Abstract
Researchers and the U.S. Department of Education regularly report on implementing the Individuals with Disabilities Education Act (IDEA) and the promising trend of increasing proportions of students with disabilities being educated in the general education classroom. Yet, much of this research is limited by an overreliance on aggregated state-level data that can mask implementation problems within states and districts. Using the Texas State Longitudinal Data System, which contains a range of connected student- and school-level data, we overcome these barriers to examine the extent to which students with disabilities across several high-incidence disability categories, student and school demographics, and district locale were provided access to the general education classroom over 22 years. We also identify Texas schools that exclude students with disabilities altogether—something we have yet to see within other empirical studies. Findings from this study add evidence to prior research focused on access and inclusion at the intersections of race and class, as well as important implications for further reauthorizations of IDEA, next-generation studies, and inquiry processes states or districts might consider to increase access to the general education classroom.
Keywords
In many locales before the 1970s, students with disabilities (SWDs) in the United States were denied entry into public schools, provided with limited access, or expelled if school personnel struggled with students (Osgood, 2009). The civil rights movement brought forth a coalition of scholars, educators, advocates, and families seeking to ensure SWDs had rights and could freely access public education. After several federal court decisions (e.g., PARC v Pennsylvania, 1971), Congress enacted the Education for All Handicapped Children Act (EAHCA) of 1975. The landmark legislation (and future reauthorizations as the Individuals with Disabilities Education Act [IDEA]) mandated that all eligible SWDs receive a free appropriate public education (FAPE) in the least restrictive environment (LRE). Researchers and the U.S. Department of Education (ED) regularly report on the law’s implementation, showing increasing proportions of SWDs educated in the general education classroom (e.g., Danielson & Bellamy, 1989; ED, 2024).
Fewer researchers focus on the extent to which students of color with disabilities and those eligible for free and reduced-price lunch (FRPL) are educated in the general education classroom relative to their White and more affluent peers (Cooc, 2022; Grindal et al., 2019; Skiba et al., 2006). These studies point to a concern expressed by the National Council on Disability (NCD, 2018): whether SWDs “participate in their neighborhood school alongside their peers without disabilities is influenced more by the zip code in which they live, their race, and disability level than by meeting the federal law defining how student placements should be made” (p. 9). Limited access to the general education classroom based on race, class, or geographic locale raises concerns about the extent to which placement decisions are centered on the individual child’s needs or broader inequities within states, districts, schools, and society (Anderson & Brock, 2020; Brock & Schaefer, 2015; Westling, 2019). Yet, as Cooc (2022) pointed out, the current literature is limited by an overreliance on aggregated state-level data from annual federal reports that cannot disentangle biases and constraints within schools and districts from differences in student learning needs.
Using the Texas State Longitudinal Data System (SLDS), which contains a range of connected student- and school-level data, we overcome these barriers to descriptively analyze the extent to which SWDs across several high-incidence disability categories, student and school demographics, and district locale were provided access to the general education classroom over 22 years. In addition, as part of this descriptive analysis, we identify Texas schools that exclude SWDs altogether—something we have yet to see in other empirical studies. Good descriptive analysis requires serious thought and expertise, as Loeb and colleagues (2017) noted, and can stand on its own as impactful and relevant research if it can “improve understanding about important phenomena” (p. v). Findings from this study add to prior research focused on inclusion at the intersections of race, class, and locale offering implications for future IDEA reauthorizations, next-generation studies, and inquiry processes states or districts might consider to increase access to the general education classroom. In what follows, we situate our study with an overview of IDEA in recognition that the LRE principle and notions of inclusion are contested (e.g., Annamma et al., 2013; Kauffman et al., 2023). We also review literature on LRE trends and provide a policy scan of special education in Texas. Next, we describe our methods, data sources, and analytical procedures. Then, we present our findings—including that student demographics and locale influence the amount of time SWDs receive in the general education classroom. Finally, we conclude with implications for policy and future research.
Literature Review
IDEA and LRE Legal Background
The core principles of the EAHCA have remained consistent since 1975 and through IDEA reauthorizations beginning in 1990. Many terms are defined in law and have been shaped by litigation and judicial interpretation, allowing states and districts flexibility in implementation. A key term, FAPE, requires that each SWD receive “a free appropriate public education [FAPE] that emphasizes special education and related services designed to meet their unique needs, and prepare them for further education, employment, and independent living” (IDEA, 20 U.S.C. §1400(d)(1)(A)). FAPE includes services “provided at the public expense, under public supervision and direction, and without charge,” aligned with SEA standards and “provided in conformity with the individualized education program” (IDEA, 20 U.S.C. §1401a). “Special education” is defined as “specially designed instruction, at no cost to the parents . . . conducted in the classroom, in the home, in hospitals and institutions, and other settings” (IDEA, 20 U.S.C. §1401(29)(A)). LRE-related decisions have often focused on the extent to which a child can “benefit educationally” from instruction (Board of Education v. Rowley, 1982). In Endrew F., the U.S. Supreme Court clarified that the benefit must be “markedly more demanding than merely de minimis” (p. 15), requiring schools to offer an IEP “reasonably calculated to enable a child to make progress appropriate in light of the child’s circumstances” (p. 11).
Of importance for this study is IDEA’s definition of special education along with the Endrew F. decision, which makes clear that a continuum of placements must be made available: To the maximum extent appropriate, children with disabilities, including children in public or private institutions or other care facilities, are educated with children who are not disabled, and special classes, separate schooling, or other removal of children with disabilities from the regular educational environment occurs only when the nature or severity of the disability of a child is such that education in regular classes with the use of supplementary aids and services cannot be achieved satisfactorily. (IDEA, 2004, sec. 1412(a)(5)(A))
Individuals with Disabilities Education Act does not define “placement,” but the ED has expanded on the term: the determination of the educational placement . . . must be based on a child’s IEP. The Department’s longstanding position is that placement refers to the provision of special education and related services rather than a specific place, such as specific classroom or school. (IDEA regulations, 2006, p. 46687)
The Endrew F. decision did not specifically address LRE or placement, but the ED posted guidance following the ruling: Consistent with the decision in Endrew F., the [U.S.] Department [of Education] continues to recognize that it is essential to make individualized determinations about what constitutes appropriate instruction and services for each child with a disability and the placement in which that instruction and those services can be provided for the child. There is no “one-size-fits-all” approach to educating children with disabilities. Rather, placement decisions must be individualized and made consistent with a child’s IEP. (ED, 2017)
A district does not have to have a full continuum within their schools but must ensure options are available outside of the district, which might include contractual arrangements with other entities. In sum, placement decisions must be individualized and based on a student’s IEP, not their race, economic status, or geographic locale. In a Dear Colleague letter, the ED’s Office of Civil Rights (2016) affirmed that “Districts must give students equitable access, without regard to race, color, or national origin to the most integrated setting appropriate for the student” (pp. 1–2). The ED monitors IDEA implementation to ensure these standards, a topic we turn to next.
IDEA Reporting, Accountability, and LRE Data
The ED has submitted annual reports to Congress on the progress of implementing EAHCA/IDEA since the late 1970s. In the report’s current form, the ED (2024) describes, our nation’s progress in (1) providing a free appropriate public education (FAPE) for children with disabilities . . . (2) ensuring that the rights of these children with disabilities and their parents are protected; (3) assisting States and localities in providing IDEA services . . . and (4) assessing the effectiveness of efforts . . . (p. xv)
The report draws data from all states and territories and includes a description of the percentage of students ages 5 through 21 served under IDEA by educational environment under the following categories: inside the regular class for 80% or more of the day; 40% through 79% of the day; and less than 40% of the day; as well as in a separate school, residential facility, homebound/hospital, correctional facilities, and parentally placed in private schools. Each report illustrates the variation in the settings in which students are educated not only by disability classification (e.g., intellectual disability, specific learning disability) but also by state, highlighting the need to look closer at individual states.
Across all disabilities, 66.7% of SWDs receive 80% or more of the day in the regular class, although some states have higher (e.g., Alabama, 83.8; Colorado, 79.4; Nebraska, 80.9%) and lower rates (e.g., New Jersey, 44.2%; Maine, 55.3%; Montana, 56.1%) of SWDs being educated in this setting (ED, 2024). The report includes educational environments for each IDEA disability classifications. Each group varies in their participation in the general education classrooms and the percentage of SWDs educated in separate schools (All states, 10.8%; Connecticut, 28.11%; Nevada, 4.3%). The report does not examine variation within states or differences across locales (e.g., rural, urban, town, suburban).
IDEA (2004) was aligned with the No Child Left Behind Act’s (2002) principles of accountability, requiring states to develop a State Performance Plan (SPP) and Annual Performance Report (APR) to evaluate IDEA implementation across 20 indicators. Indicators cover IDEA compliance and results areas, such as graduation and dropout (Indicators 1 and 2), student achievement (Indicator 3B), and suspension/expulsion (Indicator 4). Indicator 5 focuses on “Educational Environments” and the percentage of SWDs in kindergarten and ages 6 through 21 years old served: inside the regular class 80% or more of the day; inside the regular class less than 40% of the day; and in separate schools, residential facilities, or homebound/hospital placements. States set and work toward annual goals for each indicator. The ED uses SPP/APR and public meetings to determine if the state meets IDEA requirements annually, needs assistance, needs intervention, or needs substantial intervention. To improve results, the ED provides financial support and technical assistance to states and districts.
Most research on LRE trends relies on data from the ED’s annual reports to Congress and other publicly available datasets. For example, reporting on the percentage of SWDs educated in “regular schools” and “segregated facilities” between 1976–1977 and 1985–1986, Danielson and Bellamy (1989) found that just 26.5% of SWDs were educated in the regular classroom environment for 80% or more of the school day. They also found noteworthy variation by state. In preceding decades, similar studies relied on the ED’s annual reports to document a common set of patterns: more SWDs were spending 80% or more of the day in the general education classroom, although state-by-state differences remained apparent (McLeskey et al., 2004, 2011, 2012; Sawyer et al., 1994). More recently, Williamson et al. (2020) found that SWDs across all categories were increasingly educated 80% or more in the general education classroom between 1990 and 2015. In 1990, just 33.91% of SWDs were included in the regular classroom for 80% of the day or more, compared with 71.56% in 2015. Students with specific learning disabilities saw a substantial decrease in classroom removal for special education but less so for other disabilities, including autism and emotional disturbance.
A few studies using varied datasets have raised questions about the extent to which students with more significant disabilities have gained access to the general education classroom (Brock, 2018; Morningstar et al., 2017), yet little is known which student demographics and geographic locations enhance or constrain such trends. Moreover, few studies describe racial inequities in access to the general education classroom using student-level data (e.g., Cooc, 2022; Grindal et al., 2019; Skiba et al., 2006). Researchers also noted additional concerns, including data quality issues with state-reported data and the limitations of relying on ED’s annual reports for analysis purposes (Williamson et al., 2020). The lack of in-depth analysis within states, locales, and districts is concerning, especially given concerns about how racism and other forms of marginalization manifest across various public sectors (e.g., healthcare, education, housing; Phelan & Link, 2015), locales (Schell et al., 2020), and in education policies and practices (Sullivan & Bal, 2013).
Texas Special Education
Texas has unique IDEA implementation problems worth noting before considering our findings. The most notable problem was the state’s arbitrary “special education cap” which reshaped the population of students served under IDEA. In 2004, special education enrollment was 11.6% of the student population—similar to the national average. That same year, Texas began implementing the Performance-Based Monitoring Analysis System (PBMAS), an annual accountability system tracking the performance of districts and charter schools in several areas. Indicator 10 awarded districts a perfect performance score if fewer than 8.5% of students received special education (other cut-offs were 11% and 15%). By 2017, Texas identified 8.5% of the population under IDEA, with almost every district meeting the 8.5% indicator (DeMatthews & Knight, 2019). An ED (2018) investigation concluded the state failed to implement and monitor IDEA in several ways, noting: “TEA failed to ensure that all children with disabilities . . . were identified, located, and evaluated” for special education as appropriate (p. 4). TEA submitted a Corrective Action Response (CAR) in 2018 outlining a series of actions. In 2020 and 2021, the ED (2023) concluded that TEA had not taken “the necessary actions to correct all previously identified noncompliance” (p. 1). Under SPP/APR, Texas was also rated as “Needs Improvement” or “Needs Assistance” through the tenure of the current governor and TEA commissioner (2015–2023). The sum of these problems partly shaped the population of students served under IDEA.
Method
The following research questions are addressed within this study:
To answer these questions, we utilize data from the Texas SLDS, which provides comprehensive student- and school-level data from all Texas public K–12 schools (traditional and public charter), spanning the school years from 2001–2002 to 2022–2023. Texas provides an essential context for this study with nearly 1,200 districts, 9,000 schools, and 5.4 million students from diverse socioeconomic backgrounds. To examine the distribution across district locales (rural, urban, town, suburban), we incorporated locale data from the National Center for Education Statistics (NCES). The SLDS includes various connected datasets allowing a detailed descriptive analysis of educational placements. The dataset consists of a range of variables, including disability categories, educational environments for special education students, student demographics, school characteristics, and district locales.
Sample
Table 1 provides summary statistics for the sample of 10,483,680 student-year observations from 2002 to 2023. These students are all SWDs identified under IDEA from kindergarten through 12th grade in Texas public schools, categorized by their disability types, race and ethnicity, socioeconomic status, and locale. Approximately 5.661% of students with disabilities’ records in the SLDS did not match enrollment records, reflecting minor data incompleteness that does not substantially affect our main findings but may influence analyses of specific subpopulations. Regarding disability types, 41.92% had a learning disability, 12.94% had other health impairments, 8.78% had autism, 8.69% had an intellectual disability, and 6.74% had emotional disturbances. In terms of race and ethnicity, 46.00% were Hispanic, 32.63% were White, and 16.82% were Black student-year observations. In the overall sample, 64.79% were eligible for FRPL. Regarding locale, 32.86% were in a city, 23.71% were in a suburban area, 12.08% were in a rural area, and 8.52% were in a town.
LRE and Demographic Summary for Students With Disabilities, 2002–2023.
Note. Data represents student-year observations from 2002 to 2023. A student may appear multiple times across years, reflecting changes in disability status or placement. OHI = other health impairment; ID = intellectual disability; ED = emotional disturbance; LD = specific learning disability; Non-instruct. refers to students only receiving non-instructional services (e.g., speech therapy); Non-FRL refers to students not receiving Free or Reduced Priced Lunch; City, Suburban, Town, and Rural refer to NCES locales from 2007 to 2023.
The SLDS includes seven primary groups for time SWDs spend in the general education classroom in accordance with their IEP, which are more specific than federal reporting. Each group corresponds to a single setting code in the SLDS, except for group 6, which combines multiple codes. The groups represent the following:
(1) 100% of the instructional day is spent in the general education classroom;
(2) 80% to 99% of the instructional day spent in the general education classroom;
(3) 51% to 79% of the instructional day spent in the general education classroom;
(4) 40% to 50% in the general education classroom;
(5) Less than 40% of the instructional day spent in the general education classroom;
(6) Educated entirely in other settings (e.g., homebound, residential facilities1;
(7) Only receives non-instructional services (e.g., speech therapy).
For students receiving only non-instructional services (Group 7), there were 15.88% of student-year observations. To prevent distortion in the inclusion analysis, we excluded this group in our main analysis with the Inclusivity Measure because adding a student who only needs non-instructional services such as speech therapy can easily inflate the inclusivity.
Analytical Strategy
To illustrate the general trend of SWDs in different educational environments, we use a stacked area chart from 2001–2002 to 2022–2023. This chart shows the total number of students with disabilities each year, disaggregated by setting group. For the remainder of the analysis, we propose and utilize an “Inclusivity Measure” to quantify the degree of inclusion of SWDs within different educational environments. The Inclusivity Measure is a continuous variable representing the average percentage of time students with disabilities spend in general education settings. It is calculated as a weighted average of midpoints from five categorical placement ranges weighted by the proportion of students in each category (see Table 2). Sensitivity analyses using alternative value assignments, such as maximum and minimum values per category, confirmed generally consistent results. Many studies that examine inclusivity either focus solely on the proportion of SWDs included in general education classrooms for 80% or more of the day or provide general proportions across three broad settings (80% or more, 40%–79%, and less than 40% of the day; e.g., McLeskey et al., 2012; Williamson et al., 2020). These approaches offer only a basic distribution without creating a specific inclusivity measure.
The Ratio of Instructional Day in a General Education Setting for Each Setting Code.
Note. The Inclusivity Measure is a continuous variable representing the average percentage of time students with disabilities spend in general education settings. It is calculated as a weighted average of midpoints from five categorical placement ranges weighted by the proportion of students in each category using assigned midpoint values as in Table 2. For example, if the proportions of students in a school across the five groups (100%, 80%–99%, 51%–79%, 40%–50%, and less than 40%) are a, b, c, d, and e, respectively, the Inclusivity Measure (IM) is calculated as: IM = (100a + 90b + 65c + 45d + 20e) / (a + b + c + d + e). For a school with proportions a = 0.2, b = 0.3, c = 0.3, d = 0.1, and e = 0.1, the Inclusivity Measure would be: IM = (100 × 0.2 + 90 × 0.3 + 65 × 0.3 + 45 × 0.1 + 20 × 0.1)/(0.2 + 0.3 + 0.3 + 0.1 + 0.1) = (20 + 27 + 19.5 + 4.5 + 2)/1 = 73.
In contrast, our “Inclusivity Measure” provides a comprehensive view by calculating the proportion of SWDs across five groups (100%, 80%–99%, 51%–79%, 40%–50%, and less than 40% inclusion in general education), allowing for more precise comparisons across different groups and tracking trends over time. By encompassing all groups except 6 and 7, we achieve a balanced and accurate measure that reflects changes across all inclusion settings, avoiding the limitations of focusing solely on high-inclusion groups. In states like Texas, where inclusion data is detailed across five setting groups, this approach offers a distinct advantage by capturing a fuller picture of inclusivity. To validate group differences observed in the Inclusivity Measure across locale, race/ethnicity, and socioeconomic status, we employed non-parametric tests (Mann–Whitney U for two-group comparisons, Kruskal–Wallis for multi-group comparisons) due to its non-normal distribution (Shapiro–Wilk, p < .00001). These tests, supplemented by 95% confidence intervals (CIs) and exploratory analyses (t-tests, linear regression with robust standard errors), confirmed the robustness of differences reported in the text. We also utilized a line graph to depict the exclusion rates of SWDs across various disability categories. We linked enrollment records of all SWDs to their respective schools, allowing us to identify schools with no registered SWDs. For each specific disability category, we calculated the proportion of schools that completely excluded these students. This detailed data allows us to measure inclusivity more accurately and to analyze and compare time trends across various disabilities, races, socioeconomic statuses (SES), and locales.
Results
Texas Placement Trends for Students With Disabilities
Figure 1 provides the number of SWDs and their placement (see Figure 1A), and the ratios of those placements (see Figure 1B), to show trends between 2001-02 and 2022-232. Over the study period, the number of SWDs in Texas grew from 468,163 to 666,547. During this period, due to the state’s illegal special education cap, both the number and percentage of SWDs saw notable fluctuations. The number of identified students went from 468,163 in 2001–2002, decreasing to a low of 411,446 in 2011–2012, before rising again to 666,547 by 2022–2023. Similarly, the percentage of SWDs identified shifted from 11.74% in 2001–2002 to a low of 8.5% in 2013–2014, when the number was at 416,595, eventually reaching 12.70% in 2022–2023 (see Figure 1A). SWDs included in the general education classroom for 100% of the school day increased from 30.51% in 2001–2002 to 48.36% in 2022–20233. Likewise, the percentage of SWDs who spent at least 50.1% to 79.9% of the instructional school day in the general education classroom decreased from 24.55% in 2001–2002 to 10.04% in 2022–2023. The percentage of SWDs in other setting groups remained relatively stable over the study period (see Figure 1B).
Number and Proportion of Students With Disabilities by Setting.
Figure 2 presents the inclusivity measure for all SWDs across our preferred settings. The measure reflects the percentage of the school day that an average SWDs in these settings spends in a general education setting. In 2002, on average, 72.07% of instructional time for SWDs was spent in general education settings. From 2004 to 2009, there was a steady increase in time spent in the general education setting, with the proportion rising from 71.42% to 77.76%. However, from 2010 to 2014, we found a decline, with percentages fluctuating between 75.93% and 77.57%. Since 2015, we found a resurgence, with percentages gradually climbing back up and reaching 79.07% in 2023.
Statewide Inclusivity.
We also looked across disability categories to understand the extent to which SWDs were served across educational environments. Figure 3 presents the inclusivity measure for four disability categories (intellectual disability, emotional disturbance, specific learning disability, and autism). This figure displays the trends in inclusivity measures for each disability category, using dashed lines for students identified with one IDEA disability, short-dashed lines for students with more than one IDEA disability, and solid lines for the total, which includes students with one disability and students with multiple disabilities. Students with multiple disabilities exhibit lower inclusivity likely due to increased support needs. By examining the proximity of the “total” line to either the single or multiple lines, one can infer which group is more prevalent for each category. For example, in the case of specific learning disability, single disability cases are more common, whereas for autism, multiple disability cases are more common. Specific learning disability also has the highest inclusivity measure in the most recent year (92.28%), whereas intellectual disability has the lowest (48.2%).
Inclusivity Measure by Disability Classification.
Variation in Placement by Student and School Demographics and Locale
Figures 4, 5, and 6 display trends in placement by district context. Figure 4 illustrates the inclusivity measure in schools categorized as city, suburban, town, and rural. The measure is consistently highest in rural schools. For example, in the 2022–2023 school year, the inclusivity measure was 82.43% in rural schools compared with 77.99%, 78.7%, and 78.95% in city, suburban, and town schools, respectively. Town schools had the next highest measure until the late 2010s, but after that, the measure is similar across suburban, town, and city schools. Figure 5 follows a similar approach, examining inclusivity measure trends in affluent schools with a low proportion of students eligible for FRPL and economically disadvantaged schools with a high proportion of students eligible for FRPL. We sorted all schools based on the proportion of students eligible for FRPL and chose the top and bottom 25% of schools with the most and least proportions of students eligible for FRPL for analysis. The inclusivity measure is consistently higher in affluent schools. Both groups show similar trends: a slight decrease in the early years, a steep increase over 5 years, followed by slight decreases and increases. However, in recent years, the trends have increasingly diverged. Economically disadvantaged schools show little change, while affluent schools exhibit rapid growth. The difference in the inclusivity measure between the two groups narrowed from 4.99% point in 2001–2002 to 1.29% point in 2014–2015 but increased again to 5.19% point by 2022–2023.
Inclusivity Measure by District Locale in Texas.
Inclusivity Measure by District Affluence/Poverty in Texas.
Inclusivity Measure by District Student Demographics in Texas.
Figure 6 displays the trends in inclusivity measures for schools with a high proportion of White, Hispanic, and Black students. We sorted schools based on the proportions of these racial groups and chose the top 25% of schools with the highest proportions of White, Hispanic, and Black students for analysis. A school could appear in the top 25% for more than one racial group if it had high proportions of those groups. The inclusivity measure is consistently higher in majority White schools, lower in majority Black schools, and moderate in majority Hispanic schools. The patterns reveal a similar trend across the three racial/ethnic groups, but there has been a noteworthy convergence between Hispanic and Black-majority schools over the years. For example, in the 2001–2002 school year, the inclusivity measures for White, Hispanic, and Black-majority schools were 74.96, 71.31, and 68.84, respectively. However, by 2022–2023, these figures had shifted to 81.17, 77.34, and 77.27, respectively.
Figure 7 compares the inclusivity measure for Black, White, and Hispanic students across different disability categories. In most of the analyzed disability categories, White students have the highest inclusivity, whereas Black students have the lowest measure, with Hispanic students falling between the two groups. However, in the case of intellectual disability, Black students have the highest inclusivity. Recent years have seen a narrowing of the gap, particularly in categories such as learning disability, where there is not much difference across races. For example, the inclusivity measures for learning disability in 2022–2023 were 91.5, 92.73, and 92.29 for Black, White, and Hispanic students, respectively. Nonetheless, disparities persist in categories such as autism, intellectual disability, and emotional disturbance. In 2022–2023, the inclusivity measures for autism remained at 56.45, 68.47, and 60.13 for Black, White, and Hispanic students, respectively. For intellectual disability, the measures were 52.77, 44.74, and 47.57, while for emotional disturbance, they were 77.76, 83.74, and 82.85 for Black, White, and Hispanic students, respectively.
Inclusivity by Race and Disability Category.
Figure 8 compares the inclusivity measure for students with three SES statuses across disability categories. In most cases, students receiving FRPL had lower inclusivity. However, there is a noticeable upward trend in inclusivity measures across all disability categories. Despite this trend, noteworthy differences persist, particularly in the case of autism, where disparities across SES levels remain pronounced. In addition, in the case of intellectual disability, the most economically disadvantaged students have the highest inclusivity. For instance, in 2022–2023, the inclusivity measures for autism were 58.21 for students receiving Free Lunch, 61.73 for those receiving Reduced Price Lunch, and 66.75 for students not eligible for either. On the other hand, for intellectual disability, the measures were 49.48 for students receiving Free Lunch, 47.42 for those receiving Reduced Price Lunch, and 43.04 for students not eligible for either.
Inclusivity by Economic Status and Disability Category.
Texas Public Schools Without Students With Disabilities/Disability Types
Some Texas schools do not serve any students with certain disability types. Figure 9 provides the ratio of Texas schools that do not serve any students with the following disabilities: other health impairments, emotional disturbance, autism, intellectual disability, and specific learning disability. Students with autism were not educated in nearly 60% of all Texas public schools in 2002–2003, but we found a rapid decrease in the number of schools that do not educate students identified with autism. One reason may be the increased identification of students with autism during the study period (2001–2002: 8,038; 2022–2023: 100,573). Although changes have not been as rapid, more Texas schools are educating students with intellectual disabilities and other health impairments than ever before (ID: 2001–2002: 34.39%, 2022–2023: 15.87%; OHI: 2001–2002: 18.05%, 2022–2023: 7.64%). The percentage of Texas schools not enrolling students with specific learning disabilities has remained relatively stable (SLD: 2001–2002: 5.65%, 2022–2023: 4.78%). The percentage of Texas schools not enrolling students with emotional disturbance shows considerable fluctuations over the years without a clear linear trend. Starting at 26.69% in 2002, the percentages have varied annually, with notable highs and lows. For instance, the rate declined to 24.93% in 2006, rose to 33.33% in 2012 before declining to 21.93% in the most recent year.
Ratio of Texas Public Schools Without Students With Disabilities by Disability Type.
Discussion
In this study, we highlight the trends and variability of access to the general education classroom over 22 years within one state and, in doing so, make several contributions to existing research. First, our work aligns and adds nuance to prior research documenting that SWDs are increasingly educated in the general education classroom for 80% or more of the day (e.g., ED, 2024; Williamson et al., 2020). In the ED’s (2024) most recent report, 72.6% of SWDs were educated for 80% or more in the general education classroom in Fall 2021 (12.7% for 40% through 79%, 13.5% for less than 40%, and 1.2 in other settings). When using our inclusivity measure, we find that, on average, SWDs have greater access to the general education classroom. In 2002, on average, 72.07% of instructional time for SWDs was spent in the general education setting, increasing to 79.07% by 2022–2023. Like other studies (e.g., ED, 2024), we also find that access to the general education classroom increased for all disability categories with variations across disabilities (e.g., lower inclusivity rates for students with intellectual disabilities) and limited changes in access for other settings (e.g., residential, homebound).
Shifting away from state-level aggregate data and relying on longitudinal student-level data, we overcome prior methodological limitations to document less understood trends at the intersections of race, disability type, geographic locale, and school demographics. Similar to a small group of studies that include locale data (e.g., Brock & Schaefer, 2015), we find that access to the general education classroom differs by locale and by the percentage of affluent students and students of color attending a school, consistent with the NCD’s (2018) claim that participation in a neighborhood school and among peers in the general education classrooms is influenced “by the zip code in which they live, their race, and disability level” (p. 9). For example, the inclusivity measure for school year 2022–2023 was 82.43% in rural schools compared with 77.99%, 78.7%, and 78.95% in city, suburban, and town schools. Inclusivity is consistently higher for all SWDs (and disability categories we previously identified) in schools with lower ratios of students receiving FRPL, as well as in schools that consistently have a majority of White students.
We also find that Black students had the lowest inclusivity in comparison to White and Hispanic students with specific learning disabilities, emotional disturbance, and autism, although Black students classified with an intellectual disability had higher inclusivity than their peers. These findings raise concern about the extent to which Black students are appropriately identified and placed in special education as well as the levels of access to the general education classroom they are provided (Graves & Ye, 2017; Reid & Knight, 2006). Differing access rates to the general education classroom were found based on a student’s economic status. In most cases, economically disadvantaged students had lower inclusivity. These findings are consistent with previous research documenting lower levels of access to the general education classroom for Black students and students eligible for free and reduced-price lunch (Cooc, 2022; Grindal et al., 2019; Skiba et al., 2006) and extend this work by highlighting how schools with higher proportions of Black students and low-income students are less likely to be inclusive. However, we did find a narrowing of the gap over time despite persistent disparities for Black students with autism, intellectual disabilities, and emotional disturbance. We also saw a clear upward trend in our inclusivity measure across all contexts and disability classifications over the study period.
We are unaware of studies examining the extent to which SWDs are entirely excluded from public schools, which we believe is partly due to an overreliance on state-level aggregate data. This is a serious limitation given the LRE principle puts forward a strong preference for access to the general education classroom and the neighborhood school. We document a precedent of Texas schools excluding SWDs, particularly certain disability classifications (e.g., emotional disturbance, other health impairment, intellectual disability). In the 2022–2023 school year, 4.78% of Texas public schools did not educate a child with a specific learning disability, and 34.49% did not serve a child with an intellectual disability. The total exclusion of students with certain disabilities from certain Texas schools is concerning and likely has a myriad of causes, including district preferences for providing transportation services to district-organized self-contained programs for students with certain disability types or needs and public charter schools that are not adhering to IDEA’s Child Find requirements or investing in special education and related services to educate students with certain needs. Moreover, the Texas SLDS did not have a data point for marking a child’s assigned neighborhood school like some other states (e.g., California), so there are valid reasons for some schools to have few or no students with certain disability categories.
The extent to which SWDs are provided with access to the general education classroom will require further investigation, including more advanced studies to explore trends in LRE and identify possible statistical differences. Additional studies relying on longitudinal datasets with student-level data can provide important insights into trends and disparities in general education access across and within states. Research is needed to further investigate the placement of SWDs in the general education classroom and across the LRE continuum with attention to school, district, and regional context and in relation to community demographics, district priorities, state oversight and accountability mandates, and school finance policies. Further investigation is warranted to understand why some Texas schools do not serve any SWDs. Several other questions warrant further investigation and require quantitative, qualitative, and mixed methods approaches, given the complexity of providing SWDs with access to the general education classroom. First, what are the trends for access to the general education classroom across all states, across different locales, and similar districts? Second, within states and districts that have made a great deal of progress or limited progress with educating SWDs in the general education classroom, what are the barriers, opportunities, and potential pathways for increasing access to the general education classroom? Rather than fixating on problems or celebrating successes, such investigations can offer strategies for states and districts. Future studies in other states can provide important insights, considering Texas is an outlier given serious implementation concerns. Third, to what extent are SWDs included in the general education classroom meeting their IEP goals, meeting academic standards, and developing a broad range of skills associated with post-secondary transitions? We raise this question because we are weary of celebrating an increase in access to the general education classroom without information to support SWDs’ academic and social benefit.
Research is also needed to understand contexts that provide greater access to the general education classroom as well as a high-quality IEP and education leading to positive experiences and outcomes. Fourth, to what extent are traditional public schools and public charter schools provide for SWDs access to the general education classroom? This study did not separate traditional public schools from public charter schools but concerns with charter school enrollment and discriminatory practices warrant further attention (e.g., Bowman et al., 2024; Government Accountability Office, 2012). Relatedly, researchers might examine variations across districts within a state and regions given the state/district flexibility to make decisions about IDEA implementation. This study also has implications for policymakers, especially if IDEA is reauthorized and the ED reconsiders aspects of SPP/APR and its role in providing monitoring and technical assistance. Current IDEA Indicators and the SPP/APR do not sufficiently disaggregate data and consequently provide only limited insight into the extent to which all students, regardless of race or class, are being well served under IDEA. Our study clarifies that state-level trends can be positive, but serious inequities can persist and not be raised to attention for monitoring and improvement purposes. Revised regulations can require that states disaggregate data further, follow an inquiry process to interrogate IDEA data, or partner with researchers to explore IDEA implementation and set more nuanced goals. Trainings for districts to improve IEP meeting facilitation and placement decisions might help ensure students are well served under IDEA. Such efforts can strengthen SEA’s abilities to support districts and address inequities.
Limitations
This study has several limitations. The SLDS provides category-based placement data (e.g., 80%–99.9% inclusion) without detailed information on disability severity or exact inclusion percentages, limiting the precision of our Inclusivity Measure. To avoid double counting, we classified students by their first-listed disability, assuming it represents the primary concern, which may not always hold true or may change over time. Although the dataset includes a charter school indicator, we did not separately analyze charter school data, as our focus was on statewide inclusivity trends across all public schools. Future research could examine differences in inclusivity between charter and traditional public schools to explore variations in placement practices. Finally, our descriptive analysis, while suitable for identifying broad trends, could be extended through advanced statistical modeling in future studies to investigate these nuances further.
Conclusion
For far too long, researchers and policymakers have ignored the nuances of providing access to the general education classroom or gazed at this problem from afar without offering any practical insights for improvement. We hope our study and findings provokes action through more closely monitoring state implementation of IDEA through longitudinal analyses using student-level data and additional research into states, districts, and schools that are finding ways to disrupt persistent inequities and providing a high-quality education to SWDs in their neighborhood schools and in the general education classroom. We also hope quantitative and qualitative researchers work together and with SEAs and districts to explore IDEA implementation.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work was supported by James R. Yates and Alba A. Ortiz Endowed Excellence Fund for Research and Training of Public School Leadership at the University of Texas at Austin.
