Commentary

The article by Cimpian, Thompson, and Makowski is a wonderful display of the power of combining current analytic methods, present-day educational data sets, practitioner-researcher collaborative arrangements, and key policy questions pertaining to the educational outcomes of English learners (ELs). I would like to put the work in historical context to appreciate its significance and imagine this line of work going forward.

The year 2064 will mark the centennial of the Civil Rights Act—it will be 48 years after publication of this special AERJ issue. Between now and then, there will be several overhauls of the Elementary and Secondary Education Act (ESEA), which has been named the Improving America’s Schools Act of 1994, No Child Left Behind Act of 2001, and currently the Every Student Succeeds Act of 2015. The policy has included the idea of supplemental programs, separating or combining different funding streams (the “supplement-not-supplant” issue), bilingual education, introduction of state standards (including English language proficiency standards), assessment, and accountability frameworks. ELs are served through every title of the law, with the specific provision addressing their needs having been known as the Bilingual Education Act (Title VII) in its early years and subsequently turned into a focus on English language acquisition through Title III. What will a successful evolution of the law look like in the course of the next half-century?

The first thought is to look at the constants that we can expect. One constant is the civil rights framework. ESEA was born in the civil rights era, and the law has remained the bedrock of federal supports for the educational rights of students along with the major Supreme Court cases—Brown v. Board of 1954 and Lau v. Nichols of 1974. Unless we believe that the civil rights of immigrant students and children of immigrants will no longer be an issue in 2064, this will be a constant in future ESEA reauthorizations. Another constant is the continued diversification of the population along linguistic lines as well as the centrifugal force of English. A third constant are the laws of human learning—the reality that it takes most ELs four to seven years to attain reclassification status (Hakuta, Goto-Butler, & Witt, 2001; Thompson, 2015).

Looking back over history, the very idea of reclassification combines many critical considerations within the historical flow and the constants just mentioned. For one thing, the idea sits at a policy boundary between students who are “served” or “not served” by EL-specific programs. These programs, by law (such as through Lau v. Nichols as well as interpretations of the Equal Education Opportunities Act of 1974) define “appropriate” programs that are based on the so-called “Castañeda Standards” ¹ that a program should (a) be based on sound educational theory, (b) be implemented adequately, and (c) demonstrate effectiveness over time. Reclassification tries to regulate between students receiving or not receiving services through such programs.

Second, laws and policies governing ELs have long struggled with the idea that a student’s status is temporary. By nature, development is long term and conditioned on a variety of individual and socioeconomic variables. Time to reclassification should incorporate the expectation that it occurs over a long period of time and meshes poorly with policy expectations that are short term with annual accountability expectations that retainrelatively little of the longitudinal trajectory of student growth. Reclassification by its very nature is a long-term, cumulative process—much like graduation—and therefore is meaningful as an ultimate outcome in spite of its messiness.

This article takes advantage of the messiness (i.e., variability) around reclassification policies in states and districts and beautifully applies multisite regression discontinuity designs to model the effect of reclassification at thelevels of states and districts. It also looks under the hood around some of the specific data patterns using qualitative methodology and demonstrates the power of combined methods. This article’s important message is that policies and implementation (through practice) matter at the state and local levels in ways that demonstrably affect the probability of graduation for individual students.

Looking forward, it is clear that this study addresses the main constants around ELs and that these constants will remain issues as ESEA moves into the future. The article alludes to one important change in ESSA around standardized statewide entry and exit criteria. How this policy is implemented and evaluated will be a wonderful topic for future investigations, using this method of analysis across a sampling of 50 states as they adopt and adapt the policy over the next 10 years.

A second line comes from the meaning that policymakers and practitioners make of the inclusion of reclassified ELs in subgroup accountability for a period of up to four years. This specification of four years was a legislative compromise between focusing on only current ELs and focusing on any student who has ever been an EL. Understanding this shift, particularly in how it gets incorporated into new models of accountability that focus on improvement for all schools (rather than just the lowest performing schools), would make an important contribution toward future reauthorizations as they consider this policy parameter.

A third line might be in what to make of the blind spot created by the problem of the minimum N-size in accountability and reporting (decided by the state). The risk is that ELs in schools below the minimum N-size will fall “below the radar” and their needs will not be addressed. It turns out that the N-size problem shows up even in this article, where districts with fewer than 10 EL students on either side of the discontinuity were removed from the analysis. There are understandable reasons (student privacy in the case of reporting as well as statistical reliability) why these schools and districts are dropped, but the science would need to create ways of addressing small Ns, much as special education has created the field of single-subject designs. This could also be an arena where quantitative and qualitative research come together.

More generally, the field of EL research and practice is blessed by the existence of the Castañeda standards that call for continuous improvement using evidence from implementation and effectiveness (Hakuta, 2011). The research community should take advantage of this situation to continue to apply this model to the programs and policies shaped by the multiple layers of federal, state, local, schools, and classrooms into the next century.