How Can We Improve Special Education Without Asking Uncomfortable Questions?

This piece is a response to an article by Laura A. Schifter and Thomas Hehir, available here, which was itself a response to “Has Inclusion Gone Too Far?” (features, Fall 2018) by Alison Gilmour.

Laura Schifter and Thomas Hehir share with me a desire to create better educational opportunities for students with disabilities. We agree that research should identify policies and practices that lead to effective instruction for all students. Their article, unfortunately, undermines efforts to conduct this necessary work. By (1) continuing to focus on the location where students with disabilities are educated rather than the services they receive and (2) rejecting the need to study the intended and unintended consequences of special education policies, they tacitly accept the current state of educating students with disabilities.

Location Over Services

Schifter and Hehir comprehensively review the policies and court cases that have established the preference for educating students with disabilities in general education settings. The general education setting is the least restrictive environment for most students with disabilities, and the push for educating students with disabilities in their neighborhood schools and in general education classrooms is an important extension of the civil rights movement. This history should not be ignored, and the hard work to enact policies that make schools and classrooms accessible to all students should be applauded.

At the same time, Schifter and Hehir conveniently ignore the requirement that students with disabilities receive an appropriate, and individualized, education. An appropriate education means that students are receiving the services and interventions that they need in order to make progress in the general education curriculum. They note that reporting requirements under IDEA consider placement in a general education classroom for 80% or more of the day as a fully inclusive placement and contend that this placement allows for “sufficient time to provide the kinds of effective interventions” that I cited. However, they do not discuss that recent data and research both demonstrate that the interventions and services currently provided to students with disabilities are not widely successful. The 2017 National Assessment of Educational Progress shows that students with disabilities are performing far below their peers in reading and math: 11% of fourth grade students with disabilities scored at or above proficient in reading compared to 40% of their peers, and 15% of fourth grade students with disabilities scored at or above proficient in math compared to 43% of their peers. The low achievement of students with disabilities reflects the failure of these students to gain access to an appropriate education, that is the services and interventions they need to make progress, regardless of setting. Indeed observational studies of the instruction that students with disabilities receive typically find that students with disabilities are not educated using effective instructional approaches or interventions, independent of the setting in which they are educated.

While they acknowledge that settings and services are not the same, Schifter and Hehir continue to defend causal interpretations of the finding that students with disabilities have better outcomes when they spend more time in general education settings. They go so far as to compare these findings to research on smoking and lung cancer, correlational research that is accepted as causal. For correlational evidence to be accepted as causal it must meet three conditions: temporal precedence (i.e., the independent variable, such as smoking, comes before the dependent variable, such as cancer), covariation of the independent and dependent variable (i.e., when one variable changes, the other variable changes), and ruling out alternative explanations (i.e., eliminating other variables that are related to changes in the independent variable and the dependent variable). Research on tobacco use meets these conditions. Research regarding time spent in a general education classroom does not. As the authors point out, a myriad of contextual variables are associated with placement decisions. These same variables are often associated with students’ outcomes. Thus the condition of ruling out alternative explanations is not met. Policies and interventions aimed at increasing time in a general education classroom are unlikely to dramatically change the outcomes of students receiving special education services. Yet many researchers studying special education still use setting as an outcome and assume that changes in setting will translate into changes in outcomes.

Asking Uncomfortable Questions About Intended and Unintended Consequences

More problematic is Schifter and Hehir’s presumably unintentional attempt to dissuade researchers from examining both the intended and unintended results of special education policies. Policies often reflect societal values. Then the work of implementation is left to state and district policymakers, school leaders, and teachers. As Schifter and Hehir point out with the example of identification rates in Texas, the implementation of a policy may break down in ways that do not align with the letter or the spirit of the law. In these situations, researchers must ask questions about the intended and unintended consequences of policies in order to improve policies and inform their implementation.

Examining if and how students with disabilities influence their peers and teachers, and if and how peers and teachers influence students with disabilities, is a necessary step to understanding the current state of inclusive education and what works, for whom, and under what conditions. This line of research may result in findings that make people uncomfortable, such as those linking students with emotional/behavioral disorders to teacher turnover or classmates’ absences. Schifter and Hehir rightly worry that these findings could be used by critics of inclusion to undermine existing policies. The results could also be used to encourage decisions about services and placement that are not based on the individualized needs of students. Despite these concerns, such research is essential to drive implementation changes that help fulfill the intended goals of special education policies.

Improving Special Education and Inclusive Education

Improving existing systems is challenging work that involves questioning the status quo and acknowledging that those charged with implementing well-intentioned policies may need more support to realize the policies’ goals. Descriptive research focusing on the current state of special education and the appropriateness of the education that students with disabilities receive informs future improvements to implementation. Experimental work builds on this knowledge to create effective classrooms. For example, research by Erik Carter and his colleagues focuses on training paraprofessionals and students’ peers to increase the participation of students with significant disabilities in general education classrooms. Jade Wexler and her colleagues are studying a professional development program that trains general education and special education teachers to effectively provide literacy instruction embedded in general education content area classes to middle school students with reading disabilities. These researchers and many others reject the assumption that current practices in inclusion are widely effective, and their work ultimately improves the implementation of special education policies.

Schifter and Hehir seem to assume that my answer to the question posed by my article’s title—”Has Inclusion Gone too Far?”—is yes. This assumption is inaccurate. Inclusion has gone too far only in that it is being implemented without attending to the reality that many schools and teachers struggle to deliver effective services to all students and that more research is needed to identify programs and interventions to support teachers in providing instruction to students with disabilities, regardless of setting. Accepting the low achievement levels of students with disabilities leads to the replication of existing problems and inequalities that undermine diversity in classrooms and schools. Students with disabilities are not the problem. The problem is the failure to critically examine what works and what does not work in order to design educational systems that recognize students’ individual needs and result in success for all students.

Allison F. Gilmour is assistant professor of special education at Temple University.

Last Updated


Notify Me When Education Next

Posts a Big Story

Business + Editorial Office

Program on Education Policy and Governance
Harvard Kennedy School
79 JFK Street, Cambridge, MA 02138
Phone (617) 496-5488
Fax (617) 496-4428

For subscription service to the printed journal
Phone (617) 496-5488

Copyright © 2024 President & Fellows of Harvard College