/a students and students who were eligible for free or reduced-price meals, had higher achievement and lower disciplinary rates (suspensions), and were more likely to be in cities and less likely to be in suburbs or towns.

Summary and Policy Implications

This brief examined how a random sample of California school districts operationalized the state’s basic skills reclassification criterion and how that criterion impacted their multilingual students. The results point to three main conclusions and policy implications.

Inconsistency in basic skills criteria across districts creates inequities for California’s multilingual students. California’s locally determined operationalization of the basic skills reclassification criterion resulted in students facing inconsistent opportunities to reclassify depending on where they attended school. Some districts expected students to hit an ELA benchmark on the SBAC that was well above the state mean, while others set the bar considerably lower. Depending on the district, some students were required to meet multiple academic thresholds across the year to be eligible to be reclassified, while others were given multiple chances and only needed to clear one hurdle. Prior research shows that reclassification matters for students’ educational, social, and life outcomes, so these differences in basic skills expectations across districts are likely consequential.

The basic skills criterion holds particular subgroups of students back and may not be necessary. In our sample, one in five students who reached the state English proficiency threshold were not reclassified the subsequent year. Older students, students in rural settings, and students in districts with more rigorous academic criteria were more likely to remain classified as EL students even after demonstrating English proficiency. This is of policy relevance because secondary and rural EL students experience larger achievement gaps compared to other EL students, making them especially vulnerable to potential consequences of delayed reclassification. In addition, while not statistically significant, there was suggestive evidence that being held in EL status may have negatively impacted student attendance across EL subgroups and settings. This possible impact merits future examination, as it aligns with existing evidence that failing to be reclassified negatively impacts student well-being and behavioral outcomes.Lee, M. G., & Soland, J. G. (2022). Does reclassification change how English learners feel about school and themselves? Evidence from a regression discontinuity design. Educational Evaluation and Policy Analysis, 45(1), 27–51. https://doi.org/10.3102/01623737221097419

The variation in reclassification effects even among districts with the same basic skills thresholds suggests that there is no one specific basic skills threshold that would optimize students’ transitions from EL to reclassified status. Further, we see that many districts (66 out of our sampled 76 districts) appeared to be attempting to minimize barriers created by the basic skills criterion: They either offered a relatively low basic skills threshold or provided students with multiple chances to clear the threshold (or both). These patterns draw into question the efficacy of the basic skills criterion in California.

A 2026 state law, AB 2555, addresses the two challenges identified above by establishing statewide reclassification criteria. This law will eliminate local variation in criteria, which creates inequities among districts in language classification procedures and outcomes. Specifically, this law eliminates the existing multiple-criteria reclassification process. Instead, it establishes proficiency on the statewide ELP assessment (or, if applicable, the statewide alternate ELP assessment) as the sole criterion for reclassification as of the 2025–26 school year. Thus, as the findings from this study suggest, AB 2555 limits reclassification criteria to a reliable and standardized measure of ELP, bringing California’s policy into greater alignment with most other U.S. states. Given this major policy change in California, it will be critical to ensure that districts understand these changes and to support implementation.

Smoothing transitions from EL to reclassified status may require a careful look at instruction and supports. There may be a rupture in services at the time of reclassification in some districts and settings. In settings that were characterized by more difficult basic skills expectations, we saw that recently reclassified students did not grow as quickly as their peers who remained in EL services.Ma, M., & Winters, M. A. (2024). Reclassifying English learners [EdWorkingPaper, Issue 25-1234]. Annenberg Institute at Brown University. https://edworkingpapers.com/sites/default/files/ai25-1234.pdf; Umansky, I., Itoh, T., Faulkner-Bond, M., Porter, L., & Tanner-Smith, E. (Under review). English learner reclassification as an educational input: Meta-analytic evidence of the impact of exiting English learner status. This study was unable to look at characteristics of instruction, teacher preparation, or EL program design. However, we observed a rupture at the time of reclassification among otherwise English-proficient and high-performing students (because these were students in high threshold districts). This highlights the importance of smoothing supports for multilingual students at the time of reclassification. To ensure that reclassified students do not experience dramatic shifts in instructional contexts, districts and schools can provide all students, regardless of EL status, with equitable access to engaging, high-quality instruction and language supports. These supports can be tapered according to individual student needs over time. The state may also benefit from following federal guidance to monitor recently reclassified students and adjust supports for these students as needed.

California’s Academic Criterion for English Learner Reclassification: Implementation and Impacts (brief) by Ilana Umansky and Havisha Khurana is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License.

Core operating support for the Learning Policy Institute is provided by the Heising-Simons Foundation, William and Flora Hewlett Foundation, Raikes Foundation, Sandler Foundation, Skyline Foundation, and MacKenzie Scott. We are grateful to them for their generous support. The ideas voiced here are those of the authors and not those of our funders.

Cover photo by Allison Shelley for EDUimages.