The Better Question: How Can We Improve Inclusive Education?

This article is a response to “Has Inclusion Gone Too Far?” (features, Fall 2018) by Alison Gilmour. Gilmour’s response to Schifter and Hehir’s piece, titled “How Can We Improve Special Education without Asking Uncomfortable Questions?” is available here.


In her recent article, Allison Gilmour asks whether policymakers have gone too far in promoting inclusion under the Individuals with Disabilities Education Act (IDEA). She asserts that there is a lack of evidence that inclusion benefits students with disabilities, as well as insufficient research on the impact of including students with disabilities on their non-disabled peers and teachers. While several of Gilmour’s points are true on their face, her assumptions lead to problematic conclusions. In particular, she suggests that researchers should examine whether students with disabilities have negative effects on their peers and teachers in order to determine if students should be placed in inclusive settings and ultimately argues that policymakers should rethink the law’s emphasis on inclusion. In our view, these conclusions are misguided and even dangerous.

Gilmour’s positions are hardly new, but they are important to counter. Her argument is rooted in the following ableist assumptions: 1) Students with disabilities should have to prove success in order to be included in general education; 2) Students with disabilities are distinctly different from students without disabilities; and 3) Inclusion is a type of special education. Additionally, she takes a deficit-approach in which any failure of students with disabilities is attributed to inherent problems within the students—they do not succeed because they cannot succeed—and relies on this argument to justify their removal from general education.

We flip these assumptions, acknowledging: 1) There is inherent value in educating students with different backgrounds and learning needs together; 2) Students with disabilities are not distinctly different from other students, but rather require additional supports, services, interventions, and accommodations in order to make progress in general education; and 3) Inclusion is a placement where students can receive special education services and research-validated interventions such as those Gilmour cites. Above all, we assume that the success or failure of inclusion depends on the quality of instruction rather than the capabilities of the child.

Assumptions about the Least Restrictive Environment

Gilmour notes that “Inclusion did not become the widespread practice it is today because of a robust evidence base that supports its effectiveness. Rather, it is prevalent because of federal laws that establish special rights for SWDs and their parents.” She thereby implies that the least restrictive environment (LRE) provision, which requires students with disabilities be educated “to the maximum extent appropriate” with non-disabled peers, should have been evidence-based—that is, that students with disabilities should need to demonstrate success in order to be included in general education. This position fails to recognize the provision’s historical underpinnings as a demonstration of societal values.

Several events preceding IDEA’s passage demonstrated the need for a legislative push toward inclusive education for children with disabilities. Importantly, the Supreme Court’s unanimous decision in Brown vs. Board of Education (1954) set the nation against segregated education for students of color. In the years that followed, the logic of that decision was extended to children with disabilities. Parents advocated for their children’s inclusion, exposés highlighted the mistreatment of children with disabilities in institutions, congressional hearings revealed that millions of children were being excluded entirely or underserved by the public school system, and federal appellate courts ruled that children with disabilities had a right to education under the equal protection clause of the 14th Amendment.

As Martha Minow notes in In Brown’s Wake, the preference for inclusive education within IDEA was an overdue acknowledgement of the societal value of inclusion. Contrary to Gilmour’s concerns about possible adverse effects of inclusion on children without disabilities, Congressman C. D. Daniel, during the debate over IDEA, highlighted the positive spillover effects of educating children with disabilities alongside their peers: “Lessons of patience, understanding, and the ability to provide peer encouragement are just as valuable as traditional educational lessons to the future citizens of this nation.”

The courts have consistently favored inclusive placements and provided more explicit guidance as the law itself has been amended over time, with bipartisan support, to promote more inclusive education. The 1993 appellate court decision, Oberti v. Board of Education, established a notable precedent. In this case, a school district sought to place a young boy with Down syndrome in a segregated placement after he exhibited disruptive behavior. The court wrote that “if placement outside of a regular classroom is necessary for the child to receive educational benefit, the school may still be violating IDEA if it has not made sufficient efforts to include the child in school programs with nondisabled children whenever possible.” To explain its reasoning, it noted that “Congress understood that a fundamental value of the right to public education for children with disabilities is the right to associate with nondisabled peers.”

Most recently, the Supreme Court’s unanimous decision in Endrew F v. Douglas County School District (2017) strongly supported the placement of most students with disabilities in general education classes—while also recognizing that some children with complex disabilities may need more intensive services delivered in other settings. The court recognized that, for the majority of students with disabilities who are primarily in general education classrooms, IEPs should be calculated to enable them to progress from grade-to-grade. The opinion notes, “When a child is fully integrated in the regular classroom, as the Act prefers, what that typically means is providing a level of instruction reasonably calculated to permit advancement through the general curriculum.” This ruling is significant because the court clearly places the burden on the school district to provide appropriate services and accommodations, not the child.

To be sure, these court decisions emphasizing the value of inclusion are statutory in nature, not constitutional. It is plausible to argue, as Gilmour does, that Congress should amend IDEA to place less emphasis on inclusion. As we have demonstrated, however, such a change would mark a sharp departure from the values that informed the law’s enactment and have enabled it to maintain broad bipartisan support for more than four decades. We see no justification for such a change.

Misrepresentations of Students with Disabilities

Gilmour’s article characterizes students with disabilities as a separate class of students who are fundamentally different from students without disabilities and as a result require fundamentally different education and placement. Though many students with disabilities have unique needs that arise from their disabilities, the common assumption that they need to be removed from general education in order for those needs to be met is mistaken.

What distinguishes these students is that they have been found eligible to receive special education services in school. Typically, teachers or parents refer a child for services if he or she is experiencing difficulties in school. After referral, a student is evaluated and a team decides whether the child has one of 13 categories of disability and needs special education services. Importantly, children do not need to be failing in order to be found eligible. Students with disabilities can also be deemed ineligible for special education if the team decides the child does not need special education services. The majority of eligible students are found eligible during their K–12 careers. These eligibility determinations are not always clear cut and certainly do not change who students are.

Several researchers have noted the subjectivity of this evaluation process. Practitioner perceptions and bias, district practices, parental involvement, and the administration of psychological assessments can all influence a team’s eligibility determination. Texas recently made headlines because of a state policy that set a goal for the percentage of students eligible for special education. Following the establishment of this goal, eligibility for special education across the state decreased from 11.6 percent in 2004 to 8.6 percent in 2016. If identification of students with disabilities were objective, a state policy would not have as drastic an impact on enrollment in special education.

As such, it is essential to remember that students with disabilities receiving special education services are also general education students. Assuming that, once identified for special education, students with disabilities become “special education students” and not “general education students” reinforces the notion of separateness. Identification does not absolve the general education teacher of responsibility to educate the child. Other students who receive services and supports, such as low-income students and English learners, do not lose their right to be considered a part of the general education system. Neither should students with disabilities receiving special education services lose their right to be viewed as general education students.

Misrepresentations of Inclusion

Gilmour’s article also misrepresents special education as a fundamentally different kind of education that is fully individualized. She notes, “The team identifies annual goals for the student. These individualized goals determine what constitutes an ‘appropriate education’ for that particular student. Once the goals are in place, the IEP team discusses the instruction, related services, and accommodations the student requires to meet the goals.” Though this description is accurate, Gilmour fails to acknowledge the requirement under IDEA that these goals allow the student to be involved and make progress in the “general education curriculum.” The Department of Education has interpreted this phrase to mean “the curriculum that is based on the State’s academic content standards for the grade in which a child is enrolled.” The individualization comes in determining the particular services, supports, and accommodations the student may need to make progress against those standards.

Gilmour goes on to note, “A key assumption of IDEA is that including SWDs in the regular classroom will expose them to grade-level, general-education curriculum. Yet exposure may not result in progress in that curriculum.” However, the theory is not that exposure will result in progress, but rather that placement in general education, in conjunction with special education services, will result in progress in the curriculum. In fact, as Gilmour acknowledges, a full inclusion placement for a student with a disability is defined as education with students without disabilities for 80 percent or more of the school day. This leaves sufficient time to provide the kinds of effective interventions she cites. Students can and should receive special education services and intensive interventions in the context of inclusive placements. Asking whether students should be included in the general education classroom or receive intensive interventions is to conflate placement with services. This can easily become a justification for placing children in segregated classes for most of the day, a highly questionable practice.

Inclusion and Student Outcomes

To justify the position that inclusion has “gone too far,” Gilmour argues that research on inclusion and student outcomes is “methodologically flawed.” However, this research is not methodologically flawed; rather the research is correlational. It does not claim that inclusion causes improved outcomes, but rather notes that students with disabilities included in general education tend to have better outcomes even when controlling for other student, school, and district characteristics. Just because research is correlational rather than causal does not mean that its findings should be dismissed.

In other fields, correlational evidence has been used to direct policy action. For instance, public health relies heavily on carefully controlled correlational studies because, in many instances, this is the best evidence that can be acquired. Correlational evidence of the connection between smoking and lung cancer (see the Cancer Prevention Study) and heart disease (see the Framingham Heart Study) have appropriately resulted in warning labels on cigarette boxes. Denying the results of large-scale correlational studies can be dangerous.

In research conducted in Massachusetts, one of us (Schifter) examined whether students with disabilities in inclusive placements were more or less likely to graduate within four years of entering high school. The analysis controlled for factors including student disability category, race, income-status, English learner status, gender, percent of low-income students in the district, and percent of students with disabilities in the district. The results showed that the probability that a student with a high-incidence disability (such as a learning disability) in a full inclusion placement graduated on time was about 84 percent, whereas an otherwise similar student in a substantially separate placement had a probability of 43 percent. The differences for students with emotional disabilities were narrower but still substantial: 61 percent vs. 35 percent. Not only are these differences large, but the outcomes for substantially separate settings are undeniably poor—a 43 percent probability for on-time graduation for students with high-incidence disabilities and 35 percent for students with emotional disabilities should not be acceptable.

To minimize the relevance of this kind of research, Gilmour points to “selection bias.” She argues that students are often placed in separate settings because they have more intensive needs than other students with the same disability classification. In her view, relying on correlational evidence to establish benefits from inclusion is problematic because “it is likely that SWDs who would be expected to have better academic and social outcomes are more often included in general-education classrooms than their peers with more-intensive needs.” But this logic assumes that placement decisions are always based objectively on students’ level of need.

In reality, placement decisions are influenced by contextual factors such as teachers’ perception and bias, parental opinion, available programs, district and state policies, and available funding. If placement decisions were simply based on student need, we would not expect to see much variation in placement rates across systems. According to the annual report to Congress, California educates 21.5 percent of students with disabilities in substantially separate classrooms. Florida, on the other hand, educates 13.7 percent in substantially separate classrooms. Variability like this exists across states and districts, within disability categories, and across race, ethnicity, English learner, and income-status. If a student can have a separate placement in one district, move to another district and have an inclusive placement, why should we assume that the large differences in performance researchers have documented are only attributable to differences in student need?

Even if these differences were driven in part by “selection bias” related to student need, given that systemic factors can influence placement decisions, accepting the lower performance in substantially separate classrooms is dangerous. It situates children and their disabilities as the problem and lets the system off the hook for examining systemic inequities, policies, and biases that factor into placement decisions. The policy preference toward inclusive placement is clearly appropriate.

Continue to Emphasize Inclusive Education

Gilmour concludes that policymakers should no longer rely on the share of students with disabilities in inclusive settings when evaluating IDEA compliance, suggesting that doing so may “unduly” influence IEP teams’ decisions. In particular, she notes that educational placement data are broken out by state in the Department of Education’s annual report to Congress, while data on academic outcomes are not. However, the annual report to Congress is merely a report—not an accountability metric. The indicators the Department of Education uses for measuring compliance with IDEA are in fact based on student outcomes, including student test results and graduation data. Achievement data for students with disabilities are also disaggregated at the school level under the requirements of the Every Student Succeeds Act (ESSA). These school-level accountability metrics are more likely to influence IEP team decisions than state-level placement data in a report to Congress.

Inclusive education is also emphasized under the Americans with Disabilities Act (ADA). In 2016, the Department of Justice filed suit against the state of Georgia for operating a segregated program for students with emotional and behavioral disabilities. DOJ noted that under the ADA it is discriminatory to segregate students with disabilities unnecessarily and that the state has an obligation to provide services in the “most integrated setting appropriate to their needs.” Its investigation found that the segregated program provided opportunities to students with disabilities that were “unequal” to other students across the state. These students lacked access to rigorous academics, extracurricular activities (including music, art, and sports), and even decent buildings and facilities.

In short, policymakers must continue to measure and emphasize educational placement under IDEA. As we all know, we measure what we value. Measuring educational placements forces policymakers and practitioners to consider whether students with disabilities are being afforded equal opportunity. If society continues to value inclusivity, we should continue to measure the share of students with disabilities who are being educated in inclusive settings.

The Real Value of an Ecological Perspective

Gilmour contends that future research on inclusion should take an “ecological perspective” that encompasses the impact of inclusion not only on students with disabilities, but also on their peers without disabilities and teachers. Previous research has addressed the impact of including students with disabilities on both teachers and non-disabled peers with generally positive findings. Indeed, qualitative research in schools where inclusive practices have been done well has documented positive impacts on both teachers and non-disabled peers. Gilmour’s claim not only ignores this evidence but also sends an ableist message that children with disabilities are the problem, not their schools. Would anyone suggest that we study the potential negative impact of having low-income children in class on non-low-income peers and teachers—and then use that evidence to question whether low-income children should be included?

Considering the ecology of the classroom can be a useful approach for research if framed an alternative way. How do various approaches to teacher preparation improve outcomes for students in inclusive classrooms? How does principal training and support improve inclusive practices and outcomes for students? Further, we should acknowledge all students learn differently whether or not they have been identified for special education. As such, all teachers teach inclusive classrooms of students with diverse learning needs. Researchers should seek to understand how we can improve inclusive education rather than whether inclusion has gone too far.

Laura A. Schifter is a lecturer on education at the Harvard Graduate School of Education, where Thomas Hehir is a professor of practice.

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